Understanding space weather to shield society: A global ... · Earth’s magnetic ﬁeld and upper atmosphere. This results in a variety of manifestations, including geomagnetic vari-ability,

Available online at www.sciencedirect.com

www.elsevier.com/locate/asr

ScienceDirect

Advances in Space Research 55 (2015) 2745–2807

Understanding space weather to shield society: A global road mapfor 2015–2025 commissioned by COSPAR and ILWS

Carolus J. Schrijver a,⇑, Kirsti Kauristie b,*, Alan D. Aylward c, Clezio M. Denardini d,Sarah E. Gibson e, Alexi Glover f, Nat Gopalswamy g, Manuel Grande h, Mike Hapgood i,

Daniel Heynderickx j, Norbert Jakowski k, Vladimir V. Kalegaev l, Giovanni Lapenta m,Jon A. Linker n, Siqing Liu o, Cristina H. Mandrini p, Ian R. Mann q, Tsutomu Nagatsuma r,

Dibyendu Nandy s, Takahiro Obara t, T. Paul O’Brien u, Terrance Onsager v,Hermann J. Opgenoorth w, Michael Terkildsen x, Cesar E. Valladares y, Nicole Vilmer z

a Lockheed Martin Solar and Astrophysics Laboratory, 3251 Hanover Street, Palo Alto, CA 94304, USAb Finnish Meteorological Institute, FI-00560, Helsinki, Finland

c University College London, Dept. of Physics and Astronomy, Gower Street, London WC1E 6BT, UKd Instituto Nacional de Pesquisas Espaciais, S. J. Campos, SP, Brazil

e HAO/NCAR, P.O. Box 3000, Boulder, CO 80307-3000, USAf RHEA System and ESA SSA Programme Office, 64293 Darmstadt, Germany

g NASA Goddard Space Flight Center, Greenbelt, MD, USAh Univ. of Aberystwyth, Penglais STY23 3B, UK

i RAL Space and STFC Rutherford Appleton Laboratory, Harwell Oxford, Didcot OX11 0QX, UKj DH Consultancy BVBA, Diestsestraat 133/3, 3000 Leuven, Belgium

k German Aerospace Center, Kalkhorstweg 53, 17235 Neustrelitz, Germanyl Skobeltsyn Institute of Nuclear Physics, Lomonosov Moscow State University, Moscow 119991, Russian Federation

m KU Leuven, Celestijnenlaan 200B, Leuven 3001, Belgiumn Predictive Science Inc., San Diego, CA 92121, USA

o National Space Science Center, Chinese Academy of Sciences, Haidian District, Beijing 100190, Chinap Instituto de Astronomia y Fisica del Espacio, C1428ZAA Buenos Aires, Argentina

q Dept. of Physics, Univ. Alberta, Edmonton, AB T6G 2J1, Canadar Space Weather and Environment Informatics Lab., National Inst. of Information and Communications Techn., Tokyo 184-8795, Japan

s Center for Excellence in Space Sciences and Indian Institute of Science, Education and Research, Kolkata, Mohanpur 74125, Indiat Planetary Plasma and Atmospheric Research Center, Tohoku University, 6-3 Aoba, Aramaki, Aoba, Sendai 980-8578, Japan

u Space Science Department/Chantilly, Aerospace Corporation, Chantilly, VA 20151, USAv NOAA Space Weather Prediction Center, Boulder CO 80305, USA

w Swedish Institute of Space Physics, 75121 Uppsala, Swedenx Space Weather Services, Bureau of Meteorology, Surry Hills NSW, Australia

y Institute for Scientific Research, Boston College, Newton, MA 02459, USAz LESIA, Observatoire de Paris, CNRS, UPMC, Universite Paris-Diderot, 5 place Jules Janssen, 92195 Meudon, France

Received 1 December 2014; received in revised form 3 March 2015; accepted 18 March 2015Available online 3 April 2015

Abstract

There is a growing appreciation that the environmental conditions that we call space weather impact the technological infrastructurethat powers the coupled economies around the world. With that comes the need to better shield society against space weather by

http://dx.doi.org/10.1016/j.asr.2015.03.023

0273-1177/� 2015 COSPAR. Published by Elsevier Ltd.

This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

⇑ Corresponding authors.E-mail address: [email protected] (C.J. Schrijver).


http://creativecommons.org/licenses/by-nc-nd/4.0/

mailto:[email protected]


http://crossmark.crossref.org/dialog/?doi=10.1016/j.asr.2015.03.023&domain=pdf

2746 C.J. Schrijver et al. / Advances in Space Research 55 (2015) 2745–2807

improving forecasts, environmental specifications, and infrastructure design. We recognize that much progress has been made and con-tinues to be made with a powerful suite of research observatories on the ground and in space, forming the basis of a Sun–Earth systemobservatory. But the domain of space weather is vast – extending from deep within the Sun to far outside the planetary orbits – and thephysics complex – including couplings between various types of physical processes that link scales and domains from the microscopic tolarge parts of the solar system. Consequently, advanced understanding of space weather requires a coordinated international approach toeffectively provide awareness of the processes within the Sun–Earth system through observation-driven models. This roadmap prioritizesthe scientific focus areas and research infrastructure that are needed to significantly advance our understanding of space weather of allintensities and of its implications for society. Advancement of the existing system observatory through the addition of small to moderatestate-of-the-art capabilities designed to fill observational gaps will enable significant advances. Such a strategy requires urgent action: keyinstrumentation needs to be sustained, and action needs to be taken before core capabilities are lost in the aging ensemble. We recom-mend advances through priority focus (1) on observation-based modeling throughout the Sun–Earth system, (2) on forecasts more than12 h ahead of the magnetic structure of incoming coronal mass ejections, (3) on understanding the geospace response to variable solar-wind stresses that lead to intense geomagnetically-induced currents and ionospheric and radiation storms, and (4) on developing a com-prehensive specification of space climate, including the characterization of extreme space storms to guide resilient and robust engineeringof technological infrastructures. The roadmap clusters its implementation recommendations by formulating three action pathways, andoutlines needed instrumentation and research programs and infrastructure for each of these. An executive summary provides an overviewof all recommendations.� 2015 COSPAR. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://

creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Space weather; COSPAR/ILWS road map panel

Contents

Executive summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27471. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27502. Space weather: society and science . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27513. User needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2754

3.1. Electric power sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27553.2. Positioning, navigation, and communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27563.3. (Aero) space assets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2757

4. Promising opportunities and some challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2757
4.1. The opportunity of improved CME forecasts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2757
4.1.1. Solar surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27594.1.2. Heliosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2760

4.2. Challenges and opportunities for geomagnetic disturbances. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2761
4.2.1. L1 observations: Validation of >1 hr forecasts and interaction with the magnetosphere . . . . . . . . . . . . . . 27614.2.2. Reconfigurations in the magnetosphere–ionosphere system and strong GICs . . . . . . . . . . . . . . . . . . . . . . 2761
4.3. Research for improved forecasts of ionospheric storm evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27624.4. Steps for improved radiation belt forecasts and specification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27634.5. Challenges for forecasts with lead times beyond 2 days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27644.6. Specification of extreme conditions and forecasts of the solar cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2764

5. General recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2765
5.1. Research: observational, computational, and theoretical needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2767
5.2. Teaming of research and users: coordinated collaborative environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27685.3. Collaboration between agencies and communities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2771

6. Research: observational, numerical, and theoretical recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27737. Concepts for highest-priority research and instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2775

7.1. Quantify active-region magnetic structure to model nascent CMEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27757.2. Coupling of the solar wind to the magnetosphere and ionosphere, and strong GICs . . . . . . . . . . . . . . . . . . . . . . 27767.3. Global coronal field to drive models for the magnetized solar wind . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27777.4. Quantify the state of the coupled magnetosphere-ionosphere system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27787.5. Observation-based radiation environment modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27787.6. Understand solar energetic particles throughout the Sun–Earth system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2778

8. In conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2779Appendix A. Roadmap team and process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2779Appendix B. Roadmap methodology: tracing sample impact chains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2780Appendix C. State of the art in the science of space weather . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2781

C.1. Achievements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2781C.2. Prospects for future work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2782



C.J. Schrijver et al. / Advances in Space Research 55 (2015) 2745–2807 2747

Appendix D. Research needs for the solar-heliospheric domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2783Appendix E. Research needs for the geospace domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2787

E.1. Magnetospheric field variability and geomagnetically-induced currents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2787E.2. Magnetospheric field variability and particle environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2791E.3. Ionospheric variability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2795

Appendix F. Concepts for highest-priority instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2797
F.1. Binocular vision for the corona to quantify incoming CMEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2797F.2. 3D mapping of solar field involved in eruptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2798F.3. Strong GICs driven by rapid reconfigurations of the magnetotail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2798F.4. Coordinated networks for geomagnetic and ionospheric variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2799F.5. Mapping the global solar field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2800F.6. Determination of the foundation of the heliospheric field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2801F.7. Auroral imaging to map magnetospheric activity and to study coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2802F.8. Observation-based radiation environment modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2802F.9. Solar energetic particles in the inner heliosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2803
Appendix G. Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2804

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2806

Executive summary

Space weather is driven by changes in the Sun’s mag-netic field and by the consequences of that variability inEarth’s magnetic field and upper atmosphere. This resultsin a variety of manifestations, including geomagnetic vari-ability, energetic particles, and changes in Earth’s upper-most atmosphere. All of these can affect society’stechnological infrastructures in different ways.

Space weather is generally mild but some times extreme.

Mild space weather storms can degrade electric powerquality, perturb precision navigation systems, interruptsatellite functions, and are hazardous to astronaut health.Severe space storms have resulted in perturbations in theelectric power system and have caused loss of satellitesthrough damaged electronics or increased orbital drag.For rare extreme solar events the effects could be catas-trophic with severe consequences for millions of people.

Societal interest in space weather grows rapidly: Asscience and society increasingly recognize the impacts ofspace weather on the infrastructure of the global economy,interest in, and dependence on, space weather informationand services grows rapidly. Apart from having societal rele-vance, understanding space weather is an exciting sciencerevealing how the universe around us works.

Space weather is an international challenge: Significantscientific problems require substantial resources, with obser-vations having to cover the terrestrial globe and span thevast reaches of the heliosphere between Earth and the Sun.

Mitigating against the impacts of space weather can be

improved by designing less susceptible, more resilient tech-nologies, combined with better environmental knowledgeand more reliable forecasts. This roadmap outlines howwe can achieve deeper understanding and better forecasts,recognizing that the expectations for space weather infor-mation differ between societal sectors, and that capabilitiesto observe or model space weather phenomena depend onavailable and anticipated technologies.

The existing observatories that cover much of the Sun–Earth system provide a unique starting point: Moderateinvestments now that fill key capability gaps can enablescientific advances that could not be otherwise achieved,while at the same time providing a powerful base to meetmany operational needs. Improving understanding andforecasts of space weather requires addressing scientificchallenges within the network of physical processes thatconnect the Sun to society. The roadmap team identifiedthe highest-priority areas within the Sun–Earth space-weather system whose advanced scientific understandingis urgently needed to address current space weather serviceuser requirements. The roadmap recommends actionstowards such advanced understanding, focusing on thegeneral infrastructure to support research as well as onspecific concepts for instrumentation to meet scientificneeds.

Roadmap recommendations:

Research: observational, computational, and theoretical

needs:

1. Advance the international Sun–Earth system observa-tory along with models to improve forecasts based onunderstanding of real-world events through the develop-ment of innovative approaches to data incorporation,including data-driving, data assimilation, and ensemblemodeling.

2. Understand space weather origins at the Sun andtheir propagation in the heliosphere, initially prioritiz-ing post-event solar eruption modeling to developmulti-day forecasts of geomagnetic disturbance timesand strengths, after propagation through theheliosphere.

3. Understand the factors that control the generation ofgeomagnetically-induced currents (GICs) and of harshradiation in geospace, involving the coupling of the solarwind disturbances to internal magnetospheric processesand the ionosphere below.


4. Develop a comprehensive space environment speci-fication, first to aid scientific research and engineeringdesigns, later to support forecasts.

Teaming: coordinated collaborative research environment:

I Quantify vulnerability of humans and of society’sinfrastructure for space weather by partnering withuser groups;

II Build test beds in which coordinated observing sup-ports model development;

III Standardize (meta-) data and product metrics, andharmonize access to data and model archives;

IV Optimize observational coverage of the Sun-societysystem.

Collaboration between agencies and communities:

A Implement an open space-weather data and infor-mation policy;

B Provide access to quality education and informationmaterials;

C Execute an international, inter-agency assessmentof the state of the field on a 5-yr basis toadjustpriorities and to guide international coordination;

D Develop settings to transition research models tooperations;

E Partner with the weather and solid-Earth communi-ties to share lessons learned.

The roadmap’s research recommendations are expandedin three pathways that reflect a blend of the magnitude ofsocietal impact, scientific need, technological feasibility,and likelihood of near-term success. Each pathway needsrecommendations of any preceding it implemented toachieve full success, but can be initiated in parallel. Thepathways are designed to meet the variety of differentialneeds of the user communities working with different typesof impacts. Recommendations within each pathway aregrouped into actions that can be taken now, soon, or ona few-year timescale, each listed in priority order withinsuch group.

Pathway I recommendations: to obtain forecasts morethan 12 h ahead of the magnetic structure of incomingcoronal mass ejections and their impact in geospace toimprove alerts for geomagnetic disturbances and strongGICs, related ionospheric variability, and geospace ener-getic particles:

Maintain existing essential capabilities:

� solar magnetic maps (GBO, SDO) and EUV/X-rayimages at arcsec and few-second res. (SDO; Hinode),and solar spectral irradiance observations;� solar coronagraphy, best from multiple perspectives

(Earth’s view and L1: GBO and SoHO; and well offSun–Earth line: STEREO);

� in-situ measurements of solar-wind plasma and mag-netic field at, or upstream of, Sun–Earth L1 (ACE,SoHO; DSCOVR);� for several years, continue to measure the interaction

across the bowshock-magnetopause (as now withCluster/ARTEMIS/THEMIS; soon with MMS), to bet-ter understand wind-magnetosphere coupling;� satellite measurements of magnetospheric magnetic

and electric fields, plasma parameters, soft auroral andtrapped energetic particle fluxes (e.g., Van-AllenProbes, LANL satellites, GOES, ELECTRO-L, POES,DMSP);� ground-based sensors for solar, heliospheric, mag-

netospheric, and iono-/thermo-/mesospheric data tocomplement satellite data.

Model capability, archival research, or data infrastructure:

� near-real time, observation-driven 3D solar active-re-gion models of the magnetic field to assess destabiliza-tion and to estimate energies;� data-driven models for the global solar surface-coronal

field;� data-driven ensemble models for the magnetized solar

wind;� data assimilation techniques for the global ionosphere-

magnetosphere-atmosphere system using ground andspace data for nowcasts and near-term forecasts of geo-magnetic and ionospheric variability, making optimaluse of selected locations where laboratory-like test bedsexist or can be efficiently developed;� coordinated system-level research into large-scale rapid

morphological changes in Earth’s magnetotail andembedded energetic particle populations and their link-age to the ionosphere;� system-level study of the mechanisms of the particle

transport/acceleration/losses driving currents and pres-sure profiles in the inner magnetosphere;� stimulate research to improve global geospace modeling

beyond the MHD approximation (e.g., kinetic andhybrid approaches);� develop the ability to use solar chromospheric and coro-

nal polarimetry to guide full-Sun corona-to-heliospherefield models.

Deployment of new/additional instrumentation:

� binocular imaging of the solar corona at �1-arcsecondand at least 1-min. resolution, with about 10� to 20� sep-aration between perspectives;� observe the solar vector-field at and near the surface and

the overlying corona at better than 200-km resolution toquantify ejection of compact and low-lying current sys-tems from solar active regions;� (define criteria for) expanded in-situ coverage of the aur-

oral particle acceleration region and the dipole-tail fieldtransition region (building on MMS) to determine the


magnetospheric state in current (THEMIS, Cluster) andfuture high-apogee constellations, using hosted payloadsand cubesats where appropriate;� (define needs, then) increase ground- and space-based

instrumentation to complement satellite data of themagnetospheric and ionospheric variability to covergaps (e.g., in latitude coverage and over oceans);� an observatory to expand solar-surface magnetography

to all latitudes and off the Sun–Earth line [starting withSolar Orbiter];� large ground-based solar telescopes (incl. DKIST) to

perform multi-wavelength spectro-polarimetry to probemagnetized structures at a range of heights in the solaratmosphere, and from sub-active-region to global-cor-ona spatial scales;� optical monitors to measure global particle precipitation

to be used in assimilation models for geomagnetic dis-turbances and ionospheric variability.

Pathway II recommendations: to understand the particleenvironments of (aero) space assets leading to improvedenvironmental specification and near-real-time conditions.

With the Pathway-I requirements implemented:


� LEO to GEO observations of electron and ion pop-ulations (hard/�MeV and soft/�keV; e.g., GOES, . . .),and of the magnetospheric field, to support improvedparticle-environment nowcasts;� maintain a complement of spacecraft with high res-

olution particle and field measurements (such as theVan Allen Probes).

Model capability, archival research, or data infrastructure:

� specify the frequency distributions for fluences of ener-getic particle populations [SEP, RB, GCR] for a varietyof orbital conditions, and maintain archives of pastconditions;� develop, and experiment with, assimilative integrated

models for radiation-belt particles towards forecastdevelopment including data from ionosphere, thermo-sphere and magnetosphere and below, and validate thesebased on archival data.


� deploy high- and low-energy particle and electro-magnetic field instruments to ensure dense spatial cover-age from LEO to GEO and long-term coverage ofenvironment variability (incl., e.g., JAXA’s ERG).

Pathway III recommendations: to enable pre-event forecastsof solar flares and coronal mass ejections, and related solarenergetic particle, X-ray, EUV and radio wave eruptionsfor near-Earth satellites, astronauts, ionospheric storm

forecasts, and polar-route aviation, including all-clearconditions.

Maintain existing essential capabilities (in addition to

Pathway-I list):

� solar X-ray observations (GOES);� observe the inner heliosphere at radio wavelengths to

study shocks and electron beams in the corona and innerheliosphere;� maintain for some years multi-point in-situ observations

of SEPs on- and off Sun–Earth line throughout the innerheliosphere (e.g., L1, STEREO; including ground-basedneutron monitors);� maintain measurements of heavy ion composition (L1:

ACE; STEREO; near-future: GOES-R).

Model capability, archival research, or data

infrastructure:

� develop data-driven predictive modeling capability forfield eruptions from the Sun through the innerheliosphere;� investigate energetic particle energization and prop-

agation in the inner heliosphere, aiming to develop atleast probabilistic forecasting of SEP properties [cf.Pathway-I for heliospheric data-driven modeling];� ensemble modeling of unstable active regions to under-

stand energy conversions into bulk kinetic motion, pho-tons, and particles.


� new multi-point in-situ observations of SEPs off Sun–Earth line throughout the inner heliosphere to under-stand population evolutions en route to Earth (e.g.,Solar Orbiter, Solar Probe Plus).

Concepts for new priority instrumentation:Pathway I:

1. Quantify the magnetic structure involved in nascentcoronal ejections though binocular vision of thesource-region EUV corona, combined with 3D mappingof the solar field involved in eruptions through (near-)surface vector-field measurements and high-resolutionatmospheric imaging.

2. Understand the development of strong geomagnetically-induced currents through magnetotail-to-ionosphere in-situ probes, complemented with coordinated ground-based networks for geomagnetic and ionosphericvariability.

3. Map the global solar field, and use models and observa-tions to determine the foundation of the heliosphericfield, to drive models for the solar-wind plasma andmagnetic field.

4. Image the aurorae as tracers supporting the quantificationof the state of the magnetosphere-ionosphere system.


Pathway II:

5. Combined ground- and space-based observations forthe modeling of the dynamic radiation-belt populations;

Pathway III:

6. In-situ multi-point measurements to understand solarenergetic particles in the Sun–Earth system.

1. Introduction

As technological capabilities grow, society constructs arapidly deepening insight into the forces that shape theenvironment of our home planet. With that advancingunderstanding comes a growing appreciation of our vul-nerability to the various attributes of space weather. Thevariable conditions in what we think of as “space” drivesociety to deal with the hazards associated with living inclose proximity to a star that sustains life on Earth evenas it threatens humanity’s technologies: the dynamicmagnetism manifested by the Sun powers a sustained yetvariable solar wind punctuated by explosive eruptions thatat times envelop the planets, including Earth, in spacestorms with multiple potentially hazardous types of condi-tions. Powerful magnetic storms driven by solar eruptionsendanger our all-pervasive electric power grid and disruptthe many operational radio signals passing throughour planet’s upper atmosphere (including the satellitenavigation signals that is now vital to society). Energeticparticle populations can lead to malfunctions of satellitesand put astronauts at risk.

These solar-powered effects in Earth’s environment, col-lectively known as space weather (with the shorthand nota-tion of SWx), pose serious threats to the safe and efficientfunctioning of society. In recognition of the magnitude ofthe hazards, governments around the world are investingin capabilities to increase our awareness of space weather,to advance our understanding of the processes involved,and to increase our ability to reliably forecast, preparefor, design to, and respond to space weather. Scientists witha wide variety of expertise are exploring the magnetism ofthe Sun from its deep interior to the outermost reaches ofthe planetary system, and its impacts on planetary environ-ments. Great strides forward have been made towards acomprehensive study of all that happens between the mag-netized interior of the Sun and mankind’s technologicalinfrastructure. An ensemble of in-situ and remote-sensinginstruments on the Earth and in space is uncovering andquantifying many aspects of evolving space weather fromSun to Earth. These instruments monitor the progress ofspace weather across the vast interplanetary volume knownas the heliosphere and observe how space weather impactspass through Earth’s magnetic cocoon and into Earth’supper atmosphere. But this ensemble is subject tomajor observational gaps and needs to be strengthened

by improved coordination and integration. Scientificresearch and observations are increasingly combined incomputer models and analyses that describe and forecastspace weather conditions, aiming to protect society fromthe dangers of space storms, but still lacking essentialcapabilities in many parts of the overall Sun-to-Earth chain.

Similar to terrestrial weather, space weather occurs allthe time and it frequently affects our technological infras-tructure in ways that are not catastrophic, but that arecostly in their aggregate value to the global economy.Impacts on satellite-based navigation and timing capabili-ties or satellite communications affect users of the globalnavigation satellite system (GNSS) from precision agricul-ture to national security. Space weather is a serious con-straint on the resilience of GNSS and thus a vital issuefor a huge and growing range of economic activities:impacts are seen throughout the multitude of industry sec-tors that use satellite-based navigation systems, includingon the aviation industry as it seeks to exploit GNSS as away to optimize aircraft routing and hence airspace capac-ity; on the development of self-driving cars that rely onGNSS; on the accuracy of maritime drilling operations;on the timing systems that synchronize power and cell-phone networks in some countries, and on the financeindustry as it moves to microsecond-resolution time stamp-ing of financial transactions. Satellite anomalies and fail-ures due to radiation storms impact the flow ofinformation from satellite phones to weather monitoring,and from scientific exploration to intelligence monitoring.Space-weather impacts on the all-pervasive electric powergrids have an even larger reach, with impacts ranging fromrather frequent power-quality variations (such as voltagevariations, frequency drifts and harmonics, and very shortinterruptions) to potential infrequent but major powerinterruptions.

Also similar to terrestrial weather, extreme space stormsare expected to have major impacts. With our sensitiveelectrical, electronic, and space-based technologies expand-ing rapidly into a tightly woven network of applications weare not certain about the magnitudes of the impacts ofextreme space weather, but studies suggest that we shouldbe seriously concerned. For example, a single unusuallysevere geomagnetic storm has been hypothesized to causelong-term power outages to tens of millions of citizens ofmultiple countries and is thus listed among national secur-ity risks (e.g., the UK national risk register). Permanentloss of multiple satellites by severe energetic-particle stormshas significant consequences for communications, surveil-lance, navigation, and national security.

Although small compared to the risks deemed to beinvolved, the resources required to understand spaceweather and its impacts are substantial: the domain to becovered for successful space-weather forecasting andpreparedness is vast, the physics of the intricately coupledprocesses involved is complex, and the diversity of therapidly-growing user communities crosses all aspects of


society and its globally-connected economy. Moreover, theimpacts of space weather span the globe. For those rea-sons, an international approach is paramount to success-fully advancing our scientific understanding of spaceweather. This realization prompted the Committee onSpace Research (COSPAR) of the International Councilfor Science (ICSU) and the International Living With aStar (ILWS) Steering Committee to commission a strategicassessment of how to advance the science of space weatherwith the explicit aim of better meeting the user needs aroundthe globe. This report is the outcome of that activity.

In the spring of 2013, the leadership of COSPAR andILWS appointed a team of experts charged to create thisroadmap (see Appendix A for the process followed). Themission statement by COSPAR’s Panel on Space Weather(PSW) and the ILWS steering committee asks the team to“[R]eview current space weather capabilities and identifyresearch and development priorities in the near, mid andlong term which will provide demonstrable improvementsto current information provision to space weather serviceusers”, thereby expressing a focus on the terrestrialenvironment. The charge continued with the expectationthat “the roadmap would cover as minimum:

� Currently available data, and upcoming gaps.� Agency plans for space-based space weather data

(national and international): treating both scientificand monitoring aspects of these missions.� Space and ground based data access: where current data

is either proprietary or where the geographic location ofthe measurement makes data access difficult.� Current capability gaps which would provide a marked

improvement in space weather service capability.

The outcome should centre on a recommended approachto future developments, including coordination andaddressing at least:

1. Key science challenges.2. Data needs, space and ground based.3. Smooth and organised transition of scientific develop-

ments into reliable services.”

For the purpose of this roadmap we use the followingdefinition: Space weather refers to the variable state of the

coupled space environment related to changing conditionson the Sun and in the terrestrial atmosphere, specifically

those conditions that can influence the performance and

reliability of space-borne and ground-based technological

systems, and that can directly or indirectly endanger human

well-being. Aspects of space situational awareness such asspace debris in Earth orbit and asteroids or other near-Earth objects are not considered in this roadmap. Thestructure under which the roadmap team operates impliesthat it focus exclusively on civilian needs, although manyof its conclusions are likely directly pertinent to the securityand military sectors of society.

This roadmap identifies high-priority challenges in keyareas of research that are expected to lead to a betterunderstanding of the space environment and an improve-ment in the provision of timely, reliable information perti-nent to effects on space-based and ground-based systems.Among those is the realization that we cannot at presentuse observations of the Sun to successfully model the mag-netic field in coronal mass ejections (CMEs) en route toEarth, and thus we cannot forecast the strength of the per-turbation of the magnetospheric field that will occur.Another example is that we understand too little of mag-netic instabilities to forecast the timing and energy releasein large solar flares or in intense (sub) storms in geospace.Advances in these areas will strengthen our ability tounderstand the entire web of physical phenomena that con-nect Sun and Earth, working towards a knowledge level toenable forecasts of these phenomena at high skill scores.

The roadmap prioritizes those advances that can bemade on short, intermediate, and decadal time scales,identifying gaps and opportunities from a predominantly,but not exclusively, geocentric perspective. This roadmapdoes not formulate requirements for operational forecastor real-time environmental specification systems, nor doesit address in detail the effort required to utilize scientificadvances in the improvement of operational services.This roadmap does, however, recognize that forecasts(whether in near-real time or retrospectively) can helpuncover gaps in scientific understanding or in modelingcapabilities, and that test beds for forecast tools serve bothto enable quantitative comparison of competing modelsand as staging environment for the phased transition ofresearch tools to an operational application.

The main body of this roadmap is complemented byAppendices that contain more detailed or supplementalinformation, describe the roadmap process, summarizescientific advances and needs, and outline instrumentationconcepts. Appendix G provides a list of the acronyms andabbreviations used in this document.

2. Space weather: society and science

The task defined by COSPAR and ILWS is founded onthe realization that space weather is a real and permanenthazard to society that needs to be, and can be, addressedby combining scientific research with engineering ingenu-ity: protecting society from space weather requires thatwe adequately understand the physical processes of spaceweather, that we characterize the conditions to whichtechnological infrastructures need to be designed, that welearn to effectively forecast space weather, and that theconsequences of acting on such forecasts are accepted asnecessary for the protection of societal infrastructure.

Societal use of, and dependence on, ground-basedelectrical systems and space-based assets has growntremendously over the past decades, by far outpacingpopulation growth as society continues to grow itselectrical/electronic and space-based technologies. Global

Fig. 1. Number of publications per year with “space weather” in theabstract in NASA/ADS (blue; left axis), and the number of web sitesreturned by a Google search for “space weather” within calendar years(since 2003) in thousands (red; right axis).


electricity use has increased by a factor of about 1.6 overthe 15-year period between 1997 and 2012 (InternationalEnergy Agency, 2013). The global satellite industry revenuehas multiplied by a factor of about 4.2 over that period (toUS$190 billion per year for 2012, part of a total value ofUS$304 billion for the overall space industry, with over1000 operating satellites from over 50 countries; SatelliteIndustry Association, 2013). In contrast, the global pop-ulation grew by approximately 20% over that period(Population Reference Bureau, 2013), demonstrating ourincreasing use of electrical power and satellite-based infor-mation per capita.

With that growth in electrical/electronic and space-based technologies comes increasing vulnerability to spaceweather: where a century ago the main risk was associatedwith the telegraph systems we now see impacts in the elec-tric power grid, in satellite functionality, in the accuracy ofnavigation and timing information, and in long-range high-frequency (HF) radio communication. We see an increas-ing interest in understanding space weather impacts andthe threats these pose are spread over a variety of civiliansectors (and non-civilian sectors that lie beyond the scopeof this Roadmap). Selected reports on these impacts (thatthemselves provide information on more literature on thesubject) are compiled in an on-line resource list1 thataccompanies this report; that resource list also includes aglossary of solar-terrestrial terms,2 and links to aNational Geographic introduction to space weather acces-sible via YouTube,3 and lectures related to space weather,its impacts, and its science in the NASA HeliophysicsSummer School.4

The reality of the threat to society posed by spaceweather is increasingly acknowledged – reflected, for exam-ple, in the exponential growth of the number of web pageson space weather (totaling over 130,000 new entries in2013; see Fig. 1) and in the number of customers subscrib-ing to alert and forecast services (exceeding, for example,44,000 for the US Space Weather Prediction Center). Acore difficulty facing any study that attempts a cost-benefitanalysis for space weather is inadequate knowledge of thetechnological and economic impacts of ongoing spaceweather and of the risk posed by extreme space storms.This hampers the quantitative identification of the mostsignificant impacts of space weather and consequently theprioritization of the research areas and the deployment ofinfrastructure to protect against space weather. For ourroadmap, existing assessments of threats and impacts com-piled by organizations currently engaged in providingspace weather information to affected sectors proved ade-quate for a prioritization of recommendations, but achiev-ing a quantification of the SWx impact on societal

1 http://www.lmsal.com/�schryver/COSPARrm/SWlibrary.html.2 http://www.swpc.noaa.gov/content/space-weather-glossary.3 http://www.lmsal .com/�schryver/COSPARrm/SWlibrary.

html#youtube.4 http://www.vsp.ucar.edu/Heliophysics/.

technologies is important for the process of allocating therequired resources for research and forecasting, and ofdetermining what sectors of society should be involved inappropriating which resources. For example, althoughthe threat posed by geomagnetic storms is broadly recog-nized as real (e.g., Krausmann, 2011; Langhoff andStraume, 2012), establishing the vulnerability and conse-quences of such an event has proven to be very difficult(Space Studies Board, 2008; DHS Office of RiskManagement and Analysis, 2011; JASON, 2011), hamper-ing a cost-benefit assessment of investments that couldmake impacted systems less vulnerable by suitable engi-neering or by improved forecasting (DHS Office of RiskManagement and Analysis, 2011).

Like terrestrial weather, space weather manifests itself asa variety of distinct phenomena, and like terrestrialweather, it ranges from benign to extremely severe. Mostfrequently, space weather is very weak in intensity withapparently little impact on technology. Strong, severe, orextreme geomagnetic conditions (as measured by the Kpindex, on NOAA’s G scale5 occur only 5% of days througha solar magnetic cycle. Even though there are no reports ofcatastrophic failures in the US high-voltage power grid,one study finds that there appears to be an increase by40% ± 20% in insurance claims for industrial electricaland electronic equipment on the 5% most geomagneticallyactive days (as measured by the rate of change in the geo-magnetic field strength) relative to quiet days, and there isan increase of 30% ± 10% in the occurrence frequency ofsubstantial disturbances in the US high-voltage power grid(Schrijver et al., 2014). Another study suggests that, over-all, approximately 4% of the disturbances in the US high-voltage power grid reported to the US Department ofEnergy appear to be attributable to strong but not extremegeomagnetic activity and its associated geomagneticallyinduced currents (GICs; Schrijver and Mitchell, 2013).More such studies are needed to validate these findings

5 http://www.swpc.noaa.gov/NOAAscales/.

http://www.lmsal.com/~schryver/COSPARrm/SWlibrary.html

http://www.lmsal.com/~schryver/COSPARrm/SWlibrary.html

http://www.swpc.noaa.gov/content/space-weather-glossary

http://www.lmsal.com/~schryver/COSPARrm/SWlibrary.html#youtube



http://www.vsp.ucar.edu/Heliophysics/

http://www.swpc.noaa.gov/NOAAscales/


and particularly to help uncover the pathways by whichthese impacts occur (a point to which we return in ourrecommendations below).

Other aspects of space weather can adversely affect satel-lites. Severe solar energetic-particle (SEP) storms, forexample, impact satellites directly, while expansion of theterrestrial upper atmosphere by magnetospheric variability(through Joule heating) may affect low-orbiting satellitesby modifying their orbits through increased drag which isan issue for on-orbit operations, collision avoidance, andcould eventually lead to early re-entry (e.g., Rodgerset al., 1998). A series of such storms during the 2003October–November time frame, for example, saw con-siderable impacts on satellites through electronic single-event upsets (SEUs), solar-array degradation, modifiedorbit dynamics for spacecraft in low-Earth orbits, andnoise on both housekeeping data and instrument data.Another manifestation of space weather is the enhance-ment of radiation belt (RB) particles and of mag-netospheric plasma that cause charging/dischargingphenomena or state upsets in satellite electronics. For 34Earth and space science missions from NASA’s ScienceMission Directorate, for example, 59% of the spacecraftexperienced such effects (Barbieri and Mahmot, 2004). Agraphic laboratory demonstration of discharging insidedielectric materials (a critical space weather impact forsatellites in geosynchronous and middle Earth orbit) isavailable on YouTube.6

A third impact for space weather occurs via severe modi-fication of trans-ionospheric signals by highly variableplasma density in space and time, thus affecting customersof GNSS services. The economic impact of this type ofspace weather has yet to be investigated, being complicatedby the fact that it will mostly occur well downstream of theimmediate service providers and also by the fact that GNSStechnology is rapidly evolving even as the total numbers ofusers and uses increases, increasingly in layered applicationsthat may hide just how GNSS-dependent a system is.

The threat posed to society by the most severe spacestorms that occur a few times per century is largelyunknown and the magnitude of such a threat is conse-quently highly uncertain: the technological landscapeevolves so rapidly that our modern-day highly-intercon-nected societal infrastructure has not been subjected tothe worst space storms that can occur. Some reports putthe threat by the most severe space storms among the sig-nificant threats faced by our technology-dependent society.Geomagnetic disturbances (GMDs) on electrical systemshave been known to impact technology for over 150 years,starting with the telegraph systems, and showing a clearcorrelation with the sunspot cycle (Boteler et al., 1998).Among the insurance-industry reports that review thespace weather risk landscape from the industry perspective(e.g., Hapgood, 2011a,b), one (Lloyd’s, 2013) concludes for

6 http://youtu.be/-EKdxzZ52zU.

the US in particular that “the total population at risk ofextended power outage from a Carrington-level storm[i.e., unusually strong but likely to occur approximatelyonce per century] is between 20–40 million, with durationsof 16 days to 1–2 years”, while recognizing that even forweaker storms “the potential damage to densely populatedregions along the Atlantic coast is significant.” The WorldEconomic Forum (2013) includes vulnerability to geomag-netic storms explicitly in its listing of top environmentalrisks deemed to be able to significantly impact the globaleconomy. The US National Intelligence Council (2013)noted that “[u]ntil cures’ are implemented, solar super-storms will pose a large-scale threat to the world’s socialand economic fabric”. A report on a risk analysis for spaceweather impacts in the UK (Royal Academy ofEngineering, 2013) stated that “the reasonable worst casescenario would have a significant impact on the nationalelectricity grid.” Impacts of geomagnetic storms are notlimited to grids at high latitudes: although geomagneticlatitude is an important factor in GMD strengths, “geo-logical conditions tend to override the effect of latitude”

(NERC, 2012a). In addition, lower-latitude regions canexperience GICs arising from fluctuations in the mag-netospheric ring current as demonstrated by the seriousimpact of the Halloween 2003 storms on the power gridin South Africa (e.g., Gaunt, 2013).

Even if we disregard the uncertain impact of extremespace weather, we know that impacts of moderate toextreme space storms that occur a few hundred times per11-year solar cycle have consequences that merit substantialattention and investment. One study of the overall cost tothe US economy alone of non-catastrophic disturbancesin the US power grid attributable to geomagneticallyinduced currents suggests that the impacts may be as largeas US$5–10 billion/year (Schrijver et al., 2014; we note thatmore detailed follow-up studies are needed to validate thisresult, because that study had to combine its significantfindings on insurance-claim statistics with other sources ofinformation to estimate the financial impact). Economicimpacts of space weather through other technologicalinfrastructures have yet to be established, but threat assess-ments suggest “that space weather is the largest contributorto single-frequency GPS errors and a significant factor fordifferential GPS” (American Meteorological Society,2011), for an industry that is worth of order US$100 bil-lion/year worldwide (American Meteorological Society,2011; Pham, 2011). A recent study (Schulte in denBaumen, 2014) made a first attempt to couple a GIC impactmodel with an economic model of global trade showing howGIC impact in three different regions (China, Europe andNorth America) would drive impacts across the world econ-omy. It reinforces the message that space weather is a globalproblem – that a physical impact in one region can damageeconomies far from the impact site.

Given the persistent presence of the threat, society’sincreasing exposure to space weather, and the likely low-fre-quency but high-impact extreme-storm scenarios, it is not

http://youtu.be/-EKdxzZ52zU


surprising that calls for the preparation for, forecasting of,mitigation against, and vulnerability assessment for spaceweather impacts by the international community echo invarious studies over the past decade, including reports fromacademia (Hapgood, 2011a,b), from the US NationalResearch Council (2008), the UN Committee on thePeaceful Uses of Outer Space (2013), and from theOrganisation for Economic Co-operation andDevelopment (OECD, 2011). Recognition of the societalhazard posed by space weather is also reflected in reportsby, for example, the North American Electric ReliabilityCorporation (NERC, 2012b; NERC, 2013a,b) and in theorder by the US Federal Energy Regulatory Commission(FERC, 2013) to develop reliability standards for geomag-netic disturbances.

The user base interested in forecasts of space weather isgrowing rapidly with the increased awareness of space-weather threats and impacts. The official US space-weatherforecast center (the Space Weather Prediction Center of theUS National Oceanographic and AtmosphericAdministration), for example, sees a continuing rapidgrowth in subscribers to its “Product SubscriptionService7” that was initiated in 2005 and that exceeded40,000 individual subscribers early in 2014. A survey ofthe subscribers to the SWPC service in 2013 enabled anassessment of the interests from the user side (Schrijverand Rabanal, 2013), which concluded that “[s]pace weatherinformation is most commonly obtained for reasons of[indirect impacts through interruptions of power or com-munications on] human safety and continuity or reliabilityof operations. The information is primarily used for situa-tional awareness, as aid to understand anomalies, to avoidimpacts on current and near-future operations by imple-menting mitigating strategies, and to prepare for potentialnear-future impacts that might occur in conjunction withcontingencies that include electric power outages or GPSperturbations. Interest in, anticipated impacts from, andresponses to the three main categories of space weather [–geomagnetic, radiation, and ionospheric storms –] are quiteuniform across societal sectors. Approximately 40% of therespondents expect serious to very serious impacts fromspace weather events if no action were taken to mitigateor in the absence of adequate space weather information.The impacts of space weather are deemed to be substan-tially reduced because of the availability of, and theresponse to, space weather forecasts and alerts.” It appearsthat many users of space weather forecasts apply the fore-cast information to avoid impacts on their systems andoperations, either by increased monitoring given situationalawareness or by taking preventive mitigating actions. Aswith terrestrial weather, this means that the economic valueof space weather forecasts likely significantly exceeds thetotal costs of detrimental impacts, and that this value wouldincrease as forecast accuracy and specificity would increase.

7 http://pss.swpc.noaa.gov.

Other valuable uses of space weather information as indi-cated by the subscribers to space-weather information liein anomaly analysis and system design specification(Schrijver and Rabanal, 2013).

The study of space weather is important because of itssocietal relevance and much headway has been made inrecent years (see a brief discussion in Appendix C). Spaceweather also teaches us about the physical processes ofthe local cosmos that is our home within the Galaxy.More commonly known as the field of Sun–Earth connec-tions, or as heliophysics particularly within the US, this isthe science of the astrophysical processes that occur in thedeep solar interior, in the vast reaches of the solar atmo-sphere that extend beyond the furthest planet out to theinterstellar medium, and that includes the variety of cou-pling processes to the planets and natural satellites of ourown solar system. Understanding all of these processesand interactions between diverse environments is our step-ping stone to understanding what happens elsewhere in theuniverse in similar environments, as much as to under-standing the distant past and future of our own planetarysystem and its host star.

3. User needs

The complexity of the interconnected system of physicalprocesses involved in space weather precludes a single,comprehensive, yet understandable and concise explo-ration of the network as a whole. To effectively and effi-ciently address its charge, the roadmap team thereforedeveloped a strategy that focuses on three largely distinctspace weather phenomena with largely complementaryimpact chains into societal technologies: (a) geomagneticdisturbances that drive electric currents through the electricpower infrastructure, (b) variability in the ionospheric elec-tron density that impacts positioning and navigation sys-tems, and (c) energetic particles for space assets,astronauts and stratospheric air traffic (with possibleimpacts on passengers, crews, and avionics). Many othersystems are susceptible to these and other manifestationsof space weather, whether directly or indirectly, buttogether these three chains encompass the most significantimpacts and together they cover most of the research andforecast needs that would be needed for the other typesof impacts. Fig. 2 lays out these three chains side by side,with an approximate vertical time axis to map the userneeds and the physical domains involved. Some details ofthe impact tracing exercise are described in Appendix B.

The societal needs for space weather information,including forecasts, are rapidly growing as witnessed bythe exponential growth in the US NOAA/SWPC productcustomer base, the growth in space-weather researchpublications (Fig. 1), and in the variety of strategic studiesand reports (see Section 2). Of the subscribers to the SWPCelectronic alert and forecast services customers, only about20% deem the quality and content of the available informa-tion adequate (Schrijver and Rabanal, 2013). Even though

http://pss.swpc.noaa.gov

Fig. 2. Overview of the primary impacts and their societal sectors of space weather. The red shading in the background indicates the priority needs for theuser communities behind each of the impacts, differentiated by time scale for forecast or for archival information as shown on the left. Text boxes identifythe primary needed observations, archival measurements, and models to complete the impact chain, differentiated (using color, see legend) by solar,heliospheric, and geospace domains.


about two-thirds of these customers considers the informa-tion “generally adequate”, the forecasts leave much to bedesired owing to a lack of scientific understanding in multi-ple sectors of the Sun–Earth connected system and to apoor knowledge of the magnitude of impacts.

When provided with warnings of significant spaceweather storms, the subscribers to space weather informa-tion services typically either increase monitoring or imple-ment preventive action in roughly comparable fractions.Increased availability of real-time information andincreased specificity and accuracy of the forecasts shouldenable more targeted and effective actions, while loweringthe costs associated with false alarms. Improved quality offorecasts and improved quantification of the system vul-nerability can put society on the desirable path of either effi-ciently acting on space-weather forecasts as one would onterrestrial weather forecasts, or by improving system designand operations that reduce vulnerability as done for othernatural forces, at least up to some value that is ideally basedon a well-founded cost-benefit risk assessment.

In the three technological infrastructures of our impacttracing exercise different approaches to mitigate the SWxrisk are used and consequently also the needs of SWx ser-vices supporting the societal sectors using these infrastruc-tures are different:

3.1. Electric power sector

The electric power sector is primarily affected by geo-magnetically-induced currents (GICs), which can pushtransformers off their linear domains leading to dissipativeheat that can damage transformers, and to generation ofharmonics of the primary 60-Hz or 50-Hz wave. These har-monics can cause protective systems to trip and can affectsystems operated by power customers downstream in thedistribution network. Power-grid operators can reduce orprevent damage by optimizing design, by introducing pro-tective equipment, or by redistributing and changing thepower-generation resources so that fewer long-distancetransfers are needed and more near-locally-generated


power is available to counter frequency and voltage mod-ulations. Other solutions, such as that adopted in the UKwith its relatively compact power system, are to bring allavailable grid links into operation in order to maximizeredundancy and to spread GICs over the whole system,reducing the impact on individual system elements. Tooptimize system protection, improved knowledge of theimpacts is needed, including extreme-event scenarios,which requires historical and pre-historical information.

In order to effectively plan the power grid operationssubject to strong GMDs, the sector has expressed a needfor a reliable warning at least half a day ahead of comingstorms, preferably with a reliable forecast of the magnitudeand duration of the geomagnetically induced currents(GICs). The propagation time for CMEs from Sun toEarth is typically of order 2–4 days, and even for the fastestknown events exceeds 0.75 days. Hence, the need for aforecast for CME arrivals about 12 h ahead requires theformulation of a forecast well after a solar eruption hasoccurred (and should have been observed), but long beforeit approaches geospace (here defined loosely as a domainencompassing everything from Earth’s middle atmosphereout to the upstream magnetopause and downstreambeyond the magnetotail), or any sensors currently availableupstream of Earth.

In Fig. 2, the highest-priority user needs for the powersector are indicated by red-shaded areas in the bar in thebackground of the left-hand column. The deepest red areasfor GIC service needs are in the forecasts with half-day leadtimes and in the specifications of extreme conditions basedon archival information on past conditions. Interests in12 h-ahead warnings of GMDs are also expressed by indus-tries using the geomagnetic field directly, including mining,drilling, and surveying. We note that addressing the needsof the GMD/GIC and ionospheric communities also pro-vides information on modulations of satellite drag causedby current dissipation within the ITM domain of Earth’supper atmosphere.

3.2. Positioning, navigation, and communication

Positioning and navigation services such as provided bythe GNSS are most affected by ionospheric electron-densityvariability via strong coupling with magnetospheric andthermospheric changes driven by CMEs and high-speedsolar wind streams or related to X-ray and extremeultraviolet (EUV) eruptions on the solar surface or (of par-ticular importance for the low-latitude ionosphere) upwardpropagating waves originating in the lower Earth atmo-sphere. Ionospheric range errors of up to 100 m in singlefrequency GNSS applications can be largely mitigated byproper ionospheric models, or more accurately by regionaland/or global maps of the total electron content (TEC) orby 3D electron density reconstructions provided in near-real time (i.e., within minutes, or even seconds, of measure-ments being made). Precise and safety–critical positioningand navigation, as required, for instance, in aviation, suffer

in particular from steep spatial gradients and temporalvariability of the plasma density. Thus, plasma bubblesand small-scale electron density irregularities (known asplasma turbulence) may cause strong fluctuations in signalamplitude and phase that are called radio scintillations;severe phase scintillation may even cause loss of phase lockof the signal.

In professional GNSS services position accuracy is sup-ported by augmentation systems. These include (a) widearea satellite-based augmentation systems (SBAS) such asEGNOS in Europe, WAAS in North America, andMSAS in Asia), and (b) ground-based augmentation sys-tems that provide local services such as around sea- andair-ports. These systems monitor the accuracy and reliabil-ity of the positioning signal and provide an integrity flagthat can warn GNSS users when back-up solutions for nav-igation must be used (well demonstrated by the WAAS sys-tem during the Halloween storms of 2003). SBAS systemsbring a further vulnerability to space weather in that theirmessages must cross the ionosphere as both uplink to, anddownlink from, a satellite hosting an SBAS relay system.Both uplink and downlink are vulnerable to ionosphericradio scintillation. Similarly the increasing use of satellitelinks for communication with, and tracking of, aircraftopens up a vulnerability to scintillation. On the other hand,forecasts of plasma perturbations have a great potential formitigating space weather impact on radio systems by alert-ing operators and customers of telecommunication andnavigation systems and remote sensing radars as well.Thus, for example, airplanes whose HF and satellite com-munication may be disturbed heavily typically require –as is the case for weather conditions – a forecast of iono-spheric electron density variability a good fraction of aday ahead of time. Correcting for, or accommodating,these impacts on HF telecommunication and GNSS signalsrequires an extensive network of ionospheric measurementdevices, rapid modeling, and rapid dissemination of correc-tion information. The needs of most of the GNSS cus-tomers thus lands in the time domain from current state(nowcast) to of order an hour ahead (see middle columnof Fig. 2), but somewhat longer-term forecasts are valu-able, for example, for planning purposes of operationsand emergency response activities. These needs will evolveas dual-frequency and/or differential GNSS usage becomemore widespread.

Another aspect of ionospheric electron density variabil-ity driven by space weather is radio wave absorption in thefrequency range below about 30 MHz. This arises whenhigh-energy inputs to the atmosphere generate additionalionization at altitudes around 90 km altitude and belowknown as D region in the ionosphere. In this region theneutral density is high and therefore also the collision fre-quency between charged particles and neutrals. Thus,charged particles that are excited by electromagnetic wavesin the frequency range mentioned above lose their energyvery fast, i.e. the radio wave will be absorbed or at leastheavily damped. D region ionization can arise from a


variety of high-energy space weather phenomena: intensebursts of X-rays from solar flares and the precipitation ofsolar energetic particles (SEP). These have the possibilityto disrupt HF communications that are used by civil avia-tion in remote regions, especially over the oceans and poles.The impact of flare X-rays appears to be a nuisance ratherthan a serious problem, since they produce only short-liveddisruptions (of order an hour) that can be mitigated bywell-established procedures and by use of satcom applica-tions as an increasingly common backup. However, theimpact of SEPs is very significant as their impact is greatestin the polar regions (where satcom backup is not currentlyavailable) and is long-lived (up to days) so that proceduralmeasures are insufficient. Thus polar flights are diverted toother less vulnerable routes during SEP events, imposingsignificant extra costs on airlines (around $100 k per diver-sion; National Research Council, 2008). This problemcould be mitigated in the future with increased use of sat-com the polar regions, including use of the potentialCanadian PCW (Polar Communications and Weather)satellite system.

The presence of a strong D region also impacts low-frequency navigation systems by advancing the arrival ofskywave, i.e., the LF signal reflected from the ionospherethat interferes with ground-wave, which is the LF signalpropagating as an interface wave along the surface of theEarth. Thus older LF systems such as LORAN andDECCA were prone to major position errors, for exampleduring REP events at mid-latitudes. However, most of theseold systems have now been taken out of service in favor ofmore accurate GNSS. New LF systems, such as theeLORAN system now deployed operationally in the UK,can match the accuracy of GNSS, but like GNSS includeintegrity checks to warn when location data are not accurate.

3.3. (Aero) space assets

The impacts of high-energy solar-energetic and radia-tion-belt particles are strongest in space-based assets,including satellites and astronauts, but extend to strato-spheric aircraft as these rely increasingly on ever-smallerelectronics systems distributed throughout the vehicle.Such effects are caused by energetic particles (those of galac-tic origin all the time, and SEP where these have strongfluxes at energies >500 MeV) that penetrate to the Earth’satmosphere and interact with atmospheric neutral particles.Such collisions produce cascades of neutrons and ions thatcan interact with aircraft to produce single event effects(SEE) and increased dose rate for passengers and crew.During SEP events radiation effects at aircraft trajectoriesare strongest at high latitudes where solar particles canmore easily penetrate. Whether there are detrimental effectson the health of airline crews and passengers remains to beestablished, but assessing the risk, regardless of the outcomeof the studies, is clearly of importance.

The requirements from the involved sectors separatesomewhat by type of radiation, as indicated in Fig. 2. In

the background, with mostly longer-term and moderatevariability, is the galactic cosmic ray (GCR) population,for which archival data spanning centuries is importantand in principle available in natural records such as icesheets, rocks, sediments, and the biosphere. Radiation-belt(RB) particles need to be characterized for system designpurposes based on archival and extreme-event information,and their nowcast and forecast is important for planning ofspacecraft special operations and maneuvers. SEPs arehighly variable and dangerous during high-intensity,large-fluence events not only for satellites, but also forastronaut activities (which puts particular value on theall-clear forecasts for 1–3 days). Archival and near-real timeinformation is needed for design and anomaly resolution,while forecasts on time scales of up to a day or so are, asfor the RB particles, important for planning of spacecraftand astronaut operations. The time-scale needs for the threedifferent populations of energetic particles are shown sepa-rately in color-coded background bars in Fig. 2.

The customer requirements for each infrastructuredescribed above are based on present-day needs. We antici-pate that the requirements for GMDs and energetic parti-cles will remain in place for many years to come, perhapseven with more urgency as the power-grid reach and thepower-dependence of society grow and as we dependincreasingly on space-based assets. For navigation andpositioning services the requirements appear to be alreadychanging as technologies advance: whereas satellite-only,single-frequency navigation systems are sensitive to iono-spheric electron density variability, use of dual-frequencysolutions and GNSS augmentation systems are reducingthe impact of large scale electron density variations. Themitigation of harm from ionospheric scintillation is proba-bly now the key challenge for future GNSS systems.

4. Promising opportunities and some challenges

4.1. The opportunity of improved CME forecasts

One conclusion from the exercise with three sampleimpact chains (c.f., Appendix B) is that during recent yearsresearch in solar-terrestrial physics has advanced so much(compare Appendix C) that within the time span of next5–10 years we anticipate far more accurate and specificforecasts of incoming solar-wind perturbations that willhave lead times from some hours to a couple of days.Such forecasts (pertaining to the heliospheric evolution ofpost-eruption solar activity prior to their arrival at Earth)would improve our capabilities to respond to the userneeds described in Section 3 in several ways. Therefore,we have dedicated one branch in our recommendations(which we refer to as Pathway I) particularly to suchresearch and collaboration activities that can be coupledwith the anticipated advancements in post-eruptionforecasts.

Ongoing solar missions have given us guidance onoptimal solar surface observations to support modeling

Fig. 3. Identification of the top-priority needs to advance understanding of space weather to better meet the user needs, differentiated by impact. Withineach chain, letters identify opportunities for immediate advance (O), and the largest challenges (C) that need to be addressed.


so that improved estimates on the CME magnetic structureand energy content, as well as the propagation in theheliosphere can be achieved. With this information notonly the arrival timing of CMEs but in particular theirgeo-effectiveness could be provided more reliably thantoday for lead times well beyond 0.5–1 h (which is thecharacteristic travel time for heliospheric perturbationsfrom a sentinel placed in the solar wind upstream ofEarth at the so-called L1 point where a spacecraft can read-ily remain at a nearly fixed point essentially on the Sun–Earth line). At the same time, our capabilities to monitorand model the near-Earth energetic particle and radiationenvironment have advanced so much that forecasts of simi-lar lead times can be generated to support also operationsof (aero-) space assets.

The improved CME forecasts would be beneficial alsofor electrical power systems because the most severe inci-dences of geomagnetic disturbances are almost alwaysassociated with magnetospheric evolution in response tothe stresses imparted by the variations in the solar wind.In many cases support would come also for the

technologies in communication and navigation systems,although these applications would still need to toleratethe direct disturbances from solar eruptions which comein the form of light and particle bursts.

Below, we first describe some promising opportunities insolar and heliospheric research that support the efforts forCME forecasts with lead times well in excess of 1 hr.Harvesting these opportunities will naturally generate alsosome new challenges, not just in solar/heliospheric researchbut also in geospace research as there additional work isneeded for efficient utilization of the improved character-ization of solar driving. Luckily, there are some opportuni-ties also on the geospace side that we should now seize inorder to gain better understanding of certain processes inthe magnetosphere-ionosphere-thermosphere system thatcontrol the severity of SWx disturbances in the threeimpact areas. Challenges and opportunities in the researchon factors controlling large GMD and GIC are addressedafter the discussion of solar and heliospheric matters(Sections 4.1 and 4.2). In Sections 4.3 and 4.4 we discussthe tasks for improved forecasts and environment


specification for navigation and positioning and (aero-)space assets. Main points of our reasoning are presentedin the diagrams of Figs. 2 and 3.

The diagram in Fig. 3 does not only address the case ofCME prediction (post-eruption), it also describes sometasks associated with the attempts to identify active areason the solar surface even before their eruption in order toachieve forecasts with lead times of several days.Although our current scientific understanding does notprovide adequate support for such predictions with goodenough confidence levels, it is good to keep also this needin the picture as a long-term goal. The track for forecastswith lead times beyond a few days is described below inSection 4.5. Specification of extreme conditions andsolar-cycle forecasts are discussed in Section 4.6.

4.1.1. Solar surface

Once space weather phenomena need to be forecast withlead times that have the solar-wind driver beyond the Sun–Earth L1 site that lies one million miles upwind from ourplanet, the observations required to make such forecastsshift to the origins of the space weather that lie withinthe solar corona and the innermost domains of the helio-sphere. The reason for that shift all the way towards theSun is that we do not presently have the technologicalmeans to station a spacecraft on the Sun–Earth line some-where halfway the Sun and the Earth (solar sails are beingconsidered, but have yet to be demonstrated to betechnologically feasible), nor do we have the financialresources to maintain a large fleet or near-Sun orbiters withat least one always near the Sun–Earth line (although con-stellations of spacecraft have been considered for this).Given this situation, we conclude that forecasts beyondone hour for solar-wind-driven magnetospheric and iono-spheric variability are necessarily based on observationsof the domain near the Sun and on heliospheric prop-agation models from there out to the planets, unless novelobservational techniques or instrumentation can be devel-oped and deployed.

The timing of the impacts of heliospheric storms on geo-space is currently generally forecast based on corona-graphic observations from which propagation speeds anddirections are estimated. Observations from the L1 sentinelpoint are nowadays provided by the LASCO coronagraphonboard the aging ESA/NASA SoHO mission (launched in1996), aided by the STEREO coronagraphs and helio-spheric imagers at least when in appropriate phases of theirorbits around the Sun. These observations enable estima-tion of arrival times that have recently become accurateto within a quarter to half a day, while ongoing modelingof the background solar wind enables a rough estimate ofthe densities and shock strengths of the incoming storm.Coronagraphic imaging (now routinely in use, but gener-ally not extending to close to the solar surface) or alterna-tively high-sensitivity high-altitude coronal EUV imaging(which appears technologically feasible) is therefore criticalto forecasts of solar-wind driven geomagnetic disturbances

and related phenomena in the geospace on time scales of afew hours to a few days. The potential loss of L1 coron-agraphy as now performed by SoHO’s LASCO is thus aweak link in the currently available inventory of space-weather inputs, until a replacement coronagraph islaunched or until alternative means of obtaining helio-spheric observations become routinely available.

When looking to forecast CME arrival times, one long-standing technique (Hewish, 1955) that is now maturing forspace weather purposes is interplanetary scintillation (IPS;Hewish et al., 1964). This uses the scintillation patternsobserved in the signals received from astronomical radiosources. This scintillation comes from the scattering ofradio waves by small-scale plasma density irregularitiespropagating with the bulk solar-wind outflow. Thus, IPSprovides another means of remotely sensing the solar wind.As a complement to, say, white-light heliospheric imagerssimilar to those currently aboard the STEREO spacecraftwell off the Sun–Earth line, IPS can provide informationon CME speeds and thus on CME arrival times. Whencombined with modeling techniques and/or in-situ data,other parameters such as CME masses can also beobtained, along with CME propagation directions andarrival times (e.g., Bisi et al., 2010, and references therein).

Another technique that is being developed to map theapproach of the CME is based on loss-cone anisotropyof ground-based muon observations: behind the shock (ifpresent) and inside the CME there is a cosmic-ray densitydepleted region (associated with a Forbush decrease) thatsome times results in precursory signatures observableupstream of the shock (1 to 7 h ahead). A detailed studyhas recently been undertaken by a group of Canadianexperts (Trichtchenko et al., 2013) who have delivered aroadmap for development of a pre-operational system.

Successful prediction of the arrival times of heliosphericperturbations is but one factor in the forecast of the mag-nitude or duration of geomagnetic perturbations. Criticalfor the forecasting of magnitude and duration of GMDsis the knowledge of the magnetic field structure withinapproaching solar-wind perturbations that none of theabove methods can provide. Most important in determin-ing the magnitude of the geospace response (be it immedi-ate or delayed) is commonly said to be the direction of themagnetic field at the leading edge, while the trailing struc-ture affects the storm duration. There is, however, as yetonly a poor understanding of the coupling of the helio-spheric field into geomagnetic activity and the data showthat more characteristics of the solar-wind field are defi-nitely required (e.g., Newell et al., 2007).

As a prospect for the future, a radio technique known asFaraday rotation (FR) may offer a means of remotely sens-ing the heliospheric magnetic field. The FR technique isalready used by astronomers to remotely sense galacticmagnetic fields and there are now growing efforts to applythe technique to study the heliospheric field. This is a sig-nificant research challenge but may be within reach withcutting edge radio telescope technologies, such as


LOFAR and those being developed for the SquareKilometer Array (SKA).

Despite some potential for the future, for the near-term,however, the only path towards successfully forecasting theintensity and duration of strong magnetospheric storms, ofrelated ionospheric current and electron density variability,and of GICs beyond the next hour or so, requires that wecan model the field that left the Sun. Specifically, we needto develop the means to model the internal properties ofthe configuration of twisted magnetic field that was injectedinto the heliosphere in the earliest phases of a CME and thesubsequent evolution of that field configuration in its inter-action with the high-coronal and inner-heliospheric mag-netic field into which it is thrust.

Establishing the detailed geometry and strength of thecoronal magnetic field prior to eruptions (c.f. the yellowbox at the top of Fig. 3), be it from strong-field compactactive regions or from weak-field extended quiet-Sunregions, is an area of active research. For active regions,the key ingredients to this are high-resolution magneticfield observations of the solar surface layers and thenear-surface (chromospheric) layers, and high-resolutionEUV imaging in narrow thermal regimes of the overlyingcorona (flagged as “opportunities” in Fig. 3).Polarimetric imaging is used to derive the vector-magneticfield, while coronal imaging provides the pathways of coro-nal loops that outline the magnetic field. At present, fieldextrapolations based on photospheric vector-magneticmaps are limited to phases of slow evolution, often usingthe approximation that all forces within the magnetic fieldare balanced (in the so-called non-linear force-free approx-imation in which curvature and pressure-gradient Lorentzforces cancel against each other). But even when limitingthis modeling to phases of slow evolution, models com-monly fail to reproduce the observed coronal loop geome-try, revealing the sensitivity of the field to boundaryconditions above, below, and in the vicinity of an eruptingsolar region, or to the assumed initial conditions, or both(e.g., DeRosa et al., 2009). For weak-field regions, vectorfields at the surface and near-surface are difficult to estab-lish, so X-ray or EUV imaging and coronal polarimetricmeasurements sensitive to field direction (see, e.g., Bk-Ste�slicka et al., 2013) may be used to constrain the mag-netic field and, as called out as an opportunity for largetelescopes (such as DKIST) polarimetric measurementssensitive to coronal magnetic field strength may also beused in future. In general, it appears that combined mea-surements of the surface magnetic field and of the coronalfield as traced by its plasma are needed to significantlyadvance beyond present-day limited and often unsuccessfulfield modeling (for example, Malanushenko et al., 2014).

The most urgent challenge (labeled accordingly “C” inthe top yellow box in Fig. 3) to be tackled to achieve reli-able forecasts of intensities and durations of geomagneticstorms on time scales of a few hours to a few days is thusthe realization of algorithms and observations needed for

coronal field modeling. These algorithms require at leastfull-Sun magnetograms combined with active-region scalevector-magnetic data and coronal high-resolution nar-row-band imaging to provide crisp images of the thermalplasma glow: that combination provides information onthe solar surface magnetic field and, though the tracingof the coronal field in the coronal emission patterns, alsoon the field in the coronal volume above. Tracing the coro-nal structures requires not only good angular resolution,but also separation of images into narrow thermal bandsso that structures stand out well from their surroundings.Single-perspective imaging may be shown to be adequate,but binocular imaging, necessarily involving an observa-tory off the Sun–Earth line by approximately 10–20�,would substantially aid in correctly interpreting the3-dimensional coronal field structure that is otherwise seenonly in projection against the sky. Thus a second point ofview for coronal imaging substantially off the Sun–Earthline is identified as offering the potential of a big leap for-ward in space-weather forecasting on the hours to daystime scale. Surface and near-surface magnetic field mea-surements and EUV coronal imaging need to be continued,because longer-term forecasts are not feasible withoutthese.

4.1.2. Heliosphere

Intermediate to the initial phases of solar eruptions andthe geospace response lies the vast expanse of the surround-ing corona coupled with the inner heliosphere (gray boxesin Fig. 3): first the CMEs propagate through the surround-ing coronal field that is strong enough to modify the pri-mary direction of the initiating eruption up to a few tosome ten solar radii, then they interact with the pre-existingmagnetized solar wind into which the eruptions plow ontheir way to the outer heliosphere. Here lies what is pre-sently a challenge in the form of the development of adata-driven heliospheric propagation model that mustinclude the properties of the magnetic field and that is, ide-ally, modified by ingestion of information on the helio-spheric state by the STEREO spacecraft or bytomographic methods relying on interplanetary radio scin-tillation. The development and validation of such codesemerged as challenge in our impact tracings, that can besuccessfully dealt with after local coronal field modelingof the state before and after eruptions has been developedto provide input to such models, and if the global coronalfield is known well enough to know how the nascent CMEmay have been modified and deflected even before enteringthe heliosphere proper. Only once such propagation mod-els are available based on the inner-boundary initial condi-tion of what has left the solar domain can the heliosphericmodel be fully tested against the observables obtained atL1 or, for the purpose of model testing and validation, atany other point within the extended heliosphere, be it clo-ser or further from the Sun, be it by planetary or helio-spheric missions with in-situ instrumentation.


Along with the development of this heliospheric MHDmodeling capability must come the test of whether it is ade-quate for these few-day forecasts to measure only theEarth-facing side of the solar magnetic field or whethermore extended surface coverage of line-of-sight orvector-magnetic field is required to adequately model thefoundation of the solar wind based on the lowest-ordercomponents of the coronal field that are determined byfield on all sides of the Sun, including its hard-to-observepolar regions. Similarly, it should be established howimportant observations are of the inner heliosphere offthe Sun–Earth line as data to be assimilated: are corona-graphic observations from well off the Sun–Earth line (suchas from around the Sun–Earth L5 or L4 regions that trailand lead Earth in its orbit around the Sun by some 40 to80 degrees where spacecraft can maintain a relatively stableposition with respect to Sun and Earth) critically impor-tant, or useful as guides, or are they mostly superfluousprovided that Earth-perspective data and advanced coro-nal and heliospheric modeling are available? Dependingon the outcome of the study of that question, coronaland inner-heliospheric imaging, such as done by theSTEREO spacecraft in some phases of their orbits, shouldbecome structural ingredients of the space-weather scienceand forecasting infrastructure. We note that the imple-mentation of a mission for such a perspective would alsohelp address the 180-degree ambiguity in the direction ofthe vector-magnetic component transverse to the line ofsight that currently is an intrinsic problem of even the mostadvanced solar spectro-polarimetric instrumentation.

4.2. Challenges and opportunities for geomagnetic

disturbances

4.2.1. L1 observations: Validation of >1 hr forecasts and

interaction with the magnetosphere

On a par with establishing the structure of the eruptingsolar field as it begins to propagate into the heliosphere isestablishing how incoming solar-wind field structures inter-act with the magnetospheric field to drive geomagneticvariability either immediately or with delay after storingenergy in the magnetotail. L1 in-situ particle and field mea-surements are critical in providing input to models fornear-term forecasts, and for validation of longer-termforecasts of storms advancing through the heliosphere.Therefore, the continuity of L1 measurements should beensured so that besides the needs of operational servicesalso the needs from the research community are taken intoaccount.

4.2.2. Reconfigurations in the magnetosphere–ionosphere

system and strong GICs

The coupled magnetosphere-ionosphere (MI) systemresponds to the solar wind through processes of energyinput at the magnetopause, followed by transport and stor-age in the magnetosphere, and finally release of such storedmagnetic energy from the magnetospheric tail either to the

inner magnetosphere, to the ionosphere/thermosphere, orejection into the solar wind outward through the distantmagnetotail. Recent discoveries from the THEMIS andCluster missions reveal the tantalizing and complex physicsof the release of such stored energy via the developmentand penetration of earthward propagating azimuthally nar-row channels of high speed plasma (bursty bulk flows,BBF) that transport magnetic flux and embedded plasmaearthward. On the other hand, nearer-Earth dynamics thatcouple these flows to the inner edge of the plasma sheet andto the flow-braking region to the ionosphere produce largefield-aligned currents and hence GICs that are still inade-quately understood. In particular, the conditions leadingto, and the exact physical processes responsible for, largefield aligned currents (FACs) to reach the ionosphere anddrive large GICs are not known.

At times, most of the magnetic energy in the mag-netosphere is stored for one to several hours, to be releasedin one drastic and intense substorm, leading to harmfulGIC effects. At other times, a large portion of the energyis transported earthward and immediately, but gradually,released through a more steady process causing quite mod-erate GIC effects. For major solar wind drivers there willalways be a resulting magnetic storm, but even such a mag-netic storm is made up of several interacting and overlap-ping release mechanisms including extreme convection,recurrent substorms, and recurrent plasma injections tonear-Earth region in a variety of combinations, thus mak-ing the prediction of exactly when GIC effects will occur,and exactly how intense they will be, very difficult.

It has been recognized that the exact path of energy dis-sipation in the MI system is dependent on details in thetemporal development and other characteristics of thesolar wind driver disturbance, the pre-history of solar windconditions before the arrival of a major event, and the pre-conditioning of the ionosphere, thermosphere, plasmas-phere, radiation belts and, indeed, the magnetotail itself.For example, the composition of low-ionization heavy(ionospheric) ions versus high-ionization low-mass (solar-wind) ions in the magnetotail depends entirely on thepre-event history of magnetospheric activity, but plays animportant role in the release of stored solar-wind energyin the form of substorms, through various instabilities.Likewise, the readiness of the ionosphere (and the magneticfield lines connecting the ionosphere to the magnetosphere)to redirect magnetospheric current through ionosphericchannels is pre-conditioned by auroral precipitation thatmodifies conductance during the substorm growth phase,when the energy stored in the magnetotail is at first onlypartially released.

The challenge for understanding large GICs hencerequires a deeper understanding of the dynamic evolutionof the system level state of the coupled near-Earth mag-netosphere–ionosphere–thermosphere (MIT) system.Especially important is the ability to understand – and todistinguish between – more gradual dissipation of energyand those events associated with explosive release of stored


energy. However, the physical processes that distinguishbetween extremely rapid and more gradual energy releaseare very poorly understood. A number of important pro-cesses for energy transport and dissipation are known,but their inter-relationships and the pre-existing state ofthe system which is the most likely to lead to large GICsis also not known.

Almost certainly the main driver of near-Earth taildynamics is the conversion and release of stored magneticenergy in that region. The NASA magnetospheric multi-scale (MMS) is targeted to address details of the kineticphysics leading to fast magnetic reconnection and therelease of earthward flows mainly in the distant tail at25 RE (RE is the symbol used for a distance equal to oneEarth radius of 6371 km). However, the nature of thedynamical processes operating at the critical transitionregion between dipole-like (closed) and tail-like (stretched)magnetic fields (between about 6 RE and 12 RE) remains tobe properly explored, because this region has been mark-edly undersampled by previous missions. We have seenglimpses of some of the processes that couple the tail tothe ionosphere; these indicate the importance of activeMIT coupling in determining the overall ability of the sys-tem to drive large GICs: some times localized intensifica-tions in auroral arcs are seen prior to substorm onset,which indicate changes in FACs close to the location ofwhere the substorm onset occurs later. The exact characterof pressure gradients at the earthward edge of the plasmasheet certainly also plays an important role in this cou-pling. Perhaps most significant is the fact that it is currentlyimpossible to distinguish between events which only resultin small-scale expansion of bright and dynamic aurorae,often called pseudo-breakups, and those which lead tolarge and rapid reconfigurations of the tail and full sub-storm development with bright aurorae covering the wholemidnight sector of the auroral oval. Likely plasma sheetkinetic plasma instabilities play an important role, but theirability to communicate with the ionosphere, in particularthe overall nature of Alfven wave exchange with the iono-sphere, and the balance between Ohmic dissipation and theacceleration of field-aligned electrons, and their conse-quences for destabilizing the tail and driving large FACsand hence GICs are not known.

From the GIC perspective, whether local instabilities atthe inner edge of the plasma sheet precede and communi-cate with the near-Earth neutral line, or whether globaltail reconfigurations begin with reconnection in the mid-tail is hardly relevant. The critical missing understandinglies in the near-Earth MI coupling processes that allowlarge currents to be driven in the ionosphere, with theconsequent production of large GICs. Thus the iono-spheric ability to close major FACs in the substormcurrent wedge (SCW, a current system where themagnetospheric dawn-to-dusk currents from magnetotailare diverted to the midnight sector of the auroralionosphere) is crucial for the understanding significantGIC effects.

In order to predict the exact route of energy dissipation,or even the timing of the onset of violent energy releaseafter a certain storage period, it is absolutely essential tounderstand and consequently improve modeling on theprinciples of pre-conditioning in the coupled MIT system(c.f. “challenge” in the blue box of the left column inFig. 3). To this end, satellite- and ground-based data arerequired from all parts of the coupled system (flagged as“opportunity” in Fig. 3), both in (and below) the iono-sphere, the inner magnetosphere and the magnetotail(including the plasma sheet where the instabilities leadingto energy release take place), and the lobe region in whichthe plasma sheet is embedded, where the energy storageand release can be monitored. Overall, this leads to require-ments for multi-point characterization of the plasma in thetransition region between dipole and tail-like fields,together with multi-point monitoring of the medium-alti-tude region (at several thousand kilometers) of auroralacceleration and wave reflection on conjugate field lines,with extensive support from under-lying multi-instrumentground arrays of magnetometers, optical instruments,and HF radars, etc. Monitors of the solar wind and ofincoming flows in the tail, perhaps utilizing existing assetsare also required.

4.3. Research for improved forecasts of ionospheric storm

evolution

User communities of GNSS and HF communicationsystems have expressed needs both for nowcasts/near-termpredictions and for longer-term predictions with multi-hour to multi-day lead times (cf., background bars inFig. 2). The research and development work to fulfill theseneeds should address both rapidly evolving steep gradientsin the ionospheric electron density (global and regionalscales) and processes causing ionospheric scintillation(local and micro scales). Forecasts of ionospheric per-turbations and of upper-atmospheric variability driven bysolar flaring and CMEs rely on the advances describedabove in the two previous sections. Ionospheric per-turbations driven by either solar flares or CMEs and asso-ciated geomagnetic storm or substorm phenomena requireflare forecasts of strength and onset time of flares and ofCME arrival times and of CME field properties.Statistical ionosphere models tuned with data fromground-based networks or low Earth orbiting (LEO) satel-lites can in many cases provide relatively good results fornowcasts or short-term forecasts in regional or globalscales (labeled as “opportunity” in the blue box of middlecolumn in Fig. 3). However, with increasing lead timesmore comprehensive physics-based modeling with dataassimilation is required. We need to understand what fac-tors control electron density variations during ionosphericstorms, including neutral atmosphere composition tem-perature and dynamics, ion-neutral coupling, solar EUV,particle precipitation, plasma transport, and ion chemistry(“challenge” in Fig. 3). Operational use of data-driven


models requires the availability of various actual observa-tion data sets covering the solar energy input includingthe associated energy spectrum (solar spectral irradiance),the knowledge of magnetospheric key parameters such asFACs and the convection electric field, thermospheric con-ditions such as composition and winds, and current state ofthe ionosphere such as the electron density distribution.

The largest challenges in ionospheric modeling are mostprobably associated with the physics causing scintillation inradio signals. Theoretical studies addressing the growthand decay of plasma instabilities should be continued togain better understanding on processes causing scintillationat high latitudes and at equatorial latitudes. As these areasprovide different background conditions for the instabili-ties, for example due to different internal magnetic fieldorientation and solar forcing, there is no single generalapproach to address them both.

Ground-based observations of the Earth’s upper atmo-sphere have been expanded extensively in the last decade,especially with new coherent and incoherent radars, newionosondes, new all-sky photometers, and new digital ima-gers from several different generations, mainly set close andaround the magnetic equator and at high latitudes toobserve and study plasma bubbles and auroral instabilitiesand their driving mechanisms. Despite such increases in theobservations of these space weather phenomena, there arestill scientific issues that are not solved, including, forexample, the day-to-day forecast of plasma bubbles as wellas teh day-to-day neutral wind that drives theirdevelopment.

4.4. Steps for improved radiation belt forecasts and

specification

Near-real time solar-wind conditions are the controllingfactors for the near-Earth energetic particle population,but the preceding evolution of that system is also critical;the result is an energetic particle environment in whichthe outer radiation belts are dominated by locally acceler-ated particles. For example, efficient energetic particle pro-duction mechanisms appear to need a seed population ofenergetic electrons. Relativistic electrons enhancementsare an important space weather factor with a strong influ-ence on satellite electronics. Around half of all magneticstorms are followed by a corresponding enhancement ofrelativistic electron fluxes. However, GCRs- and SEPscan also penetrate into the terrestrial magnetosphere.SEPs can originate either by direct acceleration in the cor-ona, which can result in very prompt fluxes in the case ofdirect interplanetary linkage, or by acceleration at inter-planetary shocks, from where SEPs can propagate towardsEarth through direct magnetic-field linkage or by beingcarried within the CME so that their SEP populationsmay eventually pass by, and then temporarily envelop,Earth.

SEPs and GCRs originate from within the solar systemand beyond, in the Galaxy, respectively, and propagate to

the Earth’s orbit. The geomagnetic field controls the pene-tration of SEP and GCR particles into the near Earthenvironment, such that lower energy particles penetrateonly into polar regions (for example, energies below about1 GeV penetrate only poleward of about 55� magnetic lati-tude), but progressively lower latitudes are penetrated asenergies increase (an energy approaching 20 GeV isrequired to penetrate to the equator; Beer, 2010). Thusthe high-energy GCR particles penetrate much more deeplythan do most solar particles; only rarely do SEPs have suf-ficient energy to reach deep into the equatorial regions. Thelatitude to which particles penetrate is known as the geo-magnetic cutoff and depends on both particle propertiesand on the geomagnetic field strength. This geomagneticcutoff filtering is important for satellite and aviation opera-tors, as it controls the particle properties to which their sys-tems are exposed.

GCRs and very high energy SEPs that penetrate deepinto the atmosphere collide with atmospheric species toproduce neutrons, muons and other secondary particles.Ground-level and aircraft measurements of these sec-ondary particles give useful insights into the very high-en-ergy part of SEP energy spectra. However, the transportof this particle radiation through the magnetosphere andatmosphere needs to be better understood; measurements,especially on aircraft, are needed to challenge and stimulateGCR and SEP radiation transport models.

Trapped energetic particles in the inner magnetosphereform radiation belts. The inner proton belt is fairly stable,but the outer radiation belt comprises a highly variablepopulation of relativistic electrons – with fluxes oftenincreasing abruptly but generally decaying only slowly.Solar-wind changes produce changes to trapped-particletransport, acceleration, and loss. However, the pre-existingmagnetospheric state is also important and therefore manyof the challenges and opportunities listed above for theGMD/GIC case are relevant also for improved under-standing on SWx effects in the radiation environment.Continuous solar-wind and solar observations are neededto predict magnetospheric disturbances and correspondingauroral precipitations, field-aligned current variations,trapped particle variations and any SEP penetration.

Intensification of magnetospheric convection duringstorm times, local particle acceleration due to substormactivity, and resonant wave-particle interaction are themain fundamental processes that cause particle energiza-tion and loss. However, the details of the mechanismswhich may be able to contribute to these processes –including a variety of wave particle interactions, enhancedconvection and radial diffusion – remain a subject foractive research. While progress has been made in the theo-retical understanding of these competing processes, there isas yet no consensus on which of these will be significant inparticular situations, and no real predictive methods thatcan give precise fluxes at different local and universal times.There is therefore a pressing need to confront theoreticalmodels with detailed measurements in order to resolve


these shortfalls (see right hand column in Fig. 3). Dataincorporation is needed to take into account local plasmaprocesses affecting the magnetic field variations andcorresponding particle accelerations, losses etc. As muchas possible data are needed for nowcasting of energetic par-ticle fluxes. GEO, LEO, MEO, GTO data are very impor-tant. Data calibration is needed to provide the correctresult. It is essential for satellite operators to know the past,current and future condition of the space environmentaround their satellites, especially in case of satellite anom-aly and/or before critical operation.

Finally, there are short-lived populations of ring-currentparticles produced by substorms and enhanced mag-netospheric convection. Currently, our understanding ofsubstorm onset is evolving rapidly, in particular with respectto local dipolarization structures. However, we are still along way from an ability to predict precise events. Currentspacecraft measurement configurations are probably notsufficient to make accurate predictions. Because these effectsare the major contributor to spacecraft failures due tocharging, it is essential that current investigations aremaintained and extended (reflected as highest-priorityopportunity in Fig. 3). Magnetic field models (empirical,numerical) with particle tracing codes are needed to now-cast/forecast particle distribution. They require current orpredicted solar-wind information for nowcasting/forecast-ing. Data sources are solar wind monitoring data and solarUV images, models of solar wind propagation from the Sun.

4.5. Challenges for forecasts with lead times beyond 2 days

Forecasts of substantial space weather events that extendbeyond a few days require, by definition, forecasts of ener-getic events well before they occur on the Sun. This requiresadvancing our understanding of the storage and instabilitymechanisms for the solar magnetic field. Many of the issuesinvolved in that will be addressed by the science described inSection 4.1.1, including knowledge of the 3D structure ofthe field and estimations of energy and helicity budgets.With the advancement of that understanding will also comethe ability to experiment with the field configurations todeepen insights into the conditions under which such fieldsbecome unstable for flares and CMEs (see Appendix D formore discussion and a listing of requirements). Support ofthe research described in Section 4.1 will thus also supportpre-event forecasts for that reach some hours to a daybefore the event occurs on the sun.

For forecasts of CME arrivals well beyond the few daysthat it takes to travel through the heliosphere to reachEarth, this means that forecasts need to be made evenbefore the source regions of potentially geo-effectiveCMEs have rotated from near the east limb towards theEarth-perspective central meridian. In contrast, promptSEP storms around Earth often originate from solarregions that have rotated several days past central-meridian passage towards the west limb. This adds somecomplication in that the magnetic maps of such regions

suffer from perspective changes, but that can be dealt withas long as the regions are not too close to the limb (i.e.,within about 4–5 days of central meridian passage). Whatis a much larger challenge here is that in many cases suchlonger-term forecasts for active regions need to be madeeven before much of the involved field has emerged ontothe solar surface. It appears that field configurations lead-ing to the most intense space weather originating in active-region flares and eruptions form and decay within a day ortwo of the emergence of twisted bundles of magnetic flux orthe shearing by convective flows of such bundles. Onceemerged and interacting with the pre-existing field, theavailable non-potential energy can be processed eitherthrough a rapid conversion in a flare/eruption or can begradually dissipated, likely contributing somehow to thethermal energy of the solar atmosphere (therein findingan intriguing analogy with the options for magnetosphericrelaxation discussed above).

The successful quantification of the subsurface magneticfield that will be involved in possible solar eruption uponemergence by helioseismic techniques currently lies beyondour capabilities, and even beyond a proof of principle inadvanced helioseismic magneto-convective simulations. Itthus appears premature to invest in instrument opportuni-ties solely for helioseismic investigations of the eastern/leading solar limb for the purpose of exploring the poten-tial of, say, five-day pre-event forecasts. We do recom-mend, however, continued support for the developmentand testing of helioseismic methods that may reveal howwe can detect and characterize magnetic flux bundles aboutto emerge onto the solar surface a day or more ahead oftime. The background solar-wind sector structure involvedin recurring geomagnetic perturbations is driven mostly bythe rotation of the largest-scale field structures on the Sun.The forecasts of the sector structure and the non-CME-re-lated patters of fast and slow solar wind streams will bene-fit from the advanced knowledge of the global solar fieldand the associated heliospheric modeling described inSection 4.1.2. The more extensive the observational magne-tograph coverage is of the solar active-region belt, the bet-ter multi-day forecasts will become for the overall solarwind stream and sector structures. As the requirementsfor this overlap with those articulated in Section 4.1.2, wedo not repeat these here.

4.6. Specification of extreme conditions and forecasts of the

solar cycle

Quantitative knowledge of low-frequency, high-impactspace weather cannot be derived by observing the Sun–Earth system in modern times, simply because too few ofthe severe events have occurred, and the most extrememay not yet have been observed at all as yet. Rather, wemust gather data that tell us about solar activity over manycenturies in order to reach useful, reliable conclusionsabout events that happen once per century or once per mil-lennium. For energetic particle storms and the GCR


background, such information can be obtained extendingover tens of millennia by combining terrestrial and lunarradionuclide studies based on rocks, ice cores, and bio-sphere-modulated radio-nuclide records (including, forexample, carbon-14). To quantify the frequency spectrumfor solar flares we can look at the multitude of Sun-likestars accessible by nighttime ground-based or space-basedastronomical telescopes, as demonstrated by, for example,NASA’s Kepler mission and by X-ray and EUV spaceastrophysics missions (Schrijver and Beer, 2014). Currentevidence suggests that even for our aging Sun events thatare tens of times more intense than the most energetic solarones observed during the space era may occur on timescales of once per century to millennium. We recommendthat the solar-stellar links be strengthened to better quan-tify the probability distribution of the most intense butinfrequent solar flares by using archival data on Sun-likestars already available and by supporting possible futureopportunities to add to that statistical information.

As to energetic particle storms, it is prudent to make themoderate investment needed to harvest the radio-nuclideinformation stored in ice, biosphere, and the rocks broughtback from the Moon during the Apollo era. Parts of thisoverall project involve improving our knowledge of the dis-tribution of energies over particle spectra and photon spec-tra, as well as understanding how terrestrial particle spectracan be combined with photon spectra to understand howsolar/stellar flares and particle storms are related for themost intense and dangerous events.

An entirely different class of problems is to be addressedfor multi-year to multi-decade variability that is driven bythe general patterns of the solar cycle. Information on pastsolar cycles is, of course, coming from studies of sunspotpatterns, of radionuclide studies associated with GCRmodulations by the large-scale structure of the solar wind,and even from studies of potential climate impacts in thepre-industrial era on time scales from decades to millennia.The inter-disciplinary work that is required for that shouldbe stimulated as an aid to environmental specification fordesigns of long-lasting infrastructures, both in space andon the ground. Multi-year to multi-decade forecasts, onthe other hand, need to see investments in understandingthe solar dynamo. The observational requirements for thisinvolve continued observations of the solar magnetic fieldand of the flows involved in transporting it across the solarsurface and throughout its interior. Here, continued mag-netic observations as required for the science described inSection 4.1 supports the need for surface coverage.Continued helioseismic observations to study the variabil-ity of deep flow patterns involved in the dynamo areneeded in addition. Moreover, strengthening ties with theastrophysical community looking into dynamo activity inSun-like stars is needed.

Involvement in the study of habitability of exoplanets isa natural stimulus for both the extreme event knowledgeand improved dynamo understanding because exoplanetspace weather is a dynamic emerging field that can benefit

from Sun–Earth connections knowledge as much as its dis-coveries can inform us about extremes and long-termtrends in space weather in our own planetary system. Weencourage stimulus of these research fields, but as theselie at the fringe of our central charge, we do not at presentmake any explicit recommendations for investments inresearch and its observational infrastructure, other thanemphasizing that funding agencies would be well advisedto use the synergy between solar and stellar research andbetween planetary-system and exoplanet habitability toimprove our abilities to specify and predict extremespace-weather conditions and long-term trends.

5. General recommendations

With so much yet to understand, and so much space (lit-erally) to cover observationally and in computer models,the needs readily outweigh the means we can expect to haveat our disposal. A major effort of the roadmap team there-fore focused on identifying where investments would makethe largest impact in advancing our scientific insights tobest meet the needs of the space weather users. We basethat analysis on these identified primary needs:

� For the research community: Comprehensive knowledgeof conditions throughout the Sun–Earth system in thepast based on models guided by observations, and theunderstanding of the physical mechanisms involved indetermining these conditions.� For the space-weather customer community:

Knowledge of historical and recent conditions andknowledge of current conditions and forecasts for com-ing hours to days, and specification of extreme condi-tions in geospace and throughout interplanetary space.Different user communities require different types ofdata and distinct levels of accuracy and specificity.

From these follow primary requirements:

� Comprehensive, affordable, sustained observational cov-erage of the space weather system from Sun to society.� Data archiving, sharing, access, and standardization

between the various researcher and customer communities.� Advanced data-driven and experimental models as tools

for analysis, interpretation, and visualization of eventsand their spatio-temporal context, and for forecasting/specification.� Interaction and coordination between national, regio-

nal, and international researcher, user, and agencycommunities for research guidance, prioritization,impact assessment, and funding.� Education of researchers, customers, and general public.

These requirements need to be addressed in the context of achanging paradigm in the science of space weather:whereas many studies to date tend to focus on what theyrefer to as “space weather events”, there is a rapidly


growing realization that in fact significantly longer timesequences need to be studied because there appears to bean important if not crucial impact of what is eitherdescribed as hysteresis or as pre-conditioning in the system.This is now recognized to reach from the solar environmentto deep within geospace: solar active regions emerge pref-erentially in locations where other such emergence hasoccurred before; nascent CMEs often plow through highcoronal field that is still relaxing from preceding eruptions;interplanetary solar-wind structures are often mergers ofmultiple sequential CMEs; geo-effectiveness of CMEs isdependent on pre-conditioning of the magnetosphere byactivity over the preceding days; substorms appear torelease stresses in the magnetotail built-up over some per-iod of time; preceding conditions in the ionosphere controlthe timing and intensity of substorm onsets; and so forth.Consequently, there must be a shift away from the studyof short time intervals that attempt to study space weatheras if pre-conditioning were unimportant, moving towardsthe analysis of multi-day windows of space weather thatmakes allowance for significant effects of hysteresis in theSun–Earth system and its components.

The impact tracings discussed in the preceding section,and detailed in Appendices B,D and E, lead us to formulatethe following general priorities for actions that guide us tothe detailed implementation suggestions for developmentsof new observational, computational, and theoreticalcapabilities that are discussed in the subsequent section.We begin with a concise listing of the priorities for threedistinct target audiences, followed in the next threeSections 5.1, 5.2, 5.3 by a point-by-point expansion intothe types of actions to be considered or implemented.

Research: observational, computational, and theoreticalneeds:

1. Advance the international Sun–Earth system observatoryalong with models to improve forecasts based on under-standing of real-world events through the development ofinnovative approaches to data incorporation, includingdata-driving, data assimilation, and ensemble modeling.The focus needs to be on developing models for theSun–Earth system, at first as research tools focusing onactual conditions and later to transition to forecast tools,making use of the existing system before components arelost as instrumentation fails or is discontinued.

2. Understand space weather origins at the Sun and theirpropagation in the heliosphere, initially prioritizingpost-event solar eruption modeling to develop multi-day forecasts of geomagnetic disturbance times andstrengths: Advance the ability to forecast solar inputsinto geospace at least 12 h ahead based on observationsand models of their solar drivers as input into helio-spheric propagation models.

3. Understand the factors that control the generation ofgeomagnetically-induced currents (GICs) and of harshradiation in geospace, involving the coupling of the solar

wind disturbances to internal magnetospheric processesand the ionosphere below: Advance the ability to fore-cast the response of the geospace system to driving bysolar-wind variability.

4. Develop comprehensive space environment speci-fication: Create a reference specification of conditionsand their likelihoods for the local cosmos.

Teaming of research and users: coordinated collaborative

environment:

I Quantify the vulnerability of technological infrastruc-ture to space weather phenomena jointly with stake-holder groups.

II Build test beds in which coordinated observing sup-ports model development: Stimulate the developmentof (a) state-of-the-art environments for numericalexperimentation and (b) focus areas of comprehensiveobservational coverage as tools to advance under-standing of the Sun–Earth system, to validate forecasttools, and to guide requirements for operationalforecasting.

III Standardize (meta-) data and product metrics and har-monize access to data and model archives: Definestandards for observational and model data products,for data dissemination, for archive access, for inter-calibration, and for metrics. Define data sets neededto test physical models and forecast systems.

IV Optimize observational coverage: Increase coverage ofthe Sun–Earth system by combining observations withdata-driven models, by optimizing use of existingground-based and space-based resources, by develop-ing affordable new instrumentation and exploringalternative techniques, and through partnershipsbetween scientific and industry sectors.

Collaboration between agencies and communities:

A Implement an open space-weather data and informationpolicy: Promote data sharing through (1) open datapolicies, (2) trusted-broker environments for access tospace-weather impact data, and (3) partnerships withthe private sector.

B Identify, develop, and provide access to quality educa-tion and information materials for all stakeholdergroups: Collect and develop educational materials onspace weather and its societal impacts, and create andsupport resource hubs for access to these materials,and similarly for space-weather related data and dataproducts. Stimulate collaboration among universitiesin order to promote homogenized space-weather educa-tion as part of the curricula.

C Execute an international, inter-agency assessment of thestate of the field to evolve priorities subject to scientific,technological, and user-base developments to guideinternational coordination: Identify an organizationalstructure (possibly the COSPAR/ILWS combination


leading to this roadmap) to perform comprehensiveassessments of the state of the science of space weatheron a 5-year basis to ensure sustained development andavailability of high-priority data, models, and researchinfrastructure. This activity can be a foundation to alignthe plans of research agencies around the world if, as forexample for this roadmap, agency representatives areengaged in the discussions.

D Develop settings to transition research tools to opera-tions. Establish collaborative activities to host, evaluate,and compare numerical models (looking at theCommunity Coordinated Modeling Center [CCMC]and the Joint Center for Satellite Data Assimilation[JCSDA] as examples, setup, staffed appropriately fromresearch and user communities) and to assess quantita-tively their skill at forecasting/specifying parameters ofhigh operational value. Determine the suitability ofresearch models for use in a space weather service cen-ter. Foster continuous improvement in operationalcapabilities by identifying the performance gaps inresearch and operational models and by encouragingdevelopment in high priority areas.

E Partner with the weather and solid-Earth communitiesto share infrastructure and lessons learned. To improveunderstanding of the couplings between weather andspace-weather variability and to quantify potential cli-mate impacts by effects related to space weather.Another type of partnership here involves the transferof knowledge and “lessons learned” from the climate/weather communities on techniques for data-drivenassimilative ensemble modeling and on the developmentof forecasts and their standards based on that. For thesolid-Earth community: translation of geomagnetic vari-ability into electric fields involved in GICs.

The above “General Recommendations” (listed in prior-ity order within each target group) as formulated by thisroadmap team align with many found in other studiesand workshops (e.g., Committee on Progress andPriorities of U.S. Weather Research and Research-to-Operations Activities, 2010; American MeteorologicalSociety, 2011; EC Joint Research Center by Krausmann,2011; Committee on a Decadal Strategy for Solar andSpace Physics, 2012; United Nations Committee on thePeaceful Uses of Outer Space, 2014). This roadmapadvances from the general and system-wide recommenda-tions in those earlier reports by including a rationale forthe prioritization of investments for the worldwide commu-nity subject to limited resources.

5.1. Research: observational, computational, and theoretical

needs

1 Advance the international Sun–Earth system observa-tory along with models to improve forecasts based on

understanding of real-world events through the

development of innovative approaches to data incor-

poration, including data-driving, data assimilation,

and ensemble modeling.

The Sun–Earth system is currently observed by anunprecedented array of ground- and space-based instru-ments that together provide a comprehensive view, butone that is so sparse that the loss (through failure or ter-mination) of any one key asset would have a substantialimpact on the overall study of space weather phenomena.Note that our recommendation on access to the data fromthis system observatory is listed above as item A, andexpanded upon in Section 5.3, item A.

Recommendation: Funding agencies should place thecontinuation of existing observational capabilities withinthe context of the overall Sun–Earth System observatoryas a very high priority when reviewing instrumentationfor continued funding: the value of any one observablewithin the entire set of Sun–Earth observables should sig-nificantly outweigh whether the instrument has met itsown scientific goals or whether it, by itself only, merits con-tinuation based on its standalone scientific potential. Theprioritization of international assets should be taken froma community-based assessment, such as this roadmapand its recommended successors. Of particular importanceto be continued are: Earth-perspective magnetic maps andX/EUV images, solar spectral irradiance monitors (includ-ing X-rays to UV and radio), coronagraphy (ideally frommultiple perspectives), solar radio imaging, in-situ (near-)L1 measurements of the solar wind plasma (including com-position) and its embedded field and energetic particles,ground-based sensors for geomagnetic field and iono-spheric electron density variability and neutral-atmospheredynamics, LEO to GEO electron and ion populations, in-situ magnetospheric magnetic and electric fields, and geo-magnetic field measurements, space-based auroral imagers,and energetic-particle sensors (including ground-basedneutron monitors) where available throughout the innerheliosphere.

Most space weather models, whether for research or forforecasts, currently rely on snapshots or trends in observ-ables, or on extrapolations from one location to modelconditions in another, with little or no direct guidance byobservations of evolving conditions at the boundariesand in the interior of the modeled volume. This is in partbecause data incorporation strategies are only in theirinfancy, and in part because data types and model algo-rithms are not optimized for assimilation processes.

Recommendation: Funding agencies should immediatelybegin to give preference to modeling projects that use, orare specifically designed to eventually enable, direct or indi-rect guidance by observational data. Funding agenciesshould require efficient online access to relevant data setsby standardization of interfaces and by the creation ofinterface hubs and tools that transform data sets into stan-dard formats that are compatible with common practices indifferent disciplines. Each agency can focus on its owndata, but agencies should coordinate in the development


of data standards and standardized interface, emphasizingthe importance of near-real-time data availability. TheILWS program partnership can be a coordinating hub inthis process.

2 Understand space weather origins at the Sun and their

propagation in the heliosphere, initially prioritizingpost-event solar eruption modeling to develop multi-

day forecasts of geomagnetic disturbance times and

strengths, after propagation through the heliosphere.

This priority is associated with two scientific focus areas.The first is the need to enable a much-improved quan-tification of the non-potential magnetic field of sourceregions of solar activity pre- and post-eruption, for whichwe propose local-area binocular imaging as well as high-resolution comprehensive imaging that includes vector-magnetography of the solar surface and chromosphere onthe scale of the erupting region. The second involvesadvanced modeling capabilities for the coupling from thelocalized source-region corona into the overall heliosphere,and the computation of the propagation of plasma andembedded magnetic field through the heliosphere towardsgeospace, for which we propose increased coverage of thesolar surface magnetic field, and advanced modelingcapabilities for the global coronal and inner-heliosphericdynamic magnetic field and solar-wind plasma. The specificrecommendations associated with this research focus areaare presented in detail in Sections 6, 7.1 and 7.3.

3 Understand the factors that control the generation of

geomagnetically-induced currents (GICs) and of harsh

radiation in geospace, involving the coupling of the solarwind disturbances to internal magnetospheric processes

and the ionosphere below.

This priority is associated with several recommenda-tions. The first involves satellite coverage in the domainbetween the dipole-tail transition region at the inner edgeof the plasma sheet and in the near-Earth domain of themagnetospheric field (at the auroral acceleration region atseveral 1000 km altitude) to probe, along essentially thesame magnetic field lines, the processes that throttle begin-ning magnetotail activity and that determine if that activitymay develop into a magnetospheric (sub-) storm and asso-ciated intense GICs or not. The second recommendationfor this priority emphasizes the need for coordinatedground-based and space-based networks of instru-mentation, and the development of “testbed” locationswhere dense observational coverage provides informationon key distinct environments for geomagnetic and iono-spheric electron density variability (specifically for the aur-oral zone and the near-equatorial region). The twocomplementary sets of recommendations are described indetail below in Sections 6, 7.2, and 7.4–7.6.

In addition to these recommendations regarding mag-netospheric and ionospheric processes driven from outsideand within the M–I system, there are the processes that sub-ject the ionosphere to forcing by the neutral terrestrial

atmosphere for which we articulate this specific additionalrecommendation: Recommendation: Funding agenciesshould place the continuation of existing observationalcapabilities within the context of the upper atmosphere aswell as the computational capability that leads to its under-standing for the scientific and operational purposes at a highpriority. It also recommended that the coverage of theground-based instrumentation be coordinated in the devel-opment of new instrumentation, including integration ofground- and/or space-based data systems. Also, it is recom-mended to include that lower-atmospheric information intodata-driven modeling capabilities for the near-real-timeionospheric state. Additionally, the scientific communityshould pursue a general consensus on the underlying physicswith the processes associated with radio scintillation, in par-ticular about the coupling between neutral atmosphericwinds and waves, and development of the plasma turbulencethat leads to the bubble formation at equatorial latitudes.

4 Develop a comprehensive space environment

specification:

With robotic and manned spacecraft becoming evermore frequent and venturing into new orbits aroundEarth and the solar system, it is important that we establishin detail the conditions to be encountered by these space-craft prior to their design so that they can be built to sur-vive space weather storms and be resilient to recover fromsuch storms. Similarly, design criteria for technologicalinfrastructures used on Earth need to be set so that theycan resist, survive, and recover from effects of space storms.This requires information on space weather conditions thatmay be encountered during the lifetime of such spacecraftand ground-based systems.

Recommendation: We recommend for the near term thatdetailed specifications be compiled of all relevant space-weather phenomena (following the recommendation foran open-data policy formulated below) that may beencountered wherever human technologies are, or likelywill be, deployed; this includes properties of particlestorms, solar irradiance variability, and geomagnetic vari-ability, from low-level generally present backgrounds torare, extreme, impulsive events. We recommend for thenear- and mid-term that research funding be allocated tothis effort that includes studies of radionuclides in the ter-restrial biosphere and cryosphere, as well as in lunar andterrestrial lithospheres; that analogies between Sun andSun-like stars be explored; and that advanced modelingbe deployed to understand extremes in geomagnetic, helio-spheric, and solar conditions.

5.2. Teaming of research and users: coordinated

collaborative environment

I. Quantify the vulnerability of technological infrastruc-

ture to space weather phenomenaAs our ability to understand general properties of space

storms advances, more refined forecasts will come within

8 https://www.sievert-system.org/?locale=en.


reach: ranges of forecast magnitudes may shrink, starttimes and durations of events may be better forecast, andcharacterization of variability within a storm will beincreasingly feasible. But the value of such forecasts isdefined by the impacts that these and other attributes ofstorms have on our evolving technologies. Therefore, stud-ies of the impacts compared to observed properties shouldbe performed. Such impact studies will also guide ourunderstanding of impacts of high-impact, low-frequencyextreme events by narrowing the range over which we needto extrapolate to model their impacts before they occur.

Knowledge and understanding of space weather impactsremains uncertain in all impact areas of space weather. Forexample, GICs cause potentially the largest space weatherimpacts on terrestrial infrastructure, but the disturbanceand failure modes of power grids subjected to GICs remainpoorly known. For example, work in South Africa, stimu-lated by the GIC problems experienced there in October2003, shows the need for better management of trans-former life cycles against all forms of stress and aging:strong GIC events can significantly age transformers, thusadvancing the date on which they will fail (e.g., Gaunt,2013), unless replaced in good time. Such quantificationof the impacts of space weather events can be guided bythe equivalent of epidemiological studies of grid perfor-mance subject to geomagnetic variability (e.g., Schrijverand Mitchell, 2013; Schrijver et al., 2014), combined withsystematic engineering and impact studies.Epidemiological studies should be executed in partnershipsbetween space scientists and power-grid experts, whiledetailed impact studies should be led by the power sectorwith involvement of the science community (ideally withinternational coordination). Such statistical and case stud-ies are needed to understand the time sequence and struc-ture of the magnetospheric field dynamics that are themost geo-effective in driving GICs.

Another example is that there is a big gap between theknowledge of the space environment and the statistics ofsatellite anomalies. One aspect of this is that the occurrenceof a satellite anomaly is strongly dependent on the materialspecification of a particular satellite. A complicating factorin the study of satellite anomalies due to space environmentis that the correlation is often based on indirect evidence.Unfortunately, detailed information on satellite anomaliesis generally not released by companies that design or oper-ate satellites.

A third example of insufficient information is related tothe impacts of radiation on crews and passengers of airlin-ers and on the aircraft avionics. The UK Royal Academyof Engineering study on space weather (Royal Academyof Engineering, 2013) stressed the importance of ground-testing of avionics against exposure to neutrons of energiestypically found at 10–12 km altitude; this report noted thatappropriate test standards are being developed as are testfacilities capable of testing whole avionic systems, e.g.,the ChipIR facility in the UK. Other work in this areaincludes an aircrew dosimetry system operated by the

French Institute for Radiological Protection and NuclearSafety.8

Studies of this type do exist but are often restricted intheir feasibility or in their distribution and availability tothe general space-weather stakeholder community by con-cerns expressed as competition sensitive, proprietary infor-mation, or even national security. Removing or reducingthese barriers is an important step to be taken towardsunderstanding the true impacts of space weatherphenomena.

Recommendation: We recommend for the near term thatthe space weather community engage with research agen-cies responsible for the study of economic and societaldevelopments and encourage them to explore and supportresearch to quantify the impact of space weather phenom-ena on societal technologies. Where these agencies areseparate from space organizations, the community shouldencourage them to work with space organizations. In allcases, the support for research on impacts should enableand stimulate the trans-disciplinary research that isrequired, and should also encourage a data-sharingenvironment in which industries are given confidence thatthey can share information on space-weather impacts with-out concerns for their commercial competitiveness or theirpotential liabilities. To achieve this, we recommend that allresearch proposals on impact studies should be stronglyencouraged to combine the research and user communitiesand their joint expertise and data. We also recommend thata trusted-broker environment be given shape by a groupthat involves stakeholders from government, academia, aswell as industry. The community should also encourageinternational coordination of impact studies.

II. Build test beds in which coordinated observing supports

model development

Some domains within the Sun–Earth system are coveredbetter by observations than others. For example, there aresites for ionospheric research that are favorably locatedand well equipped with particularly valuable or state-of-the-art instrumentation. Coordination between the variousspace- and ground-based solar observatories in observingcampaigns of particularly active space-weather sourceregions on the solar surface offer another opportunity fora “laboratory” with extensive empirical coverage of regionsof interest. Test-bed data could be used to learn whichtypes of observations are most suitable for data incor-poration in modeling and which resolutions in space andtime would be suitable compromises when consideringboth instrument maintenance costs and additional gainfrom data ingestion.

Recommendation: We recommend for the mid-term thatinternational resources be optimally pooled, particularly insettings where significant investments have already beenmade: ground-based observatories of the ITM system inJicamarca and around EISCAT appear well suited for

http://https://www.sievert-system.org/?locale=en


focused further development; international coordinationbetween solar observatories (in space and on the ground)should be mandated by funding agencies for a fraction ofthe total observing time available; data assimilation effortswith the continuously developing models should be sup-ported by access to large-scale computer facilities (e.g.,the IBM ExaScience Lab in Belgium) with national, EU,NSF or space agency funding. Results from test runsshould be published in platforms that enable communitybased code development and testing (e.g., NASA CCMCand ESA virtual space weather modeling center) in orderto harvest theoretical and computational skills from theentire SWx community.

To make a step change in the international co-ordina-tion, and delivery, of space weather observing systems wemust increase the engagement of scientists working withthis instrumentation with modelers, operators of space-based sensors, and other consumers of space weather data.That should be aimed to ensure the optimum interactionbetween the collected data and state-of-the-art interna-tional models, and the synthesis of these data and modelresults into, first, research tools and, later, into operationaltools whose outputs can be made available to the applica-tions and technology community. In order to improve thesituation, the geospace observing system needs a more for-mal international structure to deliver its full scientificpotential.

Recommendation: We recommend establishing a globalprogram at inter-agency level for coordination of spaceweather observing systems, perhaps similar to coordinationof space exploration activities.

III. Standardize (meta-) data and product metrics and

harmonize access to data and model archives

Making efficient use of data, and even finding data thatare in principle useful in advancing the science of spaceweather require adequate standardization in data formatsand access protocols to on-line data archives. Knowledgeof quality of data products and of space weather applica-tions requires development of data product standards.

Recommendation: We recommend that funding agenciesrequire development of data archiving, data search, anddata access plans as part of observatory/instrument opera-tions support, retro-actively where needed; this shouldinclude standardization of data formats and of data retrie-val methods (i.e., web and application interfaces). Werecommend that funding agencies coordinate any neededreprocessing of historical datasets to make them generallyavailable in a standard way. We recommend that a studybe commissioned, possibly under the auspices of WMO,COSPAR, and/or ILWS, to address issues related to thelong-term preservation of data archives. We also recom-mend that agencies resolve issues that hamper use ofscience data or products that are not at all available forcommercial use (such as the Dst index) in an environmentwhere tailored space weather services may increasingly beprovided by commercial entities. Finally, we recommend

that standard metrics are developed for data product qual-ity and applications quality.

IV. Optimize observational coverage

The space from Sun to Earth is vast and poorly coveredby observatories. Part of that problem is limited access toexisting information – such as ionospheric information thatis available to, but not effectively shared by, the GNSS nav-igational service sector - while another part is related to thecost of instrumentation, particularly in space. The highcost of space missions percolates into several concerns:high costs lead to low frequency of deployment which (a)causes instrumentation to be deployed one or two decadesbehind the technological frontier, and (b) slows or pre-cludes filling of gaps in the vast space to be covered andin terms of observables that are accessible. The science ofthe Sun–Earth connections that underlie space weather isparticularly affected by this problem because the systemto be observed is vast and the number of instruments verysmall in comparison. But precisely because of that, it is aparticularly suitable area for which this issue should beaddressed because any new instrumentation is deployedas an augmentation to an existing Sun–Earth system obser-vatory so that even relatively small new ground-basedobservatories or space missions can significantly strengthenthe overall system observatory as it advances that into ascientifically much more capable state.

Recommendation: We recommend that research organi-zations and industry sectors partner to optimize use ofalready available data related to space weather impacts;central government organizations (such as OSTP in theUS) should play a leading role in fostering this sharing ofresources. Recommendation: We recommend that spaceagencies urgently seek to lower the costs to achieve theirscience goals by (a) emphasizing partnerships and effectiveuse of small opportunities within the context of the Sun–Earth system observatory, (b) by developing infrastruc-tures and rule sets that enable lower-cost access to spacein the mid- and long-term, and (c) by implementing sus-tained diversification of their research fleets with smallniche satellites and clusters of smallsats (perhaps evenmicrosats surrounding a larger mother ship with centralcoordination and communication abilities), all to enablemore frequent insertion of new observatories (standaloneand as hosted payloads) that can better utilize state-of-the-art technologies, where possible in “off-the-shelf”configurations. In working towards these goals, researchagencies should take care to maintain the most importantobservational capabilities in the currently operatingdistributed Sun–Earth system observatory, ideally in coor-dination with, and in consultation with, the research com-munity to most efficiently advance the system observatory’scapabilities.

As part of this recommendation, we note also thatefforts should be made to increase the availability ofground-based data on the geomagnetic field and on thestate of the ionosphere with high timeliness. This can be


accomplished by: (i) considering the deployment ofmagnetometers and other low cost ionospheric instru-mentation (radio wave based probing techniques) inregions with limited coverage; (ii) utilizing the WMO datainfrastructure to disseminate data from existing instru-mentation; and (iii) working with providers of proprietarydata to allow their data to be used in space weather prod-ucts. We also emphasize the opportunity to expandground-based coverage by using relatively small facilitiesthat enable emerging countries to become involved in spaceweather research.

5.3. Collaboration between agencies and communities

A. Implement an open space-weather data and informa-

tion policy

Open access to relevant space weather data will stimulateresearch by enabling innovative uses of observations of theSun–Earth system, of modeling data relevant to that sys-tem, of impact data for space weather phenomena, and ofsocietal consequences of such phenomena. An open datapolicy is already agreed upon in the G8 “Open DataCharter” (2013) between governments, but we argue for awider adoption of that charter beyond the G8 nations,and to explicitly define “open data” to mean that datashould be open to all users as soon as such data is archivedfor use any research group. Funding agencies supportingestablishment and maintenance of SWx infrastructuresshould favor such initiatives that follow this open data pol-icy. A general open data policy, supported by an infrastruc-ture of virtual observatories, will stimulate research ingeneral, in particular by enabling the utilization of ancillaryand adjacent data to observations being analyzed by aresearch group. Recommendation: We recommend thatobservational data pertinent to space weather be made pub-licly accessible as soon as such data is calibrated adequatelyfor scientific use. Data sets obtained with routine instru-ment setups or for monitoring or survey purposes shouldbe mandatorily archived and opened up for immediateinternet access without proprietary periods starting as soonas data are archived and provisionally calibrated andcharacterized by meta-data. We also recommend that datasharing be a formalized part of observatory planning, fund-ing, and operating, including sharing between space-basedand ground-based observatories and instrumentation.

Recommendation: Agencies that fund instrument opera-tions should provide sufficient resources to enable efficientopen data access - perhaps through a mix of direct fundingto instrument operators to generate and store data prod-ucts and their metadata, and funding for data infrastruc-tures that catalog and provide access to these products.Virtual observatories and/or general software (such asexists for SolarSoft IDL) should be sustained and devel-oped to ease data identification and retrieval from thearchives. Recommendation: We recommend that a similaropen-data policy for assimilation model data be developedand implemented where practical in the mid-term.

B. Identify develop, and provide access to quality educa-

tion and information materials for all stakeholder

groups

There is huge public interest in space weather and itsimpact on humankind, in part because it is a spectacularnatural phenomenon, and in part because it is a risk thatappeals to the human love of scare tales. As a result theinternet provides access to a vast and ever-growing multi-tude of web sites and web pages related to the topic ofspace weather (see Fig. 1). However, this information isof very variable quality and many web pages contain sig-nificant factual errors in both the science and the impactof space weather. These errors frequently propagate intoother web sites, into the media, and even into the discus-sion of space weather by policy makers. Therefore, it iscritically important to guide all stakeholders of spaceweather towards high-quality information on spaceweather; these stakeholders include policy-makers, reg-ulators, educators, engineers, system operators, scientists,and many more. The wider scientific community is also acritical target, especially the closely related scientific areassuch as plasma physics, radiation physics, planetary explo-ration, and astrophysics. Furthermore, the increasing pub-lic interest on space weather is stimulating an increasedfocus on Sun–Earth connections as part of university leveleducation. In this relatively young and specialized researchdiscipline, the level and focus of education may varybetween different universities, which can cause com-plications in student exchange or in recruitment of earlycareer scientists.

Recommendation: We recommend the development ofan international structure that can coordinate a globaleffort to identify and/or develop high-quality informationthat addresses the needs of all types of stakeholders inspace weather, and does so in a way that facilitates cus-tomization of information to local (national/regional) cir-cumstances, especially targeting diversification towardsculture and language. To deliver this we recommend astructure that engages the leading disseminators of spaceweather knowledge in each country, including space-weather science experts who have a good appreciation ofhow the science of space weather links into impacts onpractical systems of local importance, and of how theirnational structures handle both the science and the impactsof space weather. A key goal is to link with the most suit-able national structures, recognizing these will vary con-siderably from country to country. The structure shouldalso encompass dissemination activities in internationalbodies, including ISES, WMO, UN/COPUOS, ESA,IUGG, IAU, COSPAR, EGU and AOGS.

In the very near-term we recommend that a workinggroup (a) carry out a survey to identify and engage thenational experts and structures most active in developingand disseminating high-quality information, and then (b)propose an international structure that can coordinatetheir efforts on the tasks shown below. This working groupshould also engage with relevant international bodies and


ensure that the international structure can encompass theirdissemination work. The working group must also considerhow the international structure will develop the scientificauthority needed to deliver on the tasks below.

Once formed, the international structure should.

1. survey existing good practice in different countries andin international organizations, and encourage theexchange of ideas;

2. establish a procedure for validation of information,including a method for embodying that validation withinthe information, for example by a certification mark;

3. identify, validate and provide links to existing goodinformation that can assist education and awareness-raising amongst all types of stakeholders in spaceweather;

4. identify gaps in this information, promote efforts to fillthose gaps and provide links to validated outputs;

5. encourage adaptation of information to best suit theneeds of particular groups and countries, includingtranslation into local language and adaptation to localculture; and

6. compile information on sources of space weatherresearch data and of space weather applications and ser-vices, and provide links to these sources.

To support the worldwide dissemination of space weatherknowledge the international structure should establish aninternet-based system that can point to a range of validatedinformation on space weather covering the different needsof all types of stakeholders. Given the rapid evolution ofinternet technologies, we do not seek to prescribe howthe system be implemented, rather we recommend thatthe international structure, once established, review theoptions available, having regard to the need for a resilientand sustainable system, and for ease of use by contributorsand users around the world.

Consideration should be given on whether the interna-tional structure should seek to validate space weather ser-vices and data access. However, it may be best to avoidthis as it may open up significant legal issues, especiallyin respect of commercial services.

C. Execute an international, inter-agency assessment of

the state of the field to evolve priorities subject to

scientific, technological, and user-base developments

to guide international coordinationWith rapidly changing technologies and advancing

understanding the focus priorities of the science of spaceweather are likely to evolve over time, so that a periodicevaluation of the state of the field will prove useful to pro-vide guidance to the research organizations. Such roadmapstudies should identify options for agencies to align domes-tic and international priorities, to identify projects mostsuitable for direct collaboration, and to identify whereinternational activities might make domestic investmentsto address scientific priorities less urgent or superfluous.

Recommendation: We recommend that a roadmap for thescience of space weather be conducted at a 5-year cadence,ideally timed to provide information to national strategicassessments such as the Decadal Surveys in the US or to suchmultinational funding programs like EU H2020 programand its successors. These should be international projects,led by representatives from the scientific community, andinvolving at least representatives of all major space agenciesas well as research organizations working on the ground seg-ments of space weather. Moreover, agencies should regu-larly appoint inter-agency science definition study teamswith representatives from agencies, the science community,and the space-weather user community to find optimal bal-ance between scientific potential of new missions by them-selves and as part of the Sun–Earth system observatory,where the potential value as element of the system observa-tory should weigh heavily in its design and evaluation.

D. Develop settings to transition research tools to

operations

With the rapid advance of scientific knowledge and thecontinuous evolution of operational needs, it is essential thata constant cycle of improvement occur in space weather fore-casting/specification capabilities. This cycle involves thedevelopment and utilization of new capabilities that improveservices, and it involves the continuous feedback of the high-est priority operational needs. Research tools comprisescientific models and research (measurement) infrastructure.The latter may require investment in order to be consideredoperational with respect to data availability, timeliness,quality and reliability. Operational tools need to be consis-tent in order to comply with any given Service LevelAgreement and to be considered effectively operational.Recommendation: We recommend that a roadmap forresearch-to-operations (R2O) transition be conducted at a5-year cadence, timed to provide information to, and takeinformation from, a parallel science roadmap. This R2Oroadmap should focus on specific, high priority activitiesthat can achieve measurable progress on service metrics.

In order for research models to be considered suitablefor use in a forecast center, they should be confronted toa set of standardized evaluation criteria, such as: accuracyof output; confidence level and uncertainty estimation; sta-bility; maintainability and scalability; support and docu-mentation; autonomy and ease of use (forecasters andoperators may not be specialists).

Recommendation: We recommend that the internationalcommunity establish standardized practices and proce-dures to evaluate research models, similar to the GEM/SHINE/CEDAR challenges. Internationally agreed met-rics for model and forecast performance evaluation shouldbe established.

It is vital to implement and support reliable archiving ofdata collections and to guarantee streamlined access toarchives. Efficient dissemination of archived and near realtime data is a prerequisite for maintaining reliable serviceoperation.


Recommendation: We recommend that a common set ofspace weather metadata be developed in order to facilitateharvesting and interpreting data. In addition, the develop-ment of standardized interfaces to data repositories isstrongly encouraged, as well as common data analysistoolkits (standardized software packages). This shouldbuild on existing international programs that haveaddressed the provision of standard metadata for theSun–Earth domain, including the NASA-led SPASEproject, the ESPAS and HELIO projects in the EU FP7programme, ESA’s Cluster Active Archive, and theIUGONET project in Japan.

E. Partner with the weather and solid-Earth communitiesto share infrastructure and lessons learned

The uppermost troposphere and the thermosphere/mesosphere above it influence ionospheric processes.Moreover, space weather phenomena may couple intolong-term regional weather and climate, as has been pro-posed for cosmic-rays modulation and solar spectral irradi-ance variations.

Partnering with climate/weather stakeholders is ben-eficial also on another front: space weather scientists areonly beginning to learn how to assimilate observations intodata-driven ensemble models, and can learn much from themeteorological community. Space weather forecastinglooks to the example of terrestrial weather forecasting asa field with similar goals and much longer history andexperience. In particular, many of the advances in terrestrialweather forecasting have come about by the augmentationof physics-based models with advanced statistical methods,such as data assimilation and ensemble modeling. Theadvancement of space weather forecasting likely requiressimilar endeavors. However, we must recognize that forsome regimes (such as in Earth’s ionosphere) themeteorological techniques can be borrowed with littlemodification, while in other regimes (particularly wherethe data is sparse) new techniques will need to be developed.

Yet another reason for partnerships with weather andsolid-Earth sciences could be the sharing of stations forgeomagnetic, magnetotelluric, and ionospheric measure-ments in the much-needed expanded web of ground-basedmeasurement locales.

Engagement with the solid-Earth geophysics communityis vital for space weather studies of GIC, since the geomag-netic induction of electric field inside the Earth is a criticalelement in the causal chain of physics leading to adverseGIC impacts on technological systems. The solid-Earthgeophysics community has excellent expertise in this area,since geomagnetic induction is also a key technique forexploring the sub-surface structure of the Earth down todepths of at least several hundred kilometers. Thus theirengagement in space weather studies is vital.

Recommendation: We recommend intensification andformalization of coordination between the ITM and cli-mate/weather research and modeling communities, specifi-cally (a) to develop, in the mid-term, whole atmosphere

models (i.e. coupled models ranging from the troposphereinto the high ITM domain), (b) to learn about data assim-ilation, ensemble modeling and other advanced statisticaltechniques in forecasting scenarios, (c) to partner in themulti-use of ground stations for both terrestrial and spaceweather, and (d) to analyze possible pathways by whichspace weather could impact Earth global or regional climate.In the US, NSF/UCAR is the natural focal point for thisresearch, coordinated with NASA/SMD Heliophysicsresearch in ITM physics, while organizations like WMO orIUGG may be engaged as an enabling coordinator and asan existing multi-national organization for intergovernmen-tal communication. We also recommend strengthening tiesto the solid-Earth community to support transformationof geomagnetic variability to electric fields involved in GICs.

6. Research: observational, numerical, and theoretical

recommendations

The highest-priority scientific needs that were only verybriefly mentioned in Section 5.1 are presented in moredetail in this section, and complemented by specific instru-mentation and model concepts in Section 7. In this Section,we identify the needs (a) to maintain existing observationalcapabilities, (b) to develop modeling capabilities, enablearchival research, or improve the data infrastructure, and(c) to construct new space-based and ground-based instru-mentation. By their nature, these three classes of invest-ments can be implemented (a) immediately, (b) on a timescale of a year or two, or (c) on a time scale of about 4to 10 years.

The recommendations formulated in this section aretaken from the impact tracings (discussed in Appendix B)that were initially produced looking separately at thesolar-heliospheric domain and at the geospace domain (assummarized in Appendices D and E, respectively). In thissection we merge all those requirements, sorting them intothree general clusters of recommendations that we shallrefer to as pathways of research. Pathway I collects theneeds for the full Sun–Earth domain that are related tothe need to obtain forecasts more than 12 hrs ahead ofthe magnetic structure of incoming CMEs to improvealerts for GICs and related ionospheric variability.Pathway II will need at least part of the new knowledgedeveloped within Pathway I to forecast the particleenvironments of (aero-) space assets and to improveenvironmental specification and near-real time conditions.Pathway III is a particularly challenging one, aiming toenable pre-event forecasts of solar flares and CMEs, andrelated solar X-ray, EUV and energetic particle stormsfor the near-Earth satellites, astronauts, ionospheric condi-tions, and polar-route aviation, including all-clear condi-tions. From the viewpoint of solar physics, Pathway Icontains those requirements that are most feasible to dowithin next five years. From the viewpoint of geospaceresearch Pathway I contains the requirements linked withthe advancements to be achieved from solar physics and


support especially GIC/GMD. With the requirements ofPathway I fulfilled, we should be able to improve GIC/GMD forecast in all conditions. Also in GNSS and aero-space domains significant improvement will be achieved,but the issues involved will not be fully solved in Pathway I.

From the viewpoint of solar physics, Pathway I containsmuch the same things as Pathway II. The geospace require-ments are mostly separate in Pathway II because theyaddress largely complementary physics in geospace, whilethey will gain significant benefit from the solar and helio-spheric work on the variable solar wind supported byPathway I. Aero-space assets are here the main impactarea, but the models discussed in Pathway II will supportthe efforts to understand D-layer absorption (caused byRB energetic particle precipitation) and thus they will helpalso with HF communication.

Recommendations in Pathway III are from the view-point of solar physics much more challenging than thoseof Pathways I and II because they concern forecasts beforeany substantial event has occurred on the Sun.

Pathway I: To obtain forecasts more than 12 h ahead ofthe magnetic structure of incoming coronal mass ejectionsto improve alerts for geomagnetic disturbances and strongGICs, related ionospheric variability, and geospace ener-getic particles:


� magnetic maps (GBO, SDO), X/EUV images at arcsecand few-second res. (SDO; Hinode), and solar spectralirradiance observations;� solar coronagraphy, best from multiple perspectives

(Earth’s view and L1: GBO and SoHO; and well offSun–Earth line: STEREO);� in-situ measurements of solar wind and embedded mag-

netic field at, or upstream of, Sun–Earth L1 (ACE,SoHO; DSCOVR);� for a few years, measure the interaction across the bow-

shock/magnetopause (as now with Cluster/ARTEMIS/THEMIS; soon with MMS), to better understandwind-magnetosphere coupling;� satellite measurements of magnetospheric magnetic and

electric fields, plasma parameters, soft auroral andtrapped energetic particle flu variations (e.g., VanAllen Probes, LANL satellites, GOES, ELECTRO-L,POES, DMSP);� ground-based sensors to complement satellite data for

Sun, heliosphere, magnetosphere, and iono-/thermo-/mesosphere data to complement satellite data.

Archival research, develop data infrastructure, or modelingcapabilities:

� near-real time, observation-driven 3D solar active-re-gion models of the magnetic field to assess destabiliza-tion and estimate energies;

� data-driven models for the global solar surface-coronalfield;� data-driven ensemble models of the solar wind including

magnetic field;� data assimilation techniques for the global ionosphere-

magnetosphere system for nowcasts and near-term fore-casts to optimally use and to coordinate ground andspace based observations to meet user needs. Comparemodels and observations, ideally in select locationswhere laboratory-like test beds exist or can be developedat a few informative latitudes.� coordinated system-level research into large-scale rapid

morphological changes in the Earth’s magnetotail andembedded energetic particle populations (using datafrom, among others, SuperDARN, SuperMAG,AMPERE, etc.);� system-level study of the mechanisms of the particle

transport, acceleration, and losses driving currents andpressure profiles in the inner magnetosphere;� stimulate research to improve global geospace modeling

beyond the MHD approximation (kinetic, hybrid, . . .)� develop the ability to use chromospheric and coronal

polarimetry to guide full-Sun corona-to-heliospherefield models.


� binocular imaging of the solar corona at �1-arcsecondand at least 1-min. resolution, with �10–20� separationbetween perspectives;� observe the solar vector field at and near the surface and

the overlying corona at <200-km resolution to quantifyejection of compact and low-lying current systems fromsolar active regions;� (define criteria for) expanded in-situ coverage of the aur-

oral particle acceleration region and the dipole-tail fieldtransition region (building on MMS) to determine themagnetospheric state in current (THEMIS, Cluster)and future high-apogee constellations, using hosted pay-loads and cubesats where appropriate;� (define needs, then) increase ground- and space-based

instrumentation to complement satellite data ofmagnetospheric and ionospheric variability to coverobservations gaps (e.g., in latitude coverage or overoceans);� an observatory to expand solar-surface magnetography

to all latitudes and off the Sun–Earth line [for whichthe Solar Orbiter provides valuable initial experimentalviews];� large ground-based solar telescopes (incl. DKIST) to

perform multi-wavelength spectro-polarimetry to probemagnetized structures at a range of heights in the solaratmosphere, and from sub-active-region to global-cor-ona spatial scales;� optical monitors to measure global particle precipitation

[such as POLAR and IMAGE] to be used in data


assimilation models for GMD and ionosphericvariability.

Pathway II: For the particle environments of (aero)space assets, to improve environmental specification andnear-real-time conditions. In addition to the remote-sens-ing and modeling requirements for Pathway 1:


� develop particle-environment nowcasts for LEO toGEO based on observations of electron and ion pop-ulations (hard/�MeV and soft/�keV; e.g., GOES, . . .,striving for intercalibrated data sets with better back-ground rejection, for at least a solar cycle), and of themagnetospheric field [see GMD/GIC recs.];� maintain a complement of spacecraft with high resolution

particle and field measurements and defined inter-space-craft separations (e.g., the Van Allen Probes).


� specify the frequency distributions for fluences of ener-getic particle populations [SEP, RB, GCR] for the speci-fic environment under consideration, and maintainaccess to past conditions;� develop, and experiment with, assimilative integrated

models for RB particle populations towards forecastdevelopment including ionosphere, thermosphere andmagnetosphere, including the coupling from lower-atmospheric domains, and validate these based on archi-val information.


� increased deployment of high- and low-energy particleand electromagnetic field instruments to ensure densespatial coverage from LEO to GEO and long term cov-erage of environment variability (including JAXA’sERG [Exploration and energization of radiation in geo-space; launch in 2015]. Combine science-quality and moni-toring instruments for (cross) calibrations, resolution ofangular distributions, and coverage of energy range.

Pathway III: To enable pre-event forecasts of flares andsolar energetic particle storms for near-Earth satellites,astronauts, ionospheric conditions, and polar-route avia-tion, including all-clear conditions. In addition to theremote-sensing and modeling requirements for Pathway I:


� solar X-ray observations (GOES);� observe the inner heliosphere at radio wavelengths to

study shocks and electron beams in the corona and innerheliosphere;

� maintain for some years multi-point in-situ observationsof SEPs off Sun–Earth line throughout the inner helio-sphere (e.g., L1, STEREO; including ground-based neu-tron monitors);� maintain measurements of heavy ion composition (L1:

ACE; STEREO; near-future: GOES-R).


� develop data-driven predictive modeling capability forfield eruptions from the Sun through the innerheliosphere;� investigation of observed energetic particle energization

and propagation within the inner-heliospheric field, aim-ing to develop at least probabilistic forecasting of SEPproperties [see also Pathway-1 recommendations forheliospheric data-driven modeling];� new capabilities for ensemble modeling of active regions

subject to perturbations, to understand field instabilitiesand energy conversions, including bulk kinetic motion,SSI, and energetic particles.


� new multi-point in-situ observations of SEPs off Sun–Earth line throughout the inner heliosphere to improvemodels of the heliospheric field and understand pop-ulation evolutions en route to Earth (e.g., SolarOrbiter, Solar Probe Plus).

7. Concepts for highest-priority research and instrumentation

In this section we present brief summaries of the ratio-nales for the (new or additional) research projects andinstrumentation investments given the highest priorities inthe preceding section. These summaries, expanded uponin In 6, are meant to illustrate possible investments thataddress the identified scientific needs towards an integratedapproach to the space-weather science. These investmentconcepts demonstrate possible approaches that we deemfeasible in terms of technological requirements and bud-getary envelopes, but we emphasize that science definitionteams focusing on the top-level scientific needs may wellbe able to identify and shape other options.

7.1. Quantify active-region magnetic structure to model

nascent CMEs

We identified a very high-priority requirement tocharacterize the solar-wind magnetic field, and in particularthe field involved in CMEs. To address that, in-situ mea-surements upstream of Earth at the Sun–Earth L1 sentinelpoint are insufficient. Moreover, technological means andbudgetary resources are inadequate to position in-situ sen-tinels on the Sun–Earth line sufficiently close to the Sun, or


to launch a fleet of moving sentinels with always one nearto that position. The variable solar-wind magnetic fieldassociated with active-region eruptions must consequentlybe obtained from forward modeling of observed solar erup-tions through the embedding corona and inner heliosphere,based on the inferred magnetic field of erupting structures.A key science goal in this roadmap is consequently thedetermination of the origins of the Sun’s impulsive, erup-tive activity that will eventually drive magnetospheric andionospheric variability. Deriving that from surface (vec-tor-) field measurements only has been shown to yieldambiguous results at best.

One way to constrain the 3D active-region field makesuse of novel modeling methods that can utilize the coronalloop geometry to constrain the model field, most success-fully when 3D information on the corona is available. Asdescribed in Appendix F.1, binocular imaging of theactive-region corona at moderate spatio-temporal res-olution enables the 3D mapping of the solar active-regionfield structure prior to, and subsequent to, CMEs, therebyproviding information on the erupted flux-rope structure.This could be achieved by a single spacecraft some 10� to20� off the Sun–Earth line if combined with, e.g., the exist-ing SDO/AIA or other appropriate EUV imagers, or – ifthese are no longer available – by two identically equippedspacecraft off the Sun–Earth line with approximately 10–20heliocentric degrees of separation.

A second new observational capability is also prioritizedin this context to derive information on low-lying twistedfield configurations in the deep interiors of unstable activeregions before and after eruptions is key to determiningwhat propagates towards Earth to drive space weather.The magnetic stresses in these regions that are involvedin CMEs cannot be observed directly, but require modelingbased on observations of the vector magnetic field at andimmediately above the solar surface (in the solar chromo-sphere), guided by higher structures that are observedwithin the overlying atmosphere (differentiated by tempera-ture from the low, cool, dense chromosphere to the high,hot, tenuous, field-dominated corona). With the help ofthese atmospheric structures that outline the magnetic fieldand the parallel electrical currents we can determine thefraction of the system’s free energy that is converted topower eruptions. Binocular imaging is unlikely to succeedfor very low-lying, compact magnetic structures withinactive regions. Moreover, chromospheric vector-magneticmeasurements in addition to photospheric ones areexpected to aid in the mapping of the 3D active region fieldas this provides observational access to electrical currentsthreading the solar surface, and observing the details ofthe low-lying flux ropes before, during, and after eruptionsto quantify the 3D field ejected into the heliosphere.

The information on the low-lying magnetic field and theelectrical currents within can be obtained by a space mis-sion (outlined in Appendix F.2) that obtains vector-mag-netic measurements of active regions at the solar surfaceand within the overlying chromosphere, combined with

optical and EUV imaging of the solar atmosphere from10,000 K up to at least 3 MK, all at matching resolutionsof order 0.2 arcseconds, to map all field structures thatmay carry substantial electrical currents.

7.2. Coupling of the solar wind to the magnetosphere and

ionosphere, and strong GICs

Like explained in Section 4, some key processes in thechain of magnetospheric energy release leading to largefield-aligned currents and hence GICs remain poorlyknown:

� the dynamics nearer to Earth at the inner edge of theplasma sheet where bursty bulk flows brake and presum-ably give rise to a number of more or less effectiveplasma instabilities� and the coupling of these processes into the ionosphere

along magnetic field lines.

In particular, the processes that control the rate of energytransport and the partition between competing routes ofdissipation in the coupled (M–I) system remain insuffi-ciently understood. We do not know the conditions whichcontrol the partitioning and exchange of energy betweencurrents, waves, and particles that are believed to act as agate for the ability of the tail to drive FACs through to clo-sure in the ionosphere

To uncover the processes in this interface domain, atwo-constellation satellite mission architecture is proposed(detailed in Appendix F.3), where the challenge lies not inadvanced instrumentation but in the positioning of a suffi-cient number of spacecraft at the two key locations inspace. The first constellation should focus on plasma insta-bilities and flow braking at the inner edge of the plasmasheet, in the transition region from dipolar to tail-like mag-netic fields. It would probe the three-dimensional plasmaand electrodynamic fields (E and B) in the transitionregion, from close to geosynchronous orbit to around10–12 RE during the course of the mission. This con-stellation could consist of a central mother spacecraftaccompanied by some 4 smaller spacecraft approximately1 RE from the mother, providing coverage in the azimuthaland radial directions.

The second, coordinated constellation is to focus on the(M–I) coupling in the auroral acceleration region on conju-gate auroral field lines below. It would provide multi-pointplasma and electrodynamic fields in the auroral accelera-tion region to determine the dynamical M–I coupling, ataltitudes of around 4000 km to 1 Re, in a configurationdesigned to resolve spatio-temporal ambiguity of the fila-mentary FACs and to distinguish between alternative par-ticle acceleration processes.

These constellations should be supported by conjugateauroral imaging from the ground, and if possible also fromspace, to probe both global and small scale (<100 km)structures. LEO satellites to additionally monitor the


precipitating electrons as a measure of conductivitychanges below will provide valuable complementary mea-surements. On the other end of the coupled system of themagnetosphere measurements of incoming flows in the cen-tral plasma sheet from the more distant tail are required toassess incoming flow characteristics in the same meridian.These could be provided by existing assets, such asGeotail, Cluster, THEMIS, or perhaps by MMS in anextended mission phase. An upstream solar-wind monitoris, as always, required as well. The success of the con-stellations proposed here, as is true for the overall geospaceobserving system, is crucially dependent of measurementsfrom rich networks of ground-based instruments, including(a) magnetometers to observe how electric currents in theionosphere are modified by space weather, plus (b) a widevariety of radar and radio techniques to monitor changes inthe density, motion and temperatures of ionospheric plas-mas, as well as (c) optical techniques to measure thermo-spheric winds and temperatures. These data sources areall key inputs to aid the identification of the onset locationand the resolution of the spatio-temporal ambiguity of theprocesses leading to large dB/dt. Ground-based data sup-port also the development of improved models of the atmo-sphere and its response to space weather, increasingly so aswe advance assimilative approaches. As outlined inAppendix F.4, we need to promote these ground-based net-works as a global system for scientific progress on spaceweather, so that each instrument provider (and funder) seeshow their contribution fits into the wider picture, i.e. that alocal contribution builds and sustains local access to a glo-bal system.

7.3. Global coronal field to drive models for the magnetized

solar wind

The Sun’s surface magnetic field is a vital ingredient toany predictive model of the global magnetic field thatdefines the structure of the heliosphere, including the posi-tion of the heliospheric current sheet and the regions of fastand slow solar wind, and plays a key role in space weatherat Earth: (1) the interaction of CMEs with the ambient fieldimpacts their geo-effectiveness; (2) the connection of theheliospheric magnetic field to CME-related shocks andimpulsive solar flares determines where solar energetic par-ticles propagate, and (3) the partitioning of the solar windinto fast and slow streams is responsible for recurrent geo-magnetic activity. In order to obtain the magnetic-fieldinformation that is needed to forecast the field arriving atEarth over a day in advance, the interaction of that fieldevolving from near the solar surface into the field movingthrough the corona-heliosphere boundary must be known.

Models for the global solar magnetic field typically usemagnetic maps of the photospheric magnetic field builtup over a solar rotation, available from ground-basedand space-based solar observatories. Two well-knownproblems arise from the use of these “synoptic” maps.First, such maps are based on solar conditions that lie as

much as 27 days in the past of an always-evolving surfacefield. Second, the field in the Sun’s polar regions is poorlyobserved, and consequently the high-latitude fields in thesemaps are filled with a variety of interpolation and extrap-olation techniques. These two observational problems canstrongly influence the global magnetic field model: poorlyobserved active regions near the limb (as viewed fromEarth), unobserved regions beyond the limb, and inaccu-rate polar field estimates can introduce unacceptable errorsin the high-coronal field on the Earth-facing side of theSun.

To address these observational problems, we need toobtain photospheric magnetograms off of the Sun–Earthline, particularly of the east limb (just prior to becomingvisible again on the Earth-facing side of the Sun that isthe portion of the solar field with the oldest observationsas viewed from Earth), to complement magnetogramsobtained along the Sun–Earth line by SDO and ground-based observatories; Appendix F.5 describes a missiontrailing Earth its orbit to provide the needed observations.Moreover, we need to obtain magnetograms of the Sun’spolar fields over a few years to understand the evolutionof the Sun’s polar magnetic flux. As described inAppendix F.5, the Solar Orbiter will provide some of theinitial measurements needed by providing at least a goodcalibration of the highest-latitude solar field; but the orbitof the Solar Orbiter is such that magnetogram informationof the solar far side as viewed from Earth will be providedonly on occasion with long gaps between relatively shortobserving intervals and with long delays before the obser-vations can be telemetered to Earth: magnetic field cover-age at a cadence fast enough to follow active regionevolution fully around the Sun is not possible with theOrbiter.

The improved magnetographic coverage of the solarsurface is needed to improve the models for the corona-heliospheric interface and to feed heliospheric magnetohy-drodynamic models: to yield meaningful predictions ofCME magnetic structure, it is necessary to utilize an accu-rate representation of the time-evolving global solar coro-nal magnetic field as input to models of prediction,eruption and propagation of CMEs through the solarwind. Such models ultimately will need validation by sam-pling their results near Earth, but before that they needguidance as to the physical approximations made in thesolar coronal field and plasma model. Such guidance comesfrom coronagraphic polarimetric observations, which weshall need to learn to absorb into global coronal-helio-spheric models.

New observations and modeling techniques are needed(as summarized in Appendix F.6): we propose to start withsimulated multi-wavelength coronal magnetometricobservables for constraining the global magnetic field, tobe incorporated into global MHD models. In order toreach this goal, we propose model testbeds (c.f.,Section 5.2, Recommendation II) for synthetic polarimetricmeasurements related to the Zeeman and Hanle effects,


gyroresonance and gyrosynchrotron radiation, thermal(free-free) emission, and coronal seismology. These simu-lated observables can be tested as drivers for, and ulti-mately assimilation into, global MHD coronal models.Guided by the outcome of these experiments, observationsneed to be obtained, most likely requiring new instru-mentation (see Appendix F.6).

The MHD models for the global corona are the natu-ral foundation for plasma and field models for the overallheliosphere. The latter are already being worked on,specifically for the background quiescent solar wind struc-ture, but they are not yet driven by the full magneticinput from actual solar conditions simply because thatinformation is yet to be derived as described above.Once that information is available, fully dynamic modelsfor the entire inner heliosphere need to be stronglysupported.

7.4. Quantify the state of the coupled magnetosphere-

ionosphere system

The response of magnetosphere to solar wind drivingdepends on the previous state of magnetosphere. Similarsequences in energy, momentum and mass transfer fromthe solar wind to magnetosphere can lead in some casesto events of sudden explosive energy release while in othercases the dissipation takes place as a slow semi-steady pro-cess. Comprehensive understanding on the factors thatcontrol the appearance of the different dissipation modesis still lacking, but obviously global monitoring of the mag-netospheric state and system level approach in the dataanalysis would be essential to solve this puzzle.Continuous space-based imaging of the auroral oval wouldcontribute to this kind of research in several ways. The sizeof polar cap gives valuable information about the amountof energy stored in the magnetic field of magnetotail lobes.Comparison of the brightness of oval at different UV wave-lengths yields an estimate about the energy flux and aver-age energy of the particles that precipitate frommagnetosphere to ionosphere. These estimates are not asaccurate as those from particle instruments on boardLEO satellites, but the additional value comes from thecapability to observe all sectors of the oval simultaneously.Such view is useful especially in the cases where the mag-netosphere is prone to several subsequent activations inthe solar wind. The shape and size of the oval and intensityvariations in its different sectors enable simultaneous moni-toring of magnetosphere’s recovery from previous activitywhile new energy comes into the system from a new eventof dayside reconnection.

There is consequently a need to achieve continuouslyglobal UV or X-ray images to follow the morphologyand dynamics of the auroral oval, at least in theNorthern hemisphere, but occasionally also in the southernhemisphere. Imager data combined with ground-based net-works similarly as suggested above in Section 7.2 allowssolving the ionospheric Ohm’s law globally which yields a

picture of electric field, auroral currents and conductanceswith good accuracy and spatial resolution. This wouldmean a leap forward in our attempts to understand M–Icoupling, particularly the ways how ionospheric conditionscontrol the linkage, e.g., by field-aligned currents (seeAppendix F.7 for more discussion).

7.5. Observation-based radiation environment modeling

Trapped energetic particles in the inner magnetosphereform radiation belts. Solar wind variability modifiestrapped particle transport, acceleration, and loss. The innerproton belt is relatively stable, but the outer radiation beltcomprises a highly variable population of relativistic elec-trons. Relativistic electrons enhancements are an importantspace weather factor with a strong influence on satelliteelectronics. Around half of magnetic storms are followedby an enhancement of relativistic electron fluxes. It appearsthat the pre-existing magnetospheric state is a critical fac-tor in radiation belt variability; for example, efficient pro-duction mechanisms appear to need a seed population ofmedium-energy electrons. Storm-time magnetospheric con-vection intensification, local particle acceleration due tosubstorm activity, and resonant wave-particle interactionare the main fundamental processes that cause particleenergization and loss.

Current understanding of energetic electron and protonacceleration mechanisms within the radiation belts is insuf-ficient to discriminate between the effectiveness of differentenergization, transport, and loss processes at different L-shells and local times. Quantitative assessment of the pre-dictive properties of current and emerging models is thusessential. Given the current very limited understanding ofgeo-effectiveness, excellent monitoring of the real timeenvironment (nowcasting) is the most useful product thatcan be provided to satellite operators. This, and well-characterized descriptions of historical events, can be usedto interpret failures and improve the resilience of spacecraftdesign. We propose intensified support for comprehensivemodeling of the behavior of particles within the radiationbelts based on multi-point in-situ observations (seeAppendix F.8 for more details).

7.6. Understand solar energetic particles throughout the

Sun–Earth system

SEPs present a major hazard to space-based assets. Theirproduction is associated with large flares and fast CMEs inthe low corona, typically originating from complex activeregions. The prompt response can arrive at Earth in lessthan an hour from the onset of eruption, and some timesin a few minutes after the onset of the flare in the case ofa well-connected, relativistic particle event. This is often fol-lowed by continued high SEP fluxes as a CME shock propa-gates out from the Sun and may exhibit a distinct peak,rising by as much as two orders of magnitude, as theCME shock passes over the point of measurement. The


particles in this peak are often referred to as energetic stormparticles (ESPs) and are particles trapped close to the shockby plasma turbulence associated with the shock (e.g.,Krauss-Varban, 2010). They may provide most of the par-ticle fluence within a gradual SEP event. While warning ofevents in progress is certainly important, many users requiresignificantly longer warning, e.g., 24 h (e.g., all-clear periodsfor Extra-Vehicular Activities for astronauts). Recentmulti-point observations of SEP events off the Sun–Earthline show that prompt energetic particles have access to awide longitudinal extent for some events. They also showthat for a large proportion of SEP events, the prompt ener-getic particles do not propagate along the normal Parkerspiral but in the magnetic field of a pre-existing CME.ESPs can also reach energies in excess of 500 MeV and alsopose major space weather hazard. The ESPs represents a“delayed” radiation hazard traveling with the CME shock,so that there is a good chance of making accurate predic-tions of the onset times of ESP events. Thus the key researchchallenge is how the intensity, duration, and arrival time ofthe prompt response depends on the SEP event near theSun, presence of a type II or IV radio burst in the near-Sun interplanetary medium, and the source location of theflare/CME that drives the shock.

In order to further the understanding of these SEP pop-ulations, we recommend that in addition to the recommen-dations for the solar and heliospheric domains above, thereshould be multi-point in-situ observations of SEPs off theSun–Earth line and possibly closer to the Sun than theL1 distance. For this we recommend extensive use of anyplanetary and inner-heliospheric missions that may occurin the future (see Appendix F.9 for more details).

8. In conclusion

Throughout the development of this roadmap, we wereas much impressed by the observational coverage of theSun–Earth system that is currently available as by thesparseness of that coverage given the size, complexity,and diversity of conditions of the system. Our main recom-mendations are readily condensed into a few simplephrases that reflect the team’s philosophy: make good useof what we have now; bring considerable new resourcesto bear on those problem areas that have the largest lever-age throughout the overall system science; and be efficientwith the limited budgets that are available in the nearfuture. Out of this philosophy come the recommendationsthat stress the paramount importance of system modelingusing the system observatory, of collaboration and coordi-nation on instrumentation and data systems, and of a shiftfrom a few expensive, comprehensive satellite missions toinvestments in many efficient small satellites that targetspecific capability gaps in our current fleet.

The desire expressed in the charge to the team for demon-strable improvements in the provision of space weatherinformation will need to be evaluated after our recommenda-tions have started to be implemented. Our recommendations

are formulated clearly with an expectation to meet thatdesire: knowing how to model the field in solar eruptionsbefore its arrival at Earth, better understanding of the trig-gers and inhibitors of instabilities in the magnetosphericfield, improved tracing of the sources and sinks of energeticparticle populations from Sun to radiation belts, and greatermodeling capabilities of the drivers of ionospheric variabil-ity, to name a few of the targets of this roadmap, appearall within reach if targeted investments are made.

One key challenge overall is the transformation into aneffectively functioning global research community. Manyof our recommendations are meant as catalysts to this pro-cess: open data, shared investments, code development ascommunity effort, and coordinated instrumentationdeployment are all part of this.

In order to stimulate implementation of our recommen-dations as well as to assess their impact we recommend thatthe research community keep a finger on the pulse as activeplayers and as stewards of the investments: this roadmapneeds updates and course changes by the research commu-nity and their funding agencies every few years to remainrelevant.

Appendix A. Roadmap team and process

The roadmap team was appointed by the COSPAR lea-dership taking advice from the COSPAR Panel on SpaceWeather and the steering committee of the InternationalLiving With a Star (ILWS) program. The team organizedits work to ultimately summarize its findings in three keyareas: (1) observational, computational, and theoreticalcapabilities and needs; (2) coordinated collaborativeresearch environment, and (3) collaboration between agen-cies and between research and customer communities. Theteam’s deliberations proceeded hierarchically: (a) keyresearch questions to be addressed to make demonstrableadvances in understanding the space environment and thespace weather services derived from that; (b) researchmethodologies required to effectively address those researchquestions, and (c) specific crucial observables, models, datainfrastructures, and collaborative environments that enablethose research methodologies. Specific recommendationsfor observational, modeling, and data infrastructures aregrouped into three pathways. The pathways reflect theassessed scientific urgency, the feasibility of imple-mentations, and the likelihood of near-term success.Subsequent pathways need the recommendations of preced-ing ones implemented at least to some extent to achieve fullsuccess, but can be initiated in parallel. The roadmap teamdoes not mean to imply priorities in the sense of scientific orsocietal importance in these pathways, but developed themto devise a pragmatic, feasible, affordable, internationalimplementation plan to meet the roadmap’s charge.

The roadmap team met for three face-to-face workingmeetings: twice in Paris at COSPAR headquarters(November 2013 and September 2014) and once inBoulder (April 2014, in conjunction with the US Space

Fig. A.4. Logical structure of the roadmap: After tracing impact pathways, requirements are prioritized in three major aspects of the researchinfrastructure, after which opportunities and challenges are identified in each to reach a set of prioritized recommendations.


Weather Week). In between these work was coordinated intelecons every second week and email exchanges. Inputfrom the community was sought through presentations atvarious meetings, notifications of an email input channelthrough community newsletters, and by personal discus-sions with the team membership. An initial draft of theRoadmap was presented in an interdisciplinary lecture atthe COSPAR general assembly in Moscow in August2014, followed by a dedicated community discussion ses-sion. Prior to the completion of the full written report,the principal findings were summarized in an ExecutiveSummary (packaged as a brochure) and made availablefor final comments on the COSPAR web site,9 and handedout at the US LWS meeting and the European SpaceWeather Week, both in November 2014.

The roadmap team was supported in its activities by anoversight group from agencies and research organizations:Philippe Escoubet (ESA, ILWS Steering Committee Chair,ESA), Madhulika Guhathakurta (NASA, ILWS), JeromeLafeuille (WMO, ICTSW, ILWS), Juha-Pekka Luntama(ESA-SSA Applications Programme, ESA), AnatoliPetrukovich (RFSA, ILWS vice-chair), Ronald Van derLinden (International Space Environment Service), andWu Ji (Chinese National Space Science Centre, COSPARDeputy Chair).

For a partial compilation of space weather resources, werefer the reader to web sites such as at WMO/OSCAR10

(Observing Systems Capability Analysis and ReviewTool), the WMO Space Weather Portal,11 and the onlineSpace Weather Catalogue.12

9 https://cosparhq.cnes.fr/sites/default/files/executivesummary_com-pressed.pdf.10 http://www.wmo-sat.info/oscar/spacecabapilities.11 http://www.wmo-sat.info/product-access-guide/theme/space-weather.12 http://www.spaceweathercatalogue.org.

Appendix B. Roadmap methodology: tracing sample impact

chains

The complexity of the interconnected system of physicalprocesses involved in space weather precludes a single,comprehensive, yet understandable and concise explo-ration of the network as a whole. To effectively and effi-ciently address its charge the roadmap team thereforedeveloped a strategy that focuses on three largely distinctspace weather phenomena with largely complementaryimpact pathways on societal technologies: (A) electricalsystems via geomagnetic disturbances, (B) navigation andcommunication impacts from ionospheric variability, and(C) (aero-) space assets and human health via energeticparticles. These impact pathways collectively probe thephysics of the Sun–Earth system whose variability is quan-tified, for example, in the G (geomagnetic), R (radio black-out), and S (radiation) storm scales used by NOAA’s SpaceWeather Prediction Center. The physical processesinvolved in the probed pathways often map across thesestorm scales that are but one way to quantify some ofthe attributes of space weather. No three pathways orscales can characterize the rich spectrum of space weatherphenomena, just as terrestrial weather forecasts typicallydo not characterize the multitude of phenomena ofweather, but in both cases a carefully selected set willencompass much of the dynamics and will help usersunderstand the threat of space and terrestrial weather inthe coming hours and days.

Subsequent to this impact tracing, the roadmap teamintegrated these parallel assessments into a comprehensiveview with associated requirements and prioritized actionsto be taken to advance our understanding of space weatherand to improve the quality of space weather services. Wedo so by looking from the Sun, through the heliosphere,into geospace which we use to describe the domain

http://https://cosparhq.cnes.fr/sites/default/files/executivesummary_compressed.pdf

http://https://cosparhq.cnes.fr/sites/default/files/executivesummary_compressed.pdf

http://www.wmo-sat.info/oscar/spacecabapilities

http://www.wmo-sat.info/product-access-guide/theme/space-weather

http://www.spaceweathercatalogue.org


encompassing the coupling region between solar wind andthe geomagnetic field, throughout the geotail, and downinto the high layers of Earth’s atmosphere. A summaryof the organizational structure is shown in Fig. A.4.

Appendix C. State of the art in the science of space weather

C.1. Achievements

The research activities in “space weather” are increasingrapidly: for example, based on entries in the NASAAstrophysics Data System, over a recent 15-year period(1997–2012) the number of publications per year hasincreased by a factor of 12 (refereed publications) to 16(all publications) to 925 publications per year in total onthe topic for 2012. This rapid growth in interest in, andimplied support for, space weather research (see Fig. 1)has advanced our insights in and understanding of spaceweather tremendously. That advance largely coincided withthe growth of space-based observational capabilities,whose discoveries included the direct confirmation of theexistence of the solar wind in the early 1960s (with undeni-able confirmation during a six-month flight of the MarinerII spacecraft en route to Venus) and of the phenomenon ofthe coronal mass ejection first recognized as the highlydynamic process originating from the low corona in theearly 1970s (with the coronagraphs flown on OSO-7 andshortly after that on Skylab).

Since those early observations, recent achievements inthe space-weather science arena are supported by anincreased availability, accessibility, and sharing of botharchival and near-real-time space- and ground-based data.These involve, for example, the US “virtual observatories”

for solar and for geospace data, and the InformationSystem of the World Meteorological Organization[WMO] through its Interprogramme Coordination Teamon Space Weather [ICTSW] and the International SpaceEnvironment Service [ISES], and the ICSU World DataSystem and World Data Center [WDS/WDC], and e-in-frastructrue programs in the EU (e.g., ESPAS,SEPServer, HELIO) and Japan (IUGONET), as well asby the development of end-to-end modeling frameworks(including NASA’s CCMC), and various space agency pro-grams (e.g., ESA SSA; NASA iSWA).

On the solar side, we now have available continuousfull-Sun observations at multiple wavelengths, from multi-ple perspectives, from below the surface into the helio-sphere. This observational coverage includes far-sideobservations and stereoscopic imaging by STEREO,SDO, and SoHO combined with other observatories, andhelioseismic probing of the interior and even the far sideof the Sun. Unfortunately, this full-sphere coverage doesnot include the (near-) surface magnetic field which is mea-sured only from Earth’s perspective (continuously by SDO,and by some ground-based observatories). This full-Suncoverage supports the study of conditions leading to vari-able space weather, including the evolution leading up to

and following solar storms, and the coupling to adjacentflux systems. But the space to cover is vast, and the observ-ables range in size scale, energy, wavelength, and otherphysical attributes over many orders of magnitude, requir-ing multiple types of observatories and necessarily leavinggaps in our knowledge.

Years of increasing coverage and open-data policieshave led to a growing open-access event archival databaseof well-observed flares, CMEs, and particle events thatenables event comparisons and inter-domain couplingstudies. Search and visualization tools are developing toincrease the utility of and access to observational data.On the numerical front, leaps forward in the realism ofnumerical modeling that enable us to explore, for example,the generation, rise, and emergence of magnetic flux bun-dles into the solar atmosphere and the subsequent initialphases of a CME, the formation of sunspots upon fieldemergence onto the solar surface, and the analysis of chro-mospheric processes in a realm with radiative transfer, par-tial ionization, and comparable field and plasma forces.These numerical experimental settings are complementedby initial developments of data-driven models of the solarcoronal field and of the Sun-heliosphere coupling in thenascent solar wind.

Within the heliosphere, we now have multi-perspectiveobservations of propagating disturbances from Sun toEarth and beyond, often as combinations of in-situ andremote-sensing observations. These include long-term, con-tinuous, in-situ monitoring of solar wind and SEP proper-ties at the Sun–Earth L1 point, one million miles upwindfrom Earth, and multi-point, multi-scale measurements inthe near-Earth solar wind. These observations often pro-vide information on the structure of CMEs, including theshocks between CME and the preceding solar wind (itselfoften containing earlier CMEs) that are important to thegeneration of solar energetic particles (SEPs). It has beendemonstrated that approaching CMEs can be detectedthrough interplanetary scintillation and that these showsignatures in cosmic-ray muon intensities that may be usedas proximity proxies. Sets of spacecraft around the innerheliosphere have demonstrated the prevalence of wide-an-gle SEP events. Exploratory models exist of CME prop-agation, including CME-coronal-hole and CME-CMEinteractions, from near the Sun to beyond Earth orbit.

Understanding of the magnetospheric dynamics greatlybenefits for the availability of continuous remote sensingof solar activity and in-situ L1 monitoring of solar windand SEP properties. Many key advances in the researchof energy storage and release processes are being madeowing to the availability of multi-point, multi-scale mea-surements throughout the magnetosphere (Cluster,THEMIS) as well as specialist coverage of the magnetotail(ARTEMIS) and the radiation belts (Van Allen Probes),and improved coverage by ground-based observatories(radar and magnetometer networks, e.g., SuperDARNand SuperMAG). Here, as elsewhere in the science ofSun–Earth connections, significant advances in numerical


modeling are made, including the coupling between CMEsand the geomagnetic field as well as magnetotail dynamics,based on developments that utilize different approaches(magneto-hydrodynamic [MHD], hybrid, kinetic) and thatinclude energetic particles. Regional studies have improvedunderstanding of GIC effects at auroral latitudes and iono-spheric impacts in the Southern Atlantic (magnetic) anom-aly (SA(M) A-region). Although still challenging areas ofresearch, advances are being made in the characterizationof substorm dynamics (specifically their field evolutionand system couplings), understanding of the ring currentsystem, and analyzing the importance of ionospheric feed-back for geomagnetic field dynamics.

In the ionosphere–thermosphere–mesosphere (ITM)domain we see growing observational coverage andincreasing regional resolution of ITM dynamics (includingGNSS and ionosonde data), enabling near real-time mapsof total electron content (TEC) and critical-frequency mapsfor radio communication, and advances in 3D reconstruc-tion (by, e.g., GNSS radio-occultation and Beaconreceivers, AMISR, and forthcoming EISCAT_3D+instrumentation and the Swarm mission). There isimproved understanding of the contribution of theplasmasphere in the ionospheric total electron content(TEC), especially on the night side and during extremeevents (e.g., plasmaspheric plumes). Insight has been deep-ened in sub-auroral effects (e.g., SAPS), in vertical couplingbetween troposphere and ionosphere (tidal modes, gravitywaves), and in solar wind momentum storage and releasein the thermosphere. There is improved access to data-product repositories, which in part is due to empiricalmodel development based on long-term databases. All thissupports the development of physics-based models andunderstanding of the processes involved, including data-ingestion into assimilative models, some of which areapproaching transition to real-time operations.

The rapid growth of observational and numericalresources is advancing our understanding of the Sun–Earth system as a coupled whole, but many major issuesremain to be addressed on the path towards improvedscientific understanding and desired specificity in forecastsfor the various sectors of society that are impacted by spaceweather.

C.2. Prospects for future work

With regards to solar drivers of space weather, we haveto advance the diagnostic capability to forecast times, mag-nitudes, and directionalities of flares and eruptions; at pre-sent diagnostics with the best skill scores perform quitepoorly with often an unguided “no flare” forecast perform-ing best, and with trained observers unable to differentiateforecasts within 24-h increments and flare magnitudesuncertain to one or more orders of magnitude, and withonly the first advances to forecast the magnetic contentof eruptions into the heliosphere being made in recentyears. New magnetic field intruding into pre-existing active

regions (a phenomenon referred to as “active-region nests”

that one could view as pre-conditioning) is often involvedin major flaring, making even “all clear” forecasts difficulton time scale more than about a quarter of a day. Long-range couplings and large-scale forces affect how eruptionspropagate into the heliosphere, and we face the problem oflimited coverage of solar magnetic field and coronal struc-ture observations to incorporate within our growing mod-eling capabilities: only little more than a quarter of thesolar-surface is accessible for high-resolution, well-cali-brated magnetography, with vector-field data routinelymade only for the active regions of their relatively youngearly-decay products.

Another major challenge is establishing what determineswhether a solar event inserts SEPs into the heliosphere andhow such SEPs propagate through the heliosphere.Advancing on these fronts requires the development ofphysics-based, data-based models for the research commu-nity, including in particular a general-purpose observation-driven magnetic-field-modeling tool. Quantifying the prop-erties of extreme flares, CMEs, and SEP events requiresinvestments in analyzing natural archives (e.g., biosphere,ice deposits and rocks – including those from the lunar sur-face; e.g., Kovaltsov and Usoskin, 1964), the use of cool-star analogs to our Sun to capture a sufficient number ofrare extreme events to quantify the probability distributionfor such extreme events powered by our present-day Sun.From a climatic point of view, we need to advance ourunderstanding of solar spectral irradiance variations andits modeling based on magnetograms and models of thesolar magnetic atmosphere.

Once solar disturbances enter the heliosphere, we haveyet to understand what establishes the characteristics(width, direction, velocity), deflection, and propagation ofCMEs, including the interaction with the background solarwind. That problem requires the development of data-dri-ven MHD modeling through general-access communitymodels for the background solar wind and for interplane-tary CME evolution from Sun to Earth (such as increas-ingly available and supported at NASA’s CCMC). Wehave yet to establish the properties of the 3D magneticstructure of CMEs starting with CME initiation and fol-lowing their propagation through and interaction withthe magnetic field of the solar wind, required for forecast-ing CME arrival times and CME magnetic properties.Another major issue is the understanding of shock prop-agation through the pre-existing solar wind and the result-ing SEP production and propagation (distinct from theprompt relativistic particles).

Remote-sensing and in-situ observations are needed tovalidate heliospheric models, including continued in-situL1 monitoring of solar wind and SEP properties. We needto acknowledge, moreover, that the evolution of CMEs,and therefore their geo-effectiveness, is affected by the solarwind ahead of them: we need to include pre-conditioning ofthe heliosphere into account, and thus think in terms ofcontinuous evolution rather than in terms of isolated single


events. On larger geometric and temporal scales, we needimproved understanding of how heliospheric magnetic fieldand solar wind modulate galactic cosmic rays, as well asimproved understanding of solar-wind patterns that cansignificantly enhance the energy of magnetospheric elec-trons. For the terrestrial magnetosphere, we need improve-ment of knowledge of conditions in the incoming solarwind and incoming SEP populations, both by continuedand improved upstream monitoring and modeling/forecasting. At the ground-level “base” of the mag-netospheric state we need to ensure the continuity of globalmeasurements of electric field and currents by ground-based radars and magnetometers and to find accuratemethods to measure also particle precipitation and conse-quent ionospheric conductance variations in global scales.The details of the solar-wind/magnetosphere couplingremain elusive, so that understanding of the geo-effective-ness of CMEs has to include how the CME phenomenaevolve through the bow shock. An additional challengecomes from magnetospheric pre-conditioning that can sig-nificantly modulate the impact of solar activity in the geo-sphere. Here, once more, we need physical data-drivenmodeling (MHD and kinetic) through general-access com-munity models reaching across dipole-like and tail-like fielddomains, with ability to include regional impact modelingand the forecasting of substorms, of auroral currents, ofthe ring current, and of particle populations and their pre-cipitation. We need multi-scale multi-component measure-ments in shocks, reconnection, and turbulence, andunderstanding of cross-scale coupling from micro-scalephysics to meso- and macro-scale.

For particle populations in the radiation belts we needbetter understanding of energetic ion and electron injectioninto radiation belts in storms and substorms, and also bet-ter forecasting of solar-energetic particle radiation stormsat LEO and aircraft altitudes. Coupling across domainsis also important, requiring space-based multi-point explo-ration of the MIT interfaces to understand, e.g., the originof outflow ions and their roles in the magnetic storm devel-opment and the characteristics of the ionospheric-thermo-spheric phenomena caused by magnetic storms. Lackingnatural records, the specification of worst-case mag-netospheric processes leading to strong GICs will requirephysics-based models, to be somehow adequately validatedfor conditions not normally encountered in geospace.

A key goal, perhaps the key goal, in mitigating spaceweather impacts mediated by the ionosphere is to develop3D global models of the morphology of the ionosphere.Given such a model it will then be straightforward to derivemany products needed by users, e.g., nowcasts and forecastsof TEC for GNSS and space radars, of usable frequenciesfor HF communications (e.g., foF2, MUF(3000) F2) andof where there are strong spatial gradients that will leadto ionospheric scintillation. Such models must be physics-based in order to encompass all the processes and phenom-ena that can affect the morphology of the ionosphere, espe-cially in highly disturbed conditions. They must therefore

include coupling to all the systems associated with the iono-sphere, especially the co-located neutral atmosphere (ther-mosphere) as well the magnetically-coupled regions of theplasmasphere and magnetosphere. The key processes tobe studied include the coupling of solar-wind-mag-netospheric variability, ion-neutral coupling, the develop-ment and evolution of ionospheric storms, travelingionospheric disturbances (TIDs), small-scale irregularitiesand bubbles, the mechanisms behind equatorial F-regionradio scattering, and the role of small-scale structure inthe mid-latitude ionosphere. An adequate understandingof all this necessary physics requires extending coverage ofthese coupled domains by globe-spanning ground-basednetworks for continuous observation (including mid- andlow-latitude electron and neutral density, electric field,and neutral wind variations during geomagnetic events).Coordination in the development of these networks is alsoneeded, including integration of ground- and space-baseddata systems, to develop data-driven modeling capabilityfor the near-real-time ionospheric morphology. In thisdomain, too, we identify how physical data-driven model-ing feeds into general-access community models for, e.g.,regional perturbation knowledge, including scintillation,and near-term TEC forecasting. ITM studies will also feedinto improved predictions of satellite drag for the purposeof orbit calculations (including collision avoidance,scheduling of critical operations and re-entry predictions)via better modeling of the thermospheric neutral density,temperature and composition.

Finally, we note that there is much synergy with radioastronomy. Many ionospheric processes add noise to radioastronomy observations, so there is scope for sharing ofdata and knowledge on the basis of one scientist’s noiseis another’s signal. In addition, the technologies beingdeveloped for advanced radio telescopes (e.g., LOFARand the Square Kilometer Array SKA) have great potentialfor spin-out into ionospheric work, e.g., as demonstratedby the Kilpisjarvi Atmospheric Imaging Receiver Array(KAIRA) in Finland. Thus the ITM space weather thecommunity needs to strengthen the links with the radioastronomy community in this exciting new era.

Appendix D. Research needs for the solar-heliospheric

domain

In this Appendix D, and in the subsequent Appendix E,we provide a series of needs derived from detailed formula-tions of the top-level goals. In specifying the needs, we dif-ferentiate between data that we require to enable the basicprocess and data that we desire to significantly improve ourability or to accelerate our progress. We also differentiatebetween observables that are already being obtained andwhich we need to maintain, and observables for which weneed to improve instrumentation or diagnostic capability,or observables for which we need to develop new capabili-ties. Priorities in this Appendix are given for the science ofthe solar-heliospheric domain; they are merged and


renumbered with the priorities from the geospace domainfrom Appendix E in the main text of this roadmap.

Goal SH-1: Time-dependent description of the coronal

magnetic field, both for the space-weather source regions

(i.e., quiet-Sun filament environments and active regions)

and for the embedding global field.The cornerstone of any understanding or prediction of

solar activity lies in the magnetic field. We note that.

� the three-dimensional magnetic field of an eruptingregion and of the surrounding corona are essential forspecifying the characteristics of the transient solar activ-ity that causes space weather impacts;� the magnetic structure of the solar ejecta and its interac-

tion with the embedding coronal and heliospheric fieldsresult in the strength and orientation of the changingmagnetic field that will reach the Earth; and that� the coronal magnetic field and its extension into the

heliosphere affect SEP creation and transport.

Goal SH-1a: Specify magnetic structure of space-weather

sources associated with active regions

� REQUIRE: High-resolution vector photospheric fieldand plasma flow boundary condition. MAINTAIN theability (SDO/HMI) to observe the full disk at about1 arcsec resolution. IMPROVE for source regions byan order of magnitude spatially to at least 0.3 arcsec res-olution as combined 2nd priority by new solar observa-tions (as envisioned, for example, for Solar C).� REQUIRE: Vector boundary condition at a force-free

layer, e.g., the top of the solar chromosphere.IMPROVE on the current ground-based experimentalability at NSO/SOLIS and MLSO ChroMag; NEWability to be developed for space-based observations atabout 0.2 arcsec resolution as combined 2nd priorityby new solar observations (such as planned for SolarC), along with extensive theoretical/modelingdevelopments.� REQUIRE: Development and validation of algorithms

for magnetic field models using, e.g., coronalpolarimetry and imaging data, including considerationof connections between active region magnetic fieldsand the surrounding global background coronal field.IMPROVE on current modeling capabilities in MHDand nonlinear force-free codes using currently avail-able (e.g., SDO) observations [MAINTAIN] at about1 arcsec resolution. NEW: combine with informationfrom radio gyroresonance measurements (FASR,CSRH).� REQUIRE: Off-Sun–Earth line observations of the

coronal plasma structures for on-disk source regionsof solar eruptions. Develop a NEW capability to obtainand ingest binocular observations from near-Earth per-spective and at about 10–20� off Sun–Earth line as 1stpriority in solar observations with a new missionconcept.

Goal SH-1b: Specify the evolving global background

coronal magnetic field:

� REQUIRE: Full-disk, front-side photospheric bound-ary condition. MAINTAIN the ability to obtain full-disk magnetograms as now possible with SDO/HMIand some ground-based observatories (GBOs).� REQUIRE: Models for the global solar field reaching

into the heliosphere are crucially limited by magneto-graph data being available only for the Earth-facing sideof the Sun. More comprehensive boundary data for a sig-nificantly larger portion of the solar surface, especiallynear the east limb of the Sun (where data obtained fromEarth-based observations is oldest as regions evolve onthe far side) is necessary. This may occur through incor-poration of information from the backside of the Sun asseen from Earth, via helioseismology/time-shifted dataand/or by direct observations off the Sun–Earth line ofthe photospheric field, specifically at least about 50 helio-centric degrees trailing Earth (from around an L5 per-spective, or - at least for some time - with the SolarOrbiter mission). IMPROVE modeling combined withNSO/GONG and SDO/HMI helioseismological data.NEW: 3rd priority capability of solar magnetographyoff the Sun–Earth line, at first with the Solar Orbiter,while planning for sustained such observation for moreextensive coverage in the more distant future.� REQUIRE: Calibration of the high-latitude magnetic

field (using, e.g., measurements from above and belowthe ecliptic plane) to validate surface flux transport mod-els for the largest-scale field that determines the overallheliospheric structure. IMPROVE modeling combinedwith high-resolution observations from, e.g., Hinodeand GBOs; obtain observational coverage of the high-latitude regions by the out-of-ecliptic observatory SolarOrbiter as 3rd priority in NEW solar observations.� DESIRE: Full-disk, chromospheric magnetic boundary

to validate and constrain magnetic models; the first stepshould be a study to establish the potential value of full-disk chromospheric vector magnetographs as additionto surface field measurements; NEW measurementsunder development by MLSO COSMO and plannedfor Solar C.� DESIRE: Coronal polarimetry (radio, infrared, visible,

UV) to validate and constrain magnetic models. A firststep here is to IMPROVE our abilities to use present-day GBO data to guide and validate continuous MHDfield modeling of the global solar corona.� REQUIRE: Coronal imaging to validate and constrain

magnetic models. MAINTAIN the ability to obtain X/EUV imaging of the solar corona with space-basedobservatories at about 1 arcsec resolution as now possi-ble with SDO, supported by STEREO.

Goal SH-1c: Evolving description of the ambient inner-he-

liospheric magnetic field and plasma flow in the solar wind

through which CMEs propagate and with which they


interact, that is important for the generation and propagation

of energetic particles, and that drives recurrent geomagnetic

activity. The foundation of such models is derived from Goal

SH-1b.

� REQUIRE: Data-driven modeling of the solar windthroughout the inner heliosphere, improved with betterdescriptions of underlying physics, and routinelyincorporating magnetic field to go beyond present-dayhydrodynamic modeling. MAINTAIN the ability tofeed coronagraphic observations into heliospheric mod-els, but substantially IMPROVE the modeling byincluding magnetic information from the corona andadvancing it towards geospace and beyond.� REQUIRE: Development of techniques to adjust mod-

els based on heliospheric observations, including avail-able heliospheric images, and the development ofNEW tools, such as radio-based interplanetary scin-tillation (IPS) methods, to detect and map perturbationsapproaching Earth.� DESIRE: Solar and interplanetary observations from

above the ecliptic plane to validate models or to guidemodels once such capabilities are developed. SuchNEW observations could for some time be providedby the Solar Orbiter, the 3rd priority solar instru-mentation, to develop and evaluate the utility of suchobservations.� DESIRE: Continuation of multi-view, multiple EUV

emission line, full-disk photospheric vector magneto-graphs, multi-view coronagraphs and heliospheric imag-ing, and L1 solar wind measurements. These capabilitiesshould be MAINTAINed to drive modeling.

Goal SH-1d: Specify magnetic structure of space-weather

sources associated with long-lived (quiescent) filaments

� REQUIRE: Observations of the photospheric line-of-sight boundary condition. MAINTAIN the ability toobtain full-disk magnetograms, such as now observedby SDO/HMI and GBOs. NEW: a capability to covermore of the solar surface for magnetic measurementsas 3rd priority.� REQUIRE: Characterization of coronal currents using

non-potential magnetic models, and/or coronal evolu-tion models, that are validated/fit to data such as coro-nal polarimetric (radio, infrared, visible, UV) andimaging observations. IMPROVE our abilities to usedspace and ground-based imaging and GBO polarimetricobservations to guide and validate magnetic modelingcapabilities based on MAINTAINed surface field maps.IMPROVED and NEW coronal polarimetric measure-ments under development for GBOs in the U.S. andChina, and potentially off Sun–Earth line.� DESIRE: Chromospheric magnetic field (boundary con-

dition and prominence) measurements to validate andconstrain magnetic models. IMPROVED/NEW GBOobservations and data assimilation techniques.

� DESIRE: off-Sun–Earth line observations of coronalplasma; this NEW ability, the 1st priority for solarobservations, is to be combined with NEW data assim-ilation techniques.

Goal SH-2: Description of CME/flux-rope evolutionthroughout the heliosphere, enabling prediction of CME arri-

val time, kinematics, and magnetic field strength and direc-

tion as function of time.

Once the plasma and magnetic properties of the ejectafrom the solar source regions are specified, we are in a posi-tion to determine its heliospheric evolution. We note that

� quantitative knowledge of the properties of theerupting field and the field into which it eruptsenable modeling of the transport and evolution ofthat eruption throughout the heliosphere en routeto Earth;� specifying the solar wind conditions incident upon the

magnetosphere is critical for determining the spaceweather response in the geospace system; and that,� in particular, high-impact hazards of space weather

strongly depend on the strength and orientation of mag-netic fields and the density and velocity at solar wind atthe Earth.� REQUIRE: NEW data-assimilative models coupling

coronal models to solar wind propagation models (innear-real-time), explicitly incorporating magnetic fluxrope structure established from Goal SH-1a and 1d intobackground wind model (Goal SH-1c).� REQUIRE: MAINTAIN/IMPROVE on validation of

solar-wind propagation models with in-situ observationsat Earth/L1.� DESIRE: MAINTAIN/IMPROVE validation of solar-

wind propagation models with off-Sun–Earth-line in-situobservations (STEREO) and NEW off-Sun–Earth lineremote-sensing observations (e.g., L5) and above-ecliptic(Orbiter, SPP).� DESIRE: MAINTAIN/IMPROVE on validation of

solar-wind propagation models with remote imaging:e.g., Heliospheric imagers (structure), muon detection(orientation via pitch angle anisotropy), radio (Faradayrotation measurements constrain magnetic fieldstrength), IPS.

Goal SH-3: Develop the capability to predict occurrence oftransient solar activity and the consequences in the heliosphere

Given the coronal magnetic field, we are in a position tocharacterize and predict solar drivers of space weather. Wenote that

� transient solar activity leads to a conversion of magneticenergy into space-weather-driving phenomena such asflares, coronal mass ejections, and energetic particles;� interactions between the space-weather source and the

global background corona affect the properties of thesephenomena; and that


� the ability to predict transient solar activity in advance isgenerally desired, and required for warning of promptsolar energetic particles

Goal SH-3a: Predict transient solar activity in advance

� REQUIRE: MAINTAIN observations that may beused for precursor warnings, e.g., active-region com-plexity, neutral-line shear, filament activation, helioseis-mology obtained from a range space and ground-basedobservatories, and IMPROVE upon theoretical andempirical justifications and implementation strategiesfor their use.� REQUIRE: Develop NEW capabilities for ensemble

modeling in which probability of eruption is obtainedbased on perturbing coronal magnetic structure deter-mined as in Goal SH-1, through e.g., flux emergence/flows, with consideration of sympathetic eruptivelikelihood.� DESIRE: IMPROVE on theoretical understanding and

modeling capability for predicting instability by analyz-ing the topological properties of coronal magnetic struc-tures determined as in Goal SH-1.� DESIRE: MAINTAIN STEREO observations that may

be used for precursor warnings using off-Sun–Earth-lineobservations; NEW off-Sun–Earth line observations.

Goal SH-3b: Specify the consequences of transient solar

activity into the heliosphere

� REQUIRE: MAINTAIN space and ground-basedmulti-wavelength imaging (including coronagraph) dur-ing eruption including Sun–Earth-line (SoHO, SDO),off-Sun–Earth-line (STEREO), and NEW off-Sun–Earth line observations.� REQUIRE: IMPROVE on existing techniques to quan-

tify kinematic properties of ejecta including velocity andtrajectory.� REQUIRE: IMPROVE on eruptive models starting

from coronal magnetic structure determined as inGoal SH-1; NEW incorporate interactions between thesource and the background during eruption, and con-strain/validate by data – leading to quantification ofmagnetic structure of ejecta.

Goal SH-4: Prediction of particle intensities, includingall-clear forecasts, flare-driven acceleration near Sun, and

shock acceleration at CME fronts.

Energetic particles present a major hazard to space-based assets as well as possible consequences for aircraftand aircrews and aircraft passengers, particularly in polarroutes. The production of SEPs is associated with largeflares and fast CMEs in the low corona, typically originat-ing from complex active regions. The prompt-response par-ticles can arrive at Earth in less than an hour (some timesas rapidly as a few minutes) after the onset of an eruption.These prompt SEPs may be followed by a longer lasting

flux of particles originating at the CME shock as it propa-gates through the heliosphere, and a short, sharp, very highflux of energetic storm particles (ESPs) as the CME shockpasses the point of measurement. Many of the needs toimprove understanding of the prompt SEPs and ESPs arecontained in the needs for solar and inner-heliospheric phe-nomena; the purpose of this short section focusing on SEPsand ESPs is to explicitly highlight their significance.

The strong increases in fluxes of high-energy protons,alpha particles and heavier ions in SEP events can causeor contribute to a number of effects, including:

� Increases in astronaut radiation doses resulting inincreased long term cancer risk and short termincapacitation.� Single Event Effects (SEEs) in micro-electronics� Solar cell degradation.� Radiation damage in science instruments and interfer-

ence with operations.� Effects in aviation include increases in radiation dose

levels and interference with avionics through SEEs.

While warning of events in progress is certainly important,many user needs (e.g., all-clear periods for EVAs for astro-nauts) require significantly longer warning, e.g., 24 h.

Prompt energetic particles vs. energetic storm particles.

Prompt particles: At the onset of major (M or X-class)flares and CMEs, SEPs can arrive in minutes after the startof the event. These events are typically associated withlarge, magnetically complex active regions. The energysource of the particles is still under debate: it may beentirely due to shocks generated low in the corona, or itmay be that reconnection in solar flares plays a vital role,or both. The prompt SEPs represent the most difficult fore-casting problem, as they require.

� The prediction of major solar flares and eruptions priorto their occurrence.� The efficiency of the flare/CME to produce energetic

particles.� The escape conditions of the particles from the coronal

acceleration site.� The angular extent for injection of particles in the inter-

planetary medium (STEREO observations show thatprompt energetic particles can be measured on a widelongitudinal extent).� The conditions of propagation of the prompt particles in

the interplanetary medium (a large proportion of SEPevents are found to propagate not along the normalParker spiral but in the magnetic field of a CME). Inaddition to the conditions of diffusion in the medium,this evidently affects the delay between production ofenergetic particles and arrival at the earth.

Energetic storm particles (ESPs): Energetic storm parti-cles are particles detected in-situ when shocks pass byspacecraft. These are particles trapped by turbulence just


ahead of the shock front, and can be prominent events withthe particle flux jumping by 1–2 orders of magnitude.Occasionally, ESPs can reach >500 MeV and hence posemajor space weather hazard. Therefore, the ESPs representa “delayed” radiation hazard because the ESP event fore-cast is essentially a shock arrival forecast. The durationof the ESP event generally depends on the shock geometry:At quasi-perpendicular shocks (where the shock normalvector is perpendicular to the local magnetic field), gener-ally a narrow spike is observed, while a broad profile isobserved in a quasi-parallel shock.

A good indicator of an impending ESP event is a radio-loud shock, i.e. a shock that produces type II radio burstnear the Sun. Stronger shocks generally produce moreintense ESP events. Radio-quiet shocks can also produceESP events. Shock travel time ranges from about 18 h toa few days, so there is a good chance of making predictionof ESP events.

ESP events precipitate in the polar regions and thus theyare hazardous to satellites in polar orbits. They lead toradio fadeouts in the polar region and may cause commu-nication problems for airplanes in polar routes.

Understanding and predicting particle accelerationrelies on all preceding SH goals. We note that.

� Given a heliospheric field description, connectivity ofSun to Earth can be determined enabling all-clear fore-casts if no active region is connected to the Earth (thus,even if flaring occurs, prompt particles on time scales of10 s of minutes are unlikely to occur). However, thereare unsolved problems related to unusual spread ofSEPs in longitude.� Given prediction of a flare/CME, warning may be given

of prompt particles and generally of SEPs on time-scalesof 1–2 days; information about connectivity combinedwith models/observations may give warning of theirlikely geoeffectiveness.� CME driven shocks continuously accelerate particles

from the corona to 1 AU and beyond, so once a CMEoccurs, models that propagate them through the back-ground heliosphere may predict shock-related SEPs,and in combination with observations of shocks canallow prediction of geoeffectiveness.

Goal SH-4a: All clear (ongoing)

� REQUIRE: IMPROVE/NEW Evolving heliosphericmodel driven by 3D coronal field model (Goals SH-1band c). Modeling/analysis to understand unusual longi-tude spread and prompt rise of some SEP events.� DESIRE: IMPROVE/NEW Predictive models of erup-

tion, analyzed in the context of global field, to establishlikelihood of connectivity change and sympathetic erup-tions (Goal SH-3a).

Goal 4b: Predict rapidly arriving SEPs (time scale min-

utes to hours)

� REQUIRE: MAINTAIN/IMPROVE nowcast observa-tions to predict geoeffectiveness of incoming SEPs: rela-tivistic and mildly relativistic particles (neutronmonitors and GOES) to establish connectivity and fore-warning of future, less energetic particles; measurementof type III radio bursts to establish that the flare/erup-tion site has access to open magnetic fields.� REQUIRE for 1–2 day forecast: IMPROVE/NEW abil-

ity to predict flares/CMEs (Goal SH-3a).� DESIRE: IMPROVE/NEW Evolving heliosphere

model driven by 3D coronal field model (Goal SH-3),coupled with particle acceleration models and proba-bilistic/stochastic methods to characterize likelihood ofstrong SEPs from an active region.

Goal 4c: Predict SEPs from shock acceleration/transport

(time scale hour to days)

� REQUIRE: MAINTAIN radio observations of TYPEII bursts indicating shocks in front of CMEs; GBOradio observations to provide information on shocksclose to the Sun at high frequency; L1-observations ofshocks driven in interplanetary space that must be mea-sured from space because they are below ionosphericcutoff.� REQUIRE: MAINTAIN in-situ measurements of SEPs

at L1.� DESIRE: IMPROVE/NEW evolving heliosphere model

driven by 3D coronal field model with treatment ofresponse to propagation of CME (Goal SH-1b and c),coupled with particle acceleration and transport models.� DESIRE: MAINTAIN/NEW multi-point in-situ mea-

surements of SEPs off Sun–Earth line (STEREO, SolarProbe Plus, Solar Orbiter).

Appendix E. Research needs for the geospace domain

Following our impact tracings for the three differenttechnologies presented earlier in this roadmap, we cat-egorize the scientific activities as follows:

� Magnetospheric field variability and geomagnetically-induced currents.� Energetic particles, in order of impact importance:

– Solar energetic particles,– Radiation-belt energetic particles, and– Galactic cosmic rays;

� Ionospheric variability.

E.1. Magnetospheric field variability and geomagnetically-

induced currents

Goal GM1: Understand the dynamical response of the

coupled geospace system to solar wind forcing to improve

GIC forecast capabilities


The dynamical geospace environment, in the form ofthe coupled MIT system, responds to energy input fromthe solar wind through processes of energy input, stor-age, release and dissipation (in severe GIC events theenergy input from solar spectral irradiance plays atmost a minor role). Rapid reconfigurations of this cou-pled system can be sufficiently fast, so that inductivemagnetic fields produce large and unwanted GICs whichpresent a threat to power grid infrastructures on theground, which can be very serious during severe spacestorms.

The examples of GIC events with significant social andeconomic impact in recent history (GIC storms in 1989and 2003) motivate research to specify and forecast thegeo-electric field imposed on power grid and transmissionline infrastructure in order to facilitate better preparednessfor such impacts. One of the major challenges is that thefundamental physical processes which control especiallythe fast release of stored energy in this coupled systemremain poorly understood.

Under some conditions energy that is stored is releasedgradually with little GIC impacts; conversely under otherconditions stored energy is released very rapidly fromprior quasi-stable configurations. The dependence of thisrapid energy release on the current and prior state ofthe coupled geospace system is not well-known, and pre-sent-day prediction capabilities are thus severely ham-pered, both by the poor understanding of the impactsof system-level coupling, and through the lack of knowl-edge of the severity of the effects arising from the priorstate of the system through pre-conditioning. For exam-ple, the response of the coupled magnetosphere-iono-sphere system to a given solar wind input depends uponthe sequence of these drivers, and the time integratedresponse within the coupled geospace system. Whetherthere is a pre-existing ionization state in the ionosphere,a well-developed ring current, a heated atmosphere, or adense tail current etc. can all change the rate at whichthe geospace system dissipates stored energy, and exactlywhen violent dissipation might begin. In the absence of aproper understanding of these impacts, the inductive elec-tric fields arising from dB/dt from this coupled systemhave both a severity and indeed an onset timing whichcannot be accurately forecast with present observationsand understanding.

In order to improve the GIC forecasting capabilities,the required advances in the underlying physics will neces-sitate further discovery-based research. This must incor-porate both existing and new measurements of theglobal state of the system and its dynamics, including datafrom extensive networks of ground-based instrumentssupported by constellation class in-situ satellite infrastruc-ture. Such measurements, together with targeted improve-ments in model development (including data assimilation)will be key.

Goal GM-1a: Understand energy storage and release in

the MIT system in driving large dB/dt at ground.

� REQUIRE: Goal oriented research for improved physi-cal understanding of the onset of rapid and large scalemagnetotail morphology changes, including substormonset, for improved physics-based forecasting (5thpriority in Pathway I recommendations for modeling).At the moment we only see that some times the mag-netosphere decides to deliver the excess energy to theionosphere in a steady flow or small parcels (e.g., inthe form of BBFs in the tail), and some times in onebig parcel (substorm), driving GICs. We need to findout which characteristic of the coupling or which char-acteristic of the driver determines this choice. In addi-tion, we need improved understanding of the role ofpreconditioning in determining the magnetosphericresponse to a time sequence of solar wind drivers (cf.,the largest ever historical super-storms have occurredas isolated events which can (often) be during solarminimum periods).� REQUIRE: Systematic observations characterizing the

global state of the MIT system (3rd priority inPathway I recommendations for new instrumentation).Such observations require a maintained fleet of space-craft in crucial locations in the tail and close to the iono-spheric end of the conjugated field lines as described inSection 7.3. In addition, in order to solve the remainingkey physics questions, we will require 2 constellation typemissions at locations at the inner edge of the plasmasheet (from 7–12 RE), and within the so-called auroralacceleration region at several thousand km altitude.None of these satellites needs radically new and expen-sive instrumentation though. The challenge lies in popu-lating the key locations with spacecraft housing standardplasma instrumentation suites, and designing orbitswhich optimize the presence of s/c at these locationsoften enough, if not continuous. Again, the THEMISmission provides some useful perspectives for the plan-ning of the necessary configurations of future fleets orconstellations. It would be valuable to have simultaneousmeasurements from both inside and outside of the onsetregion (at it’s Earth side and further in the tail) in orderto follow energy conversion processes and associatedplasma flows towards the Earth and away from the mag-netotail in the form of plasmoids. In addition a set of 4–6probes at geostationary distances and distributed evenlyin different MLT sectors would help in understandinghow a major partition of this released energy gets storedinto the ring current and radiation belts. An importantasset supporting these missions would be a set of LEOsatellites with magnetic field and particle instruments inorder to get a better handle on the magnetosphere-iono-sphere coupling (energy carried by currents and precipi-tating particles). A particular challenge is to measure theintensity of field-aligned currents that directly control thehorizontal currents in the auroral ionosphere and conse-quently also the dB/dt values as measured on the ground.NEWMAINTAIN: When planning satellite fleets orconstellations for probing the MIT system it is important

13 http://www.wmo-sat.info/oscar/applicationareas/view/25.


to keep in mind the support that other existing satellitemissions can provide. A nice example towards this direc-tion is the AMPERE project which maintains an impres-sive set of magnetometers hosted by the Iridium satelliteson LEO orbits. Another example of fruitful collab-oration with long heritage of support to other spaceplasma missions is the DMSP program hosted by theUnited States Department of Defense. Ensuring continu-ity of this kind of missions will be an important factor inthe attempts to keep the costs of multipoint MIT-moni-toring on a tolerable level.� REQUIRE: With the help of observations described

above we need to develop magnetospheric modelingcapacity taking upstream solar wind input and derivingforecast dB/dt at the Earth‘s surface, including trans-formation and coupling at the bow shock and the mag-netopause, coupling between hot and coldmagnetospheric plasma populations, and coupling tothe ionosphere including ionospheric conductivity mod-ules and feedback (5th priority in Pathway I recommen-dations for modeling). We have still little understandingon the relative importance of such pre-states in theprime coupled elements of our system, nor the time con-stants of their potential impacts, but they are in order ofdistance from the Earth:– Atmosphere: temperature, altitude/density and

winds – long lived (a day or so).– Ionosphere: conductivity and convection (strength

and location) – short lived (from seconds to hours).– Plasmasphere: location and density – long lived

(days)– Ring current: strength, composition, energy of parti-

cles, shape and/location – long lived (days).– Tail current: strength, mass composition, tempera-

ture, shape and/or location – short lived (hours).– Lobe field: pressure from previous energy coupling to

solar wind, i.e. the energy reservoir from previousevents is not completely emptied – no intrinsic timeconstant, influenced by other system input andoutput.

� REQUIRE: Coordinated ground-based measurementcan provide crucial support for satellite missions (6thpriority in Pathway I recommendations for maintainingexisting capabilities). We propose to establish a formalbasis for collaboration and data sharing between helio-spheric and magnetospheric missions and ground-basedassets. Informal (e.g., Cluster Ground-based WorkingGroup) and formal (e.g., THEMIS mission) agreementshave been made before, but usually missions do notinvolve a requirement for GB support in their level-1science requirements for ground-based support (VanAllen Probes, MMS, Solar Probe, etc). This must be rec-tified. Satellite missions require support from a globalnetwork of– MAINTAIN/IMPROVE: ground-based magnetome-

ters, for the global and local current patterns.

Instruments operated in the auroral zone and polarcap (magnetic latitudes above about 65 degrees) yieldoften the most interesting data, but during storm per-iods also sub-auroral stations (magnetic latitudes ofsome 55–65�) are needed for grasping the entire pic-ture. According to the recommendations of theWMO Interprogramme Coordination Team onSpace Weather (WMO/ICTSW) the threshold forspatial resolution in ground-based magnetic fieldrecording is 500km for SWx monitoring while forbreakthrough science it would be �100 km.13

– MAINTAIN/IMPROVE: radars for the electric fieldand convection patterns on meso-scale. On globalscale such measurements can provide informationon the energy state of the magnetosphere from ovalsize and dynamics. Similarly as for magnetometers,the interesting magnetic latitudes are mostly above55 degrees. WMO/ICTSW recommendations for spa-tial resolutions are 300 km and 10 km for monitoringand break-through science, respectively.

– NEW: optical monitoring from space or ground forthe precipitation and conductivity state (7th priorityin Pathway I recommendations for new instru-mentation). While ground-based instruments (all-sky cameras) are useful in the research of meso-scalephysics, again as in the case of radars for system levelscience global images by space-born imagers will beof key importance at least for two reasons: (i)Space-based observations do not suffer fromcloudiness problems and (ii) global scale instanta-neous images of the auroral oval provide valuableinformation on the magnetospheric energy storageand release processes during storm and substormperiods.

� DESIRE, IMPROVE/NEW: Improved magnetosphere-ionosphere coupling models, including the impact offeedback between field-aligned current structures andenergetic particle precipitation on ionospheric conduc-tivity - which is an important aspect of the creation ofionospheric currents and hence ground dB/dt, resultingin an electric induction field.� DESIRE, IMPROVE/NEW: Improved data assim-

ilation method and models applied to the coupled mag-netospheric system including magnetic fields, currentsystems, and hot and cold plasma populations.� DESIRE, IMPROVE: Capacity to specify the global

state of the magnetosphere will ultimately be requiredto advance our knowledge of space weather to the truepoint of predictability. Even with currently operatingmulti-satellite missions in the so-called HeliophysicsGreat Observatory, geospace remains extremely sparselysampled in-situ. Global networks of ground-basedmagnetometers, radars, and optical imaging (both

http://www.wmo-sat.info/oscar/applicationareas/view/25


ground and space) provide the only currently credibleapproach to specifying the global state of the coupledmagnetosphere ionosphere system. Even supplementedby current satellite missions, the state remains sparselyspecified in the critical regions of in-situ geospace.

Future efforts to develop nano- and micro-satelliteinfrastructure could offer a cost-realistic route to a trueconstellation class mission (perhaps 50–100 satellites)needed to advance the system level science beyond its cur-rent typical case study methodology. Hosted payloads mayoffer an additional attractive route to securing constellationclass specification of the state and dynamics of the in-situgeospace environment. Note, however, that this would beonly an additional and not alternative route, as thoseinstruments would most likely have to rely on simple mea-surements of plasma characteristics and relatively coarseand “dirty” field measurements due to host instrumentationinterference. Nevertheless to monitor simple arrival times ofchanges in the plasma environment these would still be veryvaluable for science, but more so for monitoring. Suitablemissions, which could host SWx instrumentation are, e.g.,Galileo satellites, meteorological satellites on polar and geo-stationary orbits, and satellites testing new space technolo-gies (e.g., the PROBA missions). Keeping theseopportunities in mind, new SWx instrumentation shouldbe based on generic solutions with are suitable for payloadshosted by several different missions. NEW

Goal GM-1b: Understand the solar wind – bow shock –magnetosphere-interactions

� REQUIRE, MAINTAIN/IMPROVE/NEW: Ongoing,reliable and continuous upstream monitoring of incom-ing solar wind conditions and propagation to 1AU suchas from L1, or from other locations closer to the sun(4th priority in the Pathway I recommendations formaintaining existing capabilities).A suitable solar windmonitoring capability will be required. Data from L1typically gives the general idea and this location is defi-nitely suited for the primary mission for SWx forecasts,but in the near future we will also need satellites closerto the subsolar bowshock/magnetopause to understandwhat really arrives at Earth. For such remaining sciencepurposes during the years 2007–2009 a good con-stellation was formed by the Cluster and THEMISmulti-satellite missions. Systematically during springand autumn times one of these missions was monitoringthe energy feed from solar wind and the other was prob-ing the consequences in the magnetosphere. Ensuringsuch coordination between different missions will bevital for better physical understanding of energy transferprocesses between solar wind and magnetosphere. Inlonger run the research conducted with Cluster-THEMIS (or in the future MMS-Cluster-ARTEMIS)observations should pave the way to design futurecost-efficient solutions for continuous monitoring of

upwind conditions with optimally configured satelliteconstellations.Both Cluster and THEMIS have beendesigned for science purposes. For SWx monitoring aless comprehensive suite of instrumentation would mostlikely be sufficient. A basic set of instrumentation couldinclude, e.g., magnetic and electric field (at least AC inthe case of E-field) instruments and a plasma instru-ment(s) with similar specification as in Cluster (e.g., elec-tron and ion spectrometers, energy ranges 1–30 keV).The combined THEMIS and Cluster constellation pro-vides a good first basis for examining the baseline forfuture constellations, for observing the time evolutionof solar wind structures as they cross the bow shock,magnetosheath and magnetopause regions (mainly forcase study purposes).� DESIRE, IMPROVE/NEW: Improved global kinetic

and/or hybrid models of the solar wind-bow shock mag-netosphere interaction, including

1. downstream impacts of kinetic bow shock andmagnetosphere processing of upstream solar wind,

2. development and improved coupling of kineticmodules into global magnetospheric models espe-cially plasmasheet-ring current-magnetosphere-ionosphere system.

� DESIRE, NEW: Research in order to achieve a consoli-dated view whether a satellite (or satellite constellation)located closer to the Sun–Earth line than currently at L1is crucial to get better GIC predictions or not. Previousstudies presented in the literature give a controversialview about this question. As a consequence it is at pre-sent not clear:– how close to the Sun–Earth line does a monitor need

to be to get good upstream conditions for spaceweather/GIC input (in the literature the correlationlength is different for the magnetic field, for theplasma velocity, and for electric field/potential(Burke and Weimer D.R., 1999); this knowledgedetermines whether an L1 orbit (perhaps constrainedto within 60 RE only) is good enough or whetherobservations directly upstream from, and close to,the Earth on Earth orbiting spacecraft or con-stellation of spacecraft is needed.

– whether we can specify what we need in terms of SWparameters with a single satellite and some assump-tions about field orientations to constrain it(Weimer and King, 2008); or whether we wouldrather need multi point measurements in theupstream solar wind for improved predictions.

Goal GM2: Uncover patterns in geomagnetic activity

associated with disturbances in, and failure modes of the

power grid infrastructure.

� GICs have potentially the largest space weather impactson terrestrial infrastructure, with potentially long lasting


and very severe impact and the modes of impact andconsequent failure risk need to be understood.� Models of the impacts of imposed dB/dt from the cou-

pled geospace system in the form of geoelectric fieldson power grids need to be developed and improved,and the effects arising from underlying solid earth phy-sics such as sub-surface and sea-water conductivityshould be included.

To date our understanding of the temporal and geographi-cal patterns of geo-electric fields that are the most effectivein driving GIC impact on the hardware in, or the operationof, electric power grids is very limited.

Priority: Develop the tools and data needed to enableengineering and impact studies to understand the failure

modes of power grids in dependence of GMD driver

characteristics.

Goal GM-2a Understand the factors controlling the geo-

electric field in the regions of primarily critical power grids,

i.e serving dense populations at latitudes typically suffering

from large GMDs.

� REQUIRE: MAINTAIN/IMPROVE/NEW: Deploy-ment of a network of ground-based magnetometersand magnetotelluric instrumentation for geo-electricfield measurements with infrastructure for data collec-tion and distribution in near-real time. The critical den-sity of real-time magnetometer and magnetotelluricnetworks needed for the regions of critical power gridsunder the risk of SWx disturbances should be definedon the basis of knowledge about underlying gradientsin ground conductivity in the relevant regions, as thesemay cause very localized features in the geo-electricfield.� REQUIRE, IMPROVE/NEW: Systematic studies on

ground-conductivity in the regions of high GIC risk.Combining results from ground-based, air-borne andsatellite (Swarm in particular) measurement campaignsincluding both solid-earth and sea-water conductivityeffects.

Goal GM-2b: Assess and access space weather GIC risk

and impact data.

� REQUIRE, IMPROVE/NEW: Make power grid GIC(current and electric field) data available to the widerspace weather research community (c.f. General recom-mendations for collaboration between agencies andcommunities, point i). Data of actual GICs effects inpower grids are often collected by grid operators, butsuch data are for a variety of reasons often not (easily)made available for the scientific and technical study ofrisks, except internally with the power grid companies.This kind of data will be essential in the efforts to gainimproved understanding of the different modes of mal-function (non-linear behavior, extra harmonics, andheating, overload, etc.) of transformers and power

grids under GIC forcing. For space weather purposesthis is particularly important as the scientists do notalways know which type of disturbance is actuallyleading to potentially damaging impacts on variouscritical parts of the power grid infrastructure. Thisrequirement includes therefore also determining thecharacteristics of the most geo-effective time sequencesof dynamic fields (one large event, or a fluence of manysmaller repetitive or pulsating events or persistent mod-erate forcing) and their impact on infrastructure span-ning from individual single transformer failure andlifetimes to catastrophic network collapse. Knowledgeof such impact factors could both improve the rele-vance of space weather predictions (not only largedB/dt warning required) and assist in the industriesown assessments of approaches to critical infrastruc-ture protection (CIP) through design of resilient powernetworks avoiding design with single point failureunder the expectation that some elements of the net-work will likely fail.� REQUIRE, IMPROVE/NEW: Nations should consider

undertaking a vulnerability and risk assessment forspace weather impacts on their power grids and otherinfrastructure, recording the resulting risks in nationalRisk Registers and adopting appropriate policy and tak-ing appropriate action to mitigate such risks.

E.2. Magnetospheric field variability and particle

environment

The energetic particle environment in the mag-netosphere is an important factor of space weather. Themain external sources of the magnetospheric energetic par-ticle population are SEPs and GCRs, as well as solar windplasma particles. The last ones, being accelerated by mag-netospheric processes, are responsible for radiation belts,auroral FACs and particle precipitation into the upperatmosphere. The dynamics of the magnetospheric magneticfield under solar wind driving controls particle distribution,acceleration and losses. We note that:

� Solar activity is the main controlling factor for the par-ticle environment in the Earth’s magnetosphere.� Active processes on the Sun and in the heliosphere

reveal themselves in the particle distribution inside themagnetosphere. Omitting the internal magnetosphereprocesses like plasma instabilities and wave activities,which are currently not predictable, MHD or hybridmodels of solar wind – magnetosphere coupling are atpresent the prevailing means to provide space weatherrelated forecasting of the magnetospheric particle fluxesfor operational purposes.� Magnetospheric magnetic field variations determine the

particle dynamics, and disturbances in this field causedby solar wind variations can drastically change thenear-Earth’s particle environment.


� GCRs and SEPs penetrate deep into the Earth mag-netosphere. The resulting particle population dependson the rigidities of the primary particles as well as on themagnetospheric magnetic field structure and dynamics.� Enhancements of trapped electrons accelerated by inner

magnetospheric processes are usually the consequenceof interplanetary shocks, CME or high-speed streamsarrivals.� Low-energy particle precipitations in auroral regions are

usually manifestations of magnetospheric processes tak-ing place during geomagnetic disturbances.

It is essential for satellite operators to know the past, cur-rent and future condition of the space environment (parti-cles and fields) around their own satellite, especially in caseof actual satellite anomaly or before critical operation per-iods. Measurements and modeling of high and low energyparticle fluxes, and observational data assimilation are ofkey importance for operative space weather predictions innear-Earth space.

Goal MEP-1: Description of the magnetospheric state

In order to move the predictability of particle fluxes for-ward, we first need a complete science based description ofthe magnetospheric particle populations, which give rise tospace weather effects (1st and 2nd priorities in Pathway IIrecommendations for maintaining existing capabilities).Solar wind conditions are the controlling factors for thenear Earth energetic particle population, but the previousstate of the system is also critical. The details of the result-ing population always depend on the present driving condi-tion and the prehistory of the event (timescales to beconsidered are of the order of several days to a week).

Trapped energetic particles in the inner magnetosphereform the radiation belts. The inner proton belt is relativelystable, but the outer radiation belt comprises a highly vari-able population of relativistic electrons. Solar wind changesproduce changes to trapped particle transport, accelera-tion, and losses. Current understanding of the ways inwhich the magnetosphere responds to solar inputs, indicatethat the effectiveness of the input is highly dependent on theinitial state. For example, magnetospheric reconnectionrates are believed to be dependent on the oxygen contentof the tail in the reconnection region, so that substorm trig-gering may be modified during storm time, when oxygenconcentrations may be greatly elevated. Moreover, the pro-duction of relativistic electrons, is believed to depend on aseed population of high energy electrons as a result of ear-lier substorm activity. There are numerous other examplesfor such preconditioning.

Storm-time magnetospheric convection intensification,local particle acceleration due to substorm activity andresonant wave-particle interaction are the main fundamen-tal processes that cause particle fluxes energization andloss. However, the details of the mechanisms that may beable to contribute to these processes remain a subject foractive research. While progress has been made in the theo-retical understanding of these competing processes, there is

as yet no clear consensus on which of these will be signifi-cant in particular situations, and no real predictive meth-ods which can give precise fluxes at different local anduniversal times. There is therefore a pressing need to con-front theoretical models with detailed measurements, inorder to resolve these shortfalls.

Finally, there are short-term populations of ring currentparticles produced by substorms and enhanced convection.Currently our understanding of substorm onset is evolvingrapidly, in particular with respect to local dipolarizationstructures. However, we are still a long way from an abilityto predict precise events, and a new physics understandingis almost certainly required. Current spacecraft measure-ment configurations are probably not sufficient to makeaccurate predictions. Since these effects are the major con-tributor to spacecraft failures due to charging, it is essentialthat current investigations are maintained and extended.

Goal MEP-1a Magnetic field and solar wind description:

� REQUIRE: MAINTAIN current continuous solar andsolar-wind observations as a basis for the prediction ofmagnetospheric disturbances (1st, 2nd, and 3rd priori-ties in Pathway I for maintaining existing capabilities).� REQUIRE: Advanced magnetospheric modeling is

needed to take into account local plasma processesaffecting the magnetic field variations producing particleaccelerations, losses, etc. (4th priority in Pathway I formodeling).� REQUIRE: In order to provide the diverse system level

data set essential for evolving predictive models of ener-getic particles, and to provide nowcasting of energeticparticle fluxes, a required space weather product, weshould MAINTAIN similar space and ground-basedinfrastructure as required for the prediction of GMDand GIC (see above)– Simultaneous GEO, LEO, MEO, GTO near real-time

magnetic field and particle data collection and pro-cessing. (1st and 2nd priorities in Pathway II formaintaining existing capabilities)

– Measurements of current solar wind conditions andsolar UV monitoring for solar wind prediction.

– Monitor magnetospheric activity and dynamics bymeans of networks of ground based magnetometers,radars and auroral imagers, including space born glo-bal auroral imagery. (6th priority in Pathway I formaintaining existing capabilities).

� REQUIRE: Improved magnetic field models (empirical,numerical) with particle tracing codes to form a basis fornowcast/forecast particle distributions, (1st priority inPathway II for maintaining existing capabilities).� REQUIRE: Improved models for solar wind prop-

agation from the Sun. (1st, 2nd, and 3rd priorities inPathway I for modeling).� DESIRE: Accurate Magnetic field models to predict Dst

and forecast particle distributions. We suggest this as aparticularly fruitful area for modeling.


� DESIRE: Visualization tools for magnetic fields and lowand high-energy particle distribution in geospace.

Goal MEP-2a: Specify extreme condition of soft particle

(keV) environment in Geospace – Surface charging

Dynamical variations of keV plasma, such as substorminjection, auroral particle precipitation, and field-alignedcurrents are a critical factor in surface charging of satellitesat LEO and GEO. Thus soft plasma descriptions are anessential precursor for the mitigation of surface chargingproblems from extreme conditions in the soft particle(keV) environment; this information is essential for satellitedesign.

� REQUIRE: Monitoring soft particle fluxes by LEOsatellites (e.g., DMSP, POES) (1st Priority in PathwayII recommendations for maintaining existing capabili-ties; 5th priority in Pathway I recommendations formaintaining existing essential capabilities)� DESIRE: permanent auroral oval monitoring from

spacecraft in UV (like Polar) (7th priority in PathwayI recommendations for new instruments)� REQUIRE: Large databases of current and historical

data obtained from in-situ particle measurementsonboard spacecraft should be preserved (and compiledwith cross-calibration) as we need to analyze the occur-rence frequency distribution of extreme conditions ofsoft particles; Understanding the physical mechanismof particle injection and precipitation and relating solarwind conditions with extremes in the geospace environ-ment is an important aspect in order to make progress inpredicting extreme conditions in the soft particleenvironment (1st priority in the Pathway II recommen-dations for modeling).

Goal MEP-2b: Nowcast and forecast of soft particle

(keV) environment in Geospace – surface charging.

The soft particle environment in geospace is dynamicallychanging depending on solar wind conditions.Understanding the current and future soft particle environ-ment (spatial distribution, time variation, energy spectra,etc.) is important to assess the risk of space assets and toexamine on-going satellite anomalies in geospace (1stpriority in the Pathway II recommendation for maintainingexisting capabilities; 1st priority in the Pathway II recom-mendation for new instrumentation).

� DESIRE: Ability to reconstruct three dimensional parti-cle distributions using limited numbers of satellite obser-vations based on the data assimilation method withmagnetospheric models.� DESIRE: IMPROVE our understanding of the relation-

ship between the solar wind parameters and the varia-tions of injecting/precipitating soft particles for correctpredicting the space environment around GEO (2ndpriority in Pathway II recommendations for modeling).

Development of soft particle model in geospace includ-ing supply and loss of soft particle variations are impor-tant for the prediction.

Goal MEP-3a: Magnetospheric energetic particle

measurements.Trapped energetic electrons (�MeV) can penetrate

spacecraft shielding leaving their energy and charge embed-ded in devices. This energy deposit contributes to totalionizing dose and deep dielectric charging; the depositedcharge can build up leading to electrostatic discharge whena breakdown voltage is reached. Dose and damage lead tocatastrophic failure or progressive degradation of the per-formance of solid-state devices, including electronic com-ponents, solar cells, and focal planes. Electrostaticdischarge can physically damage spacecraft materials, cre-ate short circuits, or manifest itself as phantom commandsthrough electromagnetic or radio frequency interference.Trapped energetic protons and heavier nuclei (keV toGeV) cause dose and single event effects (SEE). This isespecially important near the SAA.

At the onset of most geomagnetic storms the outer zoneof the radiation belts may be depleted in several hours.Subsequently, the outer zone will some times build up tolevels higher than before the magnetospheric activity began.A variety of processes can apply in different conditions (seeMEP1), with more extreme events some times causingprompt effects. There is no clear correlation between ringcurrent enhancement, as measured by Dst, and the effectsin terms of relativistic electron penetration deeper into theouter zone, slot, and even inner zone. In the past it wasbelieved that the higher the solar wind speed, the morelikely is the event to lead to elevated post-event electronfluxes; however this has recently been called into question.Monitoring such events at GEO is routinely undertaken.Given the very limited understanding currently availableof the relative geoeffectiveness of various solar wind distur-bances (see above), the best practice consensus is that excel-lent monitoring of the real time environment (nowcasting)is the most useful product that can be provided to satelliteoperators. This, and well characterized descriptions of his-torical events, can be used to interpret failures and improvethe resilience of spacecraft design.

� We REQUIRE high quality measurements of the ener-getic particle environment in Geospace, particularlyLEO and GEO. We need to MAINTAIN the existingGOES satellite measurement, and ENSURE continuedaccess to LANL GEO data, even if not in real time.We NEED better energy resolution at GEO.� REQUIRE: maintain (and enhance) the availability of

high-energy, widely distributed, multiple local time,GEO energetic particle data (5th priority in Pathway Irecommendations for maintaining existing capabilities;1st priority in Pathway II for maintaining existingcapabilities).


� REQUIRE: We need to preserve access to data sets, cur-rent and historical, ground and space based (e.g., preser-vation of LANL, Ground magnetometers, HF radars),in standard archival formats, in order to enable detailedvalidation of physics models (1st priority in Pathway IIfor archival research).� REQUIRE We need a standardization of instrument

documentation and documentation on data processing;a minimum level of documentation required for usercommunity should be identified. (General recommenda-tion for Teaming, point g).� REQUIRE: Need to establish methodologies and met-

rics for validation and calibration models and of data.(General recommendation for Teaming, point g).� DESIRE: hosted radiation monitor payloads at least at

LEO and L1, Perhaps in the long run widespread minia-turised radiation monitors can populate all other keyregions of geospace from field aligned accelerationregions to the ring current and magnetotail.� REQUIRE: To improve our understanding of processes

leading to better understanding of the radiation beltdynamics, we need continuing support of the VanAllen Probes (RBSP) or a similar mission (2nd priorityin Pathway II for maintaining existing capabilities).

Goal MEP-4: Model of magnetospheric particle

acceleration.

It is essential for satellite operators to know if relativisticelectrons can be expected to show a major increase. Thusprocesses of acceleration, transport, and loss of energeticelectrons should be the basis for predicting the dynamicsof trapped energetic particles. However, as remarked pre-viously, understanding geoeffectiveness is a major shortfallin current theoretical understanding. Currently under-standing of Energetic Electron and Proton accelerationmechanisms is insufficient to discriminate between theeffectiveness of different energization, transport and lossprocesses at different L-shells and local times.Quantitative assessment of the predictive properties of cur-rent and emerging models is thus essential. Hence detailedtesting against fine-grained observations is needed.

� REQUIRE: We need to preserve access to data sets,current and historical, ground and space based (e.g.,preservation of LANL and ground-based magnetome-ters), in standard archival formats, in order to enabledetailed validation of physics models. In the futurecubesats should be used in detailed campaigns for thispurpose also (1st priority in Pathway II for archivalresearch).� REQUIRE: Need to establish methodologies and met-

rics for validation and calibration models and of data(General recommendation for Teaming, point g). Wesuggest the establishment of an energetic particle indexfor MEO high-energy flux. Multipoint in-situ measure-ments from MEO satellites can provide the neededinformation about trapped particle population.

The end goal should be a fully predictive dynamical modelof radiation belt particle populations, at different times, L-shells and local times, related to geomagnetic activityindices (especially Dst) and solar wind conditions (velocity,field orientation and pressure). Such a model will certainlyalso need to encompass an accurate description both of thecurrent conditioning of the system and the current and pre-dicted inputs. It should be capable of predicting extremeevents with an accuracy that increases as the event realisticapproaches, until realistic warnings can be given of majordisruptions. Moreover, such a model (or models) will haveprovided a realistic basis for improved engineering designsand procedures that mitigate potential impacts.

Goal MEP-5: Description of SEP and GCR penetration

into the magnetosphere.

SEP and GCR originate from solar system or theGalaxy, respectively, and propagate to the Earth’s orbit.Energetic particle can cause dose and SEE either throughdirect ionization events, or through the nuclear showersthey create. The terrestrial magnetic field topology in gen-eral effects the penetration of these particles into the nearEarth environment. While GCR particles of specific ener-gies can enter the Earth’s magnetic field, solar particles ofrelatively low energies cannot penetrate deep into theEarth’s magnetosphere. Magnetospheric structure changesduring geomagnetic disturbances can provide expansion ofthe region populated by energetic particles coming fromheliosphere. Knowledge of the geomagnetic cutoff is thusimportant for satellite and aviation operators, as it controlsparticle fluxes at low altitudes. The latitude of the penetra-tion region for particles of higher energies depends on theirrigidity and the geomagnetic activity level. On occasion SEPpenetrate deep into the atmosphere where they produce thesecondary particles measured by neutron monitors.Independent on-ground measurements give important infor-mation on SEP spectra, while SEP variations are alsoobserved by satellites in GEO and LEO. The slowly-varyingGCR component is a persistent threat, while SEP are themain source of variable radiation effects on commercial air-craft, particularly at high latitudes, and represent a criticalhazard to astronauts, as well as having effects on spacecraft.

Energetic particles originating from SEP and GCR cancause dose and damage in similar ways to trapped elec-trons, but also cause SEE. We need observations and mod-els to tackle these factors (3rd priority in Pathway III formaintaining existing capabilities). SEP forecasting basedon solar observations is of key importance. This opportu-nity is described in detail in the SH sections.

� MAINTAIN energetic particle monitoring to specifysolar particle access at different latitudes (1st and 2ndpriorities in Pathway II for maintaining existingcapabilities).� MAINTAIN measurements onboard polar LEO:

POES, MetOP, Meteor M, in proton channels 10MeVand above. (5th priority in Pathway I for maintaining


existing capabilities).� MAINTAIN supplemented monitoring at L1 (ACE)

and GEO (GOES, LANL, Electro-L) to establish freespace flux.� MAINTAIN solar observations in UV (See SH section).� REQUIRE: IMPROVE models of the effective vertical

cut-off rigidities dependent on local (or universal) time,and geodetic coordinates (altitude, latitude and longi-tude), as well as on the conditions of geomagnetic dis-turbances described by the Kp/Dst-indices, andpossibly also auroral electrojet indices, AE/AU/ALwhich can potentially modify access.

E.3. Ionospheric variability

The users of ionospheric space weather services have dif-ferential needs depending on the application areas. GNSSusers in precise and safety of life applications need now-casts and short-term forecasts of disturbed periods, so thatthey are aware of potential for disruption of GNSS signals,leading to higher uncertainties in measurements, and insevere cases, to loss of service. In surveying, forecasts onelectron density variations up to 1–2 days and nowcastsand 1-day forecasts on scintillation would be desirablefor planning purposes. For aviation aircraft and ground-systems supporting them we need both short-term alertsand forecasts of disturbed conditions with 0.5–1 day leadtimes (e.g., at least six hours before take-off) in order tobe prepared for potential disruptions in satcom and HFsignals. Operators of space radars require nowcasts andshort-term forecasts of TEC in order to correct for rangeand bearing errors in radar tracking of space objects inLEO.

We note that statistical ionosphere models tuned withdata from ground-based networks or LEO satellites canin many cases provide relatively good results for nowcastsor short-term forecasts in regional or global scales.However, with increasing lead times more comprehensivephysics-based modeling with data assimilation is required.

Goal IO-1: Understand/quantify the benefits of data

assimilation in ionospheric modeling.

In the work to get longer lead times in predictions withadequate reliability space weather research can benefitfrom synergies with meteorology (c.f. General recommen-dations for Collaboration between agencies and communi-ties, point m). Data assimilation is today the standardapproach in numerical weather prediction. In spaceweather assimilation is utilized in some research projects,but comprehensive understanding on how assimilationcould in optimal way support ionospheric SWx forecastsis still missing.

Goal IO-1a: Upgrading MLTI models with assimilation

capability

Variability in the ionospheric electron density is con-trolled by several processes that are coupled with solaractivity, and with the Earth’s magnetosphere and neutral

atmosphere (thermosphere). In thermosphere global cir-culation patterns, thermal conditions and chemistry allhave their impact to the evolution of ionospheric condi-tions. Therefore, priority in the upgrading work shouldbe given to such models that have the capability to takeinto account the various coupling processes.

We recommend the following:

� REQUIRE, IMPROVE/NEW: Advance projects inwhich the research community can investigate theopportunity to use research models forMagnetosphere-Lower Thermosphere–Ionosphere-(MLTI) system in extensive use with assimilationcapability (4th priority in Pathway I recommendationsfor modeling). The CCMC service maintained byNASA is an example of a platform that could beupgraded to become a forum for centralized assimilationcode development as a joint community effort.Especially in the development phases, solutions shouldbe favored and further developed that have flexibleinterfaces for adopting several types of observations,e.g., solar (spectral) irradiance, ionospheric plasma con-vection, 2D and 3D views of electron density, and neu-tral wind properties (chemical composition, density,temperature, wind).

Goal IO-1b: Defining optimal observation capabilities for

ionospheric SWx services.

For the mission to utilize both observations and modelsin ionospheric SWx forecasts it will be valuable to knowwhich instruments provide the best support for improvedmodel results. The optimal use of data requires improv-ing/developing actual/new measurement and modelingtechniques taking into account current and future customerrequirements. In particular, the development of data-dri-ven, physics-based models and associated assimilationtechniques must be supported by funding agencies.Furthermore, it has to be investigated which temporaland which spatial resolution is needed to provide the opti-mum cost-benefit ratio in operational use. Dedicatedresearch projects are needed to answer these questions.


� REQUIRE, IMPROVE/NEW: During the coming fiveyears national funding agencies, space agencies, EUand other stake holders should support research projectswhich investigate the observational needs for advancedionospheric space weather modeling and forecasts inrelation to existing and planned ground and space basedmonitoring capabilities. Funded projects should includesensitivity analyses that show how much different mea-surements with variable time and space resolutions con-tribute to improving our ability to forecast. Potentiallynew metrics need to be developed in order to quantifyin an objective way the benefits from various input datasets (General recommendations for Teaming, point g).The studies should demonstrate how a complementary


mix of measurements can be used effectively in opera-tional forecast systems. Suggestions of optimal cost-benefit solutions should be one outcome from theseresearch projects.� REQUIRE, MAINTAIN/IMPROVE: To support the

work described above the research groups and agenciesmaintaining SW instrumentation should establish testbeds for evaluating new observation methods, tech-niques and data products and for performance val-idation of ionospheric forecast codes at high, middleand low latitudes (General recommendation forTeaming, point f). Such test beds can be identified fromthe existing ground and space based ionospheric moni-toring capabilities, many of which can provide long datarecords from versatile instrumentation. Consideringground based measurement facilities only, a test bed areashould be covered by a dense network of GNSS receiversaccompanied by a wide variety of other relevant instru-mentation (ionosondes, ISR radar, magnetometers,riometers, HF-radars, Beacon receivers). Preferenceshould be given to such set-ups that can provide also neu-tral atmospheric measurements (wind, density and tem-perature, in particular at low and mid-latitudes) andare supported frequently with relevant space-basedobservations. Suitable, already existing candidates fortest-bed areas are often located in the surroundings ofISR systems. For example, the measurements of theResolute Bay and Poker Flat ISRs in northern US andCanada, of the EISCAT ISRs in Fennoscandia andSvalbard, of the Millstone Hill ISR in USMassachusetts and of the Jicamarca radar in Peru aresupported by networks of relevant instrumentation.� DESIRE, MAINTAIN/IMPROVE: Science organiza-

tions involved in SW (e.g., COSPAR, ILWS, WMO,ISES) should do overarching coordination and reviewwork in order to compose unified view from the studiesdescribed above (i.e., re-do the Roadmap task that ispresented here, but with the input from the varioussensitivity studies; see General recommendations for col-laboration between agencies and communities).Recommendations from this work should i) help inidentifying from existing assets those which deserveprioritization in maintenance ii) and support the plan-ning work for future investments in SW instrumentationboth in national and international level.

Goal IO-2: Improved understanding on physics of iono-

spheric processes disturbing trans-ionospheric radio wave

propagation.

In ionospheric research there are still several topics thatare relevant for SWx prediction but are lacking for generalconsensus on the underlying physics. Examples of suchtopics are processes associated with radio scintillation (inparticular plasma turbulence at high latitudes and bubbleformation at equatorial latitudes) and the couplingbetween neutral atmospheric dynamics (wind and waves)and ionospheric disturbances. Systematics in the

appearance of ionospheric storms in different latitudes,local times and during the different storm phases has notbeen studied comprehensively yet, which complicates theefforts even for short-time global forecasts.

Goal IO-2a: Conducting ionospheric research with

enhanced observations.The spatial resolution of measurement stations provid-

ing ionospheric data has great variability. Some regionsare populated with dense GNSS receiver networks whileno ground-based receivers are available in the ocean areas.To fill this gap that is typically for ground based measure-ments, space based monitoring techniques such as radiooccultation, and satellite altimetry data can be used.While ionospheric plasma can be probed with a wide vari-ety of different cost effective and robust instrumentation,monitoring of thermospheric properties (chemical com-position, neutral wind, density and temperature) is chal-lenging. Achieving reliable continuous estimates of theglobal magnetospheric energy input as Joule heating oras energetic particle precipitation is also difficult.

We recommend the following.

� REQUIRE: Advance combined use of ground-based,space based and airborne instrumentation in ionosphericmodeling and forecasts (6th recommendation in PathwayI for maintaining existing capabilities; 2nd recommenda-tion in Pathway III for new instrumentation). Thealready existing ground-based networks (magnetome-ters, ionosondes, riometers, HF-radars, Beacon andGNSS receivers) should be maintained and their data dis-semination systems should be homogenized and stream-lined. Solar SSI measurements, Beacon transmitters,Radio occultation receivers and altimeter radars on-board LEO satellites are examples of data sources whichcan complement the specific view of ground-based mea-surements to jointly lead to a comprehensive image ofthe physical processes. Efforts for open data policyshould be supported (MAINTAIN/IMPROVE,MAINTAIN refers to the existing instrumentation,IMPROVE refers to measurement in ocean areas andto upgrades in data dissemination and to open data pol-icy, c.f. General recommendations for collaborationbetween agencies and communities, point i). In particu-lar, efforts to get access for science community to the dataarchives of Real Time Kinematic (RTK) positioning net-works should be supported. Commercial enterprises inseveral countries maintain RTK systems (dense GNSSreceiver networks) that can provide information on iono-spheric electron content with much better time and spaceresolution than typically available in scientific missions.� REQUIRE, IMPROVE/NEW: Establish and maintain

collaboration with research groups and agencies whichconduct thermospheric and magnetospheric research inorder to search new ways to get neutral atmosphericor Joule heating data as input for ionospheric modeltesting and forecasts. Advancements in Fabry–Perotinterferometry should be harvested, especially where


advanced instrument technologies will open up daytimeobservations. Opportunities provided by some forth-coming satellite missions should also be investigated(e.g., the NASA missions ICON and GOLD for thermo-spheric measurements, the US-Taiwanese atmosphere–ionosphere mission COSMIC II and the ESA missionSwarm for MLTI research).� DESIRE: Advance the search of new data sources to

support SWx forecasts. Examples of new potentialsources are GNSS reflectometry, SAR imaging, andmulti-satellite radio occultation sounding by COSMICII satellites.

Goal IO-2b: Enhancing system level modeling of theionosphere.

Models available today provide a mosaic on processesfrom different parts of the system (with separate pieces forhigh latitudes and low latitudes, for magnetosphere–iono-sphere coupling and ionosphere–thermosphere coupling,for global processes and micro-scale processes, etc.) butknowledge on the linkage between the different pieces is miss-ing. Today ionospheric models are typically coupled withneutral atmosphere models (e.g., MSIS) that describe theclimatology quite well, but they fail to describe the processesin lower atmosphere which affect ionospheric conditions(tides, gravity waves, planetary waves). For this reason, themodels are not able to reproduce properly e.g., the back-ground and seeding conditions for bubble formation. It isimportant to have a model where ionosphere is properly cou-pled with such neutral atmosphere and magnetosphere mod-els that are capable to describe properly at least thoseprocesses that control ionospheric electron content.


� REQUIRE, IMPROVE/NEW: Advance efforts forthree-dimensional imaging and modeling of ionosphericphenomena with the goal to gain better understandingon non-equilibrium plasma processes and on causeand consequence relationships between processes in dif-ferent scale sizes (particularly between regional andmicroscales in order to gain better understanding onprocesses causing scintillation).� DESIRE, NEW: Advance efforts for the development of

a framework where the whole atmosphere (up to�400 km altitudes) is treated as one system. That shouldgive a better means to understand vertical coupling, andalso stimulate stronger links between solar-terrestrialresearch community and the meteorological community.

Appendix F. Concepts for highest-priority instrumentation

F.1. Binocular vision for the corona to quantify incoming

CMEs

Rationale: Knowledge of the magnetic structure of thesolar wind, and in particular that of coronal mass ejections,

that will interact with the magnetospheric field that in turndrives the underlying ITM is needed at least a day prior toreaching geospace, i.e., well before reaching the solar-windsentinel(s) positioned a million miles (or about an hour ofwind travel time) upstream of Earth at the Sun–Earth L1point. This knowledge, in particular of the magnetic direc-tion and strength of the leading edge of coronal mass ejec-tions [CME]s) can be obtained by forward modeling anobserved solar eruption through the embedding coronaand inner heliosphere, provided the magnetic structure ofthe erupting structure is known. Deriving the magnetic con-figuration of an erupting solar active region based on surface(vector-) field measurements alone yields ambiguous resultsat best; these are insufficient for the purpose of MHD mod-eling of CMEs. New field modeling methods have beendeveloped that can utilize the coronal loop geometry to con-strain the model field, particularly when 3D information onthe corona is available (e.g., Malanushenko et al., 2014).Binocular imaging of the active-region corona at moderatespatio-temporal resolution enables the 3D mapping of thesolar active-region field structure prior to, and subsequentto, CMEs, thereby providing information on the eruptedflux-rope structure. Combined with full-disk coronal imag-ing and - if feasible within the mission parameters – withcoronagraphic imaging provides valuable information onthe direction taken by the nascent CME en route to the innerheliosphere through the high corona.

Objectives: Obtain EUV images of solar active regions,of the solar global corona, and (if not provided by otherinstrumentation) of the inner heliosphere from a perspec-tive off the Sun–Earth line, to complement EUV imagesfrom existing instruments on the Sun–Earth line such asSDO/AIA or, if not available, by a second identically-equipped spacecraft.

Key requirements: EUV images of active regions at (atleast) two wavelengths (characteristic of 1–2 MK and 2–3 MK plasma) at �1.5-arcsec resolution and 1-min.cadence, and full-disk observations at �3 arcsec resolutionand 30-s. cadence with low noise to detect signals out to atleast 1.5 solar radii, observed from a perspective between 5and 15 degrees from the complementary second imager. Ifnot provided by other resource: coronagraphic images at 2-min. cadence from about 1.5 to 5 solar radii.

Implementation concept: A single two-channel full-diskEUV imager a 1 arcsec resolution with on-board data pro-cessing to (selected) high-resolution active region imagesand binned full-coronal images (to reduce overall telemetryrates). A compact coronagraph if needed. In a slightly ellip-tical solar-centric (“horseshoe”) orbit drifting away fromEarth at a rate of no more than about two degrees per yearfor the first three years after reaching 5-degrees of sep-aration from the Sun–Earth line. If combined with SDO/AIA a single spacecraft suffices; if a standalone mission,two similar spacecraft are needed to drift apart by therequired separation for a period of at least one year.Each spacecraft could be scoped within, e.g., the NASASMall EXplorer (SMEX) envelope.


Status: Proposed in this Roadmap. Whereas theSTEREO mission has given us a temporary glimpse ofstereoscopic coronal EUV imaging, the field modelingcapabilities at the time did not yet exist and the image res-olution was insufficient on the STEREO and SoHOspacecraft.

Supporting observations: New observational technolo-gies can support the goals of determining the 3D structureof the magnetic field in source regions of solar activity, par-ticularly when combined with the above direct observationsand modeling techniques. This includes radio observationssuch as, for example, with proposed instrumentation forFASR, etc., particularly when multi-frequency ultra-wide-band radio array(s) are available over frequency range of50 MHz to 20 GHz, with full Sun images at all wavelengthstaken once per second (partly completed in China, USdevelopment efforts continuing).

F.2. 3D mapping of solar field involved in eruptions

Rationale: A key science goal in this roadmap is to deter-mine the origins of the Sun’s activity to help predict varia-tions in the space environment. Knowing the low-lyingtwisted filament-like field configurations and their embed-ding field in the deep interiors of unstable active regionsbefore and after eruptions is key to determining what propa-gates towards Earth to drive space weather, which, in turn,is needed to forecast the dynamics and coupling of theEarth’s magnetosphere-ionosphere-atmosphere system dri-ven by the incoming CMEs. Magnetic stresses involved incoronal mass ejections cannot be observed directly, butrequire modeling based on observations of the vector mag-netic field at and above the solar surface, guided by, andcompared to, observations of the solar atmosphere in whichstructures from chromospheric to coronal temperatures arecarriers of the electrical currents that reflect the system’s freeenergy converted to power eruptions. In the context ofPathway 1, the primary goal is to observe electrical currentsthreading the solar surface, and observing the details of thelow-lying configurations known as filaments and theirembedding flux ropes before, during, and after eruptionsto quantify the 3D field ejected into the heliosphere. In thecontext of Pathway 3, the same instrumentation providesobservations of the small-scale processes involved in thetriggering of flares and eruptions, needed for short-termSEP forecasts, and for CMEs forecasts of more than 2–4 days and SEP all-clear forecasts in the coming hours.

Objectives: Obtain vector-magnetic measurements ofactive regions at the solar surface and within the chromo-sphere to measure electrical currents. Image the solar atmo-sphere from 10,000 K up to at least 3MK at matchingresolutions to observe all field structures that may carryelectrical currents. Provide observations before, during,and after eruptions to derive CME field structure to driveheliospheric models, to study how the nascent CME isrestructured as it propagates through the active-regionmagnetic field.

Key requirements: High-resolution imaging (at matchingresolutions of 0.2 arcsec or better) is needed throughoutactive-region atmospheres, spanning the entire activeregion footprint, with observations at temperatures charac-teristic of photosphere, chromosphere, and corona.Spectro-polarimetric observations for photospheric andchromospheric magnetic field measurements. Imagingcadence of approximately 10 s, or better. (Near-)Continuous solar viewing.

Implementation concept: Geo-synchronous or low-Earthorbiter in high-inclination orbit with UV-optical telescopewith polarimetric imaging capabilities, enabling photo-spheric and chromospheric imaging and polarimetry. SoftX-ray and EUV imagers. Substantial ground-based net-work to enable large effective telemetry rates, and/oronboard image selection from large memory.

Status: Considered as a multi-agency mission betweenJAXA, ESA, and NASA to share cost and expertise, insupport of international space-weather research.

F.3. Strong GICs driven by rapid reconfigurations of the

magnetotail

Rationale: The processes of energy input from the solarwind and storage in the magnetotail are now reasonablywell-understood, with further recent discoveries from theTHEMIS and Cluster missions beginning to reveal thetantalizing physics of the development and penetration ofEarthward propagating bursty bulk flows and localizeddipolarizing flux tubes which transport flux and plasmaEarthwards. However, the nearer Earth dynamics whichcouple these flows to the inner edge of the plasmasheetand how exactly the flow braking region couples to theionosphere and produces large field-aligned currents andhence GICs are poorly understood. A two-constellationsatellite mission architecture is proposed. The first willreveal the key plasma physical processes associated withplasma instabilities and flow braking at the inner edge ofthe plasmasheet, in the transition region from dipolar totail-like magnetic fields; the second will reveal the natureof magnetosphere-ionosphere (M-I) coupling in the auroralacceleration region on field lines conjugate to the inneredge of the plasma sheet. From the GIC perspective, theprocesses that control the rate of energy transport andthe actual partition between competing routes of dis-sipation in the coupled M-I system remain insufficientlyunderstood. In terms of space weather impacts, the condi-tions leading to and the physical processes responsible forenabling large field aligned currents to reach the iono-sphere and drive large GICs are not known.

Objective: Determine the M-I processes controlling thedestabilisation of the near-Earth magnetotail, which willlead to the establishment of large field-aligned currentsresulting in extreme GICs.

Key requirements: To meet this objective we need theflight of two coordinated satellite constellations in the inneredge of the plasmasheet, which marks the transition region


between tail-like and dipole fields, and on the conjugateauroral field lines below. It should be noted that the chal-lenge of these constellation missions lies in the populationof magnetospheric key regions and not in particularlyfancy or expensive instrumentation.

Transition Region Explorers: Three-dimensionalplasma and electrodynamic fields (E (at least AC) andB) in the transition region between dipole-like and tail-like fields, which is the originator for large field-alignedcurrents. Coverage is required from close to geosyn-chronous orbit to around 10–12 Re. This could be accom-plished by at least one classical well-instrumentedspinning spacecraft with magnetic field, electric field,plasma measurements for pitch angle resolved electronsand ions in the energy range from around 10’s eV (aslow as possible without ASPOC) to several hundredkeV, including species resolution. This likely requires astandard suite of particle instruments including an electro-static analyser, solid state detector, and ion compositionspectrometer. Potentially this satellite should carry suffi-cient fuel to change apogee altitude between 8–12 RE dur-ing the course of the mission.

This should be supplemented by a swarm of around 4smaller spacecraft approximately 1 RE from the mother,providing coverage in the azimuthal and radial directions.The smaller daughters could carry a more limited basicplasma payload of a magnetometer, miniaturised electro-static analyser, and Langmuir probes for total (includingcold) plasma density and temperature.

Field-Aligned Current Explorers: Multi-point plasmaand electrodynamic fields in the auroral acceleration regionin order to determine the dynamical coupling between themagnetosphere and the ionosphere including the partition-ing and exchange of energy between currents, waves andparticles which are believed to act as a gate for the abilityof the tail to drive FAC through to closure and eventualenergy dissipation in the ionosphere. Operational altitudesshould encompass the range of around 4000 km to 1 Re,utilising conjugate measurements from different altitudeson the same field line, and multiple along track satellitesproviding a capability to resolve spatio-temporal ambigu-ity related to the filamentary nature of FAC and examineand distinguish between the dynamics of Alfvenic andinverted-V auroral acceleration processes. Optionally anadditional single 3-axis stabilized satellite for in-situ aur-oral imaging.

Baseline of two or three spinning spacecraft providingelectric (via wire booms) and magnetic fields and wavesmonitoring, as well as plasma electrons and ions from ener-gies of around 10 eV to 30 keV, likely from an electrostaticanalyzer. Options to add higher energy coverage from ansolid state detector, to resolve the populations up to some50 keV should also be studied. Again, Langmuir probeswould provide significant information about very lowenergy populations below the energy range of the particleinstrument. Payload for the 3-axis stabilized satellite forin-situ auroral imaging is TBD.

Auroral Imaging and Supporting Ground Networksand LEO Satellite Constellations: The constellation mis-sions should be complemented by conjugate auroral imag-ing from the ground, as well as supporting networks ofground-based magnetometers, HF radars, riometers, etc.,to aid the identification of the onset location and the res-olution of the spatio-temporal ambiguity of the processesleading to large dB/dt. Existing or newly provided con-stellations of low-Earth orbiting satellites which canadditionally monitor the precipitating electrons as a mea-sure of ionospheric conductivity changes will provide valu-able complementary measurements; global measurementsof the background large scale FAC distributions, such asavailable from AMPERE, provide the capability to iden-tify onset locations with respect to the nightside convectionpattern.

Incoming Tail Flows and Upstream Solar WindMonitor: Measurements of incoming flows in the centralplasmasheet of the more distant tail are also required toassess incoming flows in the same meridian. These couldbe potentially be provided by pre-existing assets, such asGeotail, Cluster, THEMIS, or perhaps MMS in anextended mission phase. An upstream solar wind monitoris of course required as always.

Implementation concept: The challenge for missionimplementation is not in the instrumentation, which isreadily available and should thus not be a cost driver.Rather, the challenge lies in the positioning of a sufficientnumber of spacecraft at the two key locations in space,and thus in both the possible orbit configuration and thenumber of spacecraft. Likely this requires detailed futurestudy of at least the potential orbits which can launchand deliver the satellites into the appropriate operativeorbits at modest cost. We recommend that the formulationof an international study with representatives of nationalspace agencies is considered, perhaps in the context ofILWS.

Status: Two satellite-constellation concept proposed inthis roadmap.

Supporting observations: by existing mid-tail andupstream solar wind satellites, as well as existing comple-mentary multi-instrument ground-based networks andexisting LEO satellite constellations.

F.4. Coordinated networks for geomagnetic and ionospheric

variability

Rationale: Our understanding of space weather impactson the upper atmosphere is crucially dependent on measure-ments from rich networks of ground-based instruments,including (a) magnetometers to observe how electric cur-rents in the ionosphere are modified by space weather, plus(b) a wide variety of radar and radio techniques to monitorchanges in the density, motion and temperatures of iono-spheric plasmas, as well as (c) optical techniques to measurethermospheric winds and temperatures. These data sourcesare all key inputs into the development of improved models


of the atmosphere and its response to space weather. Thiswill become even more important in future as we focus onassimilative approaches to modeling. As in meteorology,these approaches will advance our science (e.g., throughuse of reanalysis techniques to reveal new systematic fea-tures in the data) and will also provide an efficient practicalbasis for future applications of our science. These assimila-tive approaches are significantly enhanced by the availabil-ity of diverse datasets from spatially dense networks as thesedata then provide strong constraints on the assimilation. Weneed to promote these ground-based networks as a globalsystem for scientific progress on space weather, so that eachindividual (and often independently funded) instrumentprovider (and responsible funding agency) sees how theircontribution fits into the wider picture, i.e. that a local con-tribution builds and sustains local access to a global system.

Objectives: To make a step change in the internationalco-ordination, and delivery, of ground-based spaceweather observing systems to optimize our ability toobserve space weather processes. This must includeincreased engagement with modelers, operators of space-based sensors, and other consumers of space weather data,in order to ensure the optimum interaction between the col-lected data and state-of-the-art international models, andthe synthesis of these data and model results into opera-tional tools whose outputs can be made available to theapplications and technology community. This would be amajor advance on the current situation where ground-based instruments are mostly established and supportedby individual national programs, with only a few projectsformally constituted as multi-national programs(EISCAT being the notable example). There are a goodnumber of projects that operate internationally throughworking level agreements in the scientific community(e.g., SuperDARN) but experience shows that these are dif-ficult to sustain in modern conditions; they rather reflect anolder (make-shift) way of working that dates as far back asthe IGY in 1957/58. Given modern approaches to fundingand governance, these observing systems need a more for-mal international structure that can give them the continu-ous support and stimulus that they need to deliver their fullscientific potential for a future operative space weathersystem.

Key requirements: Maintenance and extension of theSuperDARN network to measure electric fields in the high-and mid-latitude regions at least in the north hemisphere,preferably both (these measurements are a crucial factorin modeling the ionospheric, magnetospheric and radiationbelt response to space weather); high-resolution volumetricmeasurements of ionospheric properties by incoherent scat-ter radars (ISR) at several locations (to resolve the detailedion chemistry and plasma physics at work in the iono-sphere); rich networks of magnetometers, GNSS receivers,ionosondes and riometers to provide regional and globalmaps of key ionospheric properties including ionosphericcurrent systems (auroral, equator and Sq), total electroncontent and ionospheric scintillation, ionospheric critical

frequencies and layer heights and D region absorption;improved operation of Fabry–Perot interferometers (FPI)to enable daytime as well as nighttime measurements ofthermospheric temperature winds. It is highly recom-mended to ensure a concentration of networked instru-ments, including FPI, around major facilities such asincoherent scatter radars, as these provide vital contextfor that technique. These concentrations of instrumentswill allow us to use their locations as a scientific test bedwhere we can explore the detailed response of the iono-sphere to space weather.

Implementation concept: Establish a global program atinter-agency level for coordination of space weatherobserving systems, perhaps similar to coordination ofspace exploration activities. The involvement of agenciesis crucial as it is vital to involve funding bodies to developa sustainable system that is subject to periodic review andwhere there is a proper emphasis on, and awareness of,the global nature of the program. The program wouldestablish working groups with strong expert membershipto carry out its technical tasks, including detailed reviewof measurement requirements, review of advances in instru-ment technology, coordination with space-based measure-ments, recommendations on standards, exploitation ofsecondary data sources especially radio astronomy (muchof their “noise” is actually ionospheric signals that we wishto exploit), etc. The crucial aspect of this program is tobuild a framework where individual agencies can see thata modest contribution enables global science and, in par-ticular, is the logical way to enable first-class participationby the scientific community that they support.

Status: Proposed in this roadmap.

F.5. Mapping the global solar field

Rationale: The global solar magnetic field extends outinto the heliosphere. It defines the structure of the helio-sphere, including the position of the heliospheric currentsheet and the regions of fast and slow solar wind, and playsa key role in space weather at Earth: (1) The interaction ofCMEs with the ambient field impacts their geoeffectiveness.(2) The connection of the heliospheric magnetic field toCME-related shocks and impulsive solar flares determineswhere solar energetic particles propagate. (3) The partition-ing of the solar wind into fast and slow streams is respon-sible for recurrent geomagnetic activity. The Sun’s surfacemagnetic field is a vital ingredient to any predictive modelof the global magnetic field, as it is used to derive boundaryconditions. Global magnetic field models (both the simplerpotential-field source-surface (PFSS) models, and the moresophisticated MHD models) have shown significant successin describing coronal and heliospheric structure. Thesemodels typically use magnetic maps of the photosphericmagnetic field built up over a solar rotation, available froma ground-based and space-based solar observatories. Twowell-known problems arise from the use of these“synoptic” maps. First, the maps contain data that is as


much as 27 days old. The Sun’s magnetic flux is alwaysevolving, and these changes in the flux affect coronal andheliospheric structure. Second, the line-of-sight (LOS) fieldat the Sun’s poles is poorly observed, and the polar fields inthese maps are filled with a variety of interpolation/extrap-olation techniques. Unfortunately, these observationalgaps can strongly influence the solution for the global mag-netic field. In particular, poorly or unobserved activeregions at the limbs (as viewed from Earth) as well as inac-curate polar field estimates can introduce unacceptableerrors in the field on the Earth-facing side of the Sun.

Objective: Model the evolving global solar magneticfield. This requires the near simultaneous observation ofthe Sun’s magnetic field over a larger portion of the sun’ssurface than is available from the Earth view alone.Obtaining photospheric magnetograms off of the Sun–Earth line off of the east limb (portion of the Sun withthe oldest observations as viewed from Earth), to comple-ment magnetograms obtained along the Sun–Earth lineby SDO and ground-based observatories, is the most cru-cial component. Obtaining magnetograms of the Sun’spolar fields over a few years is required to understand theevolution of the Sun’s polar magnetic flux.

Key requirements: Ideally, several spacecraft wouldobserve the Sun’s magnetic field continuously, includingthe polar fields, but such a plan is unlikely to be economi-cally feasible in the foreseeable future. The processes bywhich the magnetic flux on the Sun evolves have been stud-ied for many years, and resulted in the construction of fluxtransport models capable of predicting the evolution of thefield. The incorporation of magnetograms away from theSun–Earth line would be used to augment existing,Earth-view magnetograms to capture a significantly largerportion of the Sun’s evolving flux. LOS magnetogramswith MDI resolution and approximate cadence are likelyto be adequate, although vector magnetograms withHMI resolution and cadence are desirable. Flux transportmodels also predict the evolution of the Sun’s polar fields,but are largely uncalibrated there. Observing polar fluxevolution with MDI resolution over a few years would sig-nificantly constrain these models. EUV or X-ray imaging(STEREO cadence/resolution) to capture coronal holesand evolving structures for model validation from all ofthese views are desirable.

Implementation concept: Primary instrument: Full-diskmagnetograph with MDI-like spatial resolution and hourlytime resolution in an ecliptic orbiting spacecraft reaching atleast 45 degrees off the Sun–Earth line. Orbits goingbeyond this point are desirable. Given such a spacecraft,a heliospheric imager would augment the goals of (1) and(4) by imaging earth-directed CMEs. If feasible, X-ray orEUV imaging (1 channel) at STEREO spatial resolutionand cadence are desirable. A separate, high-latitude (30�or more above the ecliptic) spacecraft mission of a fewyears with the same instrumentation is desirable.

Status: Images from Solar Orbiter may be sufficient toprovide testing of concept of far side imaging, but are

not adequate to fulfill the goal of more continuous moni-toring of a larger portion of the Sun’s magnetic flux.Solar Orbiter may partially fulfill high latitude missionrequirements at the latter stage of the mission.

Supporting observations: Direct imaging data of thesolar atmosphere (such as possible with the STEREOspacecraft) or indirect information derived from far-sidehelioseismology (such as with SDO/HMI and GONG) pro-vide useful constraints, but are no substitute for directmagnetography because these methods do not provide ade-quate information on the magnetic field.

F.6. Determination of the foundation of the heliospheric field

Rationale: The global solar magnetic field plays a crucialrole in space weather at Earth. It influences the internalmagnetic structure of interplanetary coronal mass ejec-tions; the connectivity of the magnetic field determineswhere solar energetic particles propagate, and structureof the field determines whether fast solar wind streams willcross the Earth’s location. An accurate representation ofthe time-evolving global solar coronal magnetic field is arequired input to models of prediction, eruption and prop-agation of CMEs through the solar wind.

Objectives: Determine and obtain a critical set of multi-wavelength coronal magnetometric observables for con-straining the global magnetic field. Develop methods forincorporating these observations into global MHD modelsof the solar corona.

Key requirements: Testbeds of synthetic polarimetricmeasurements at multiple wavelengths to provide diagnos-tics related to the Zeeman and Hanle (saturated andunsaturated) effects and coronal seismology. Techniquesfor efficiently modifying global MHD models of the solarcorona to match data, synthetic or observed. Ultimately,full-sun synoptic observations sufficient to enable a data-assimilative, real-time updated representation of the globalcoronal magnetic field.

Implementation concept: Different wavelengths diagnosedifferent aspects of the solar coronal magnetic field – strongvs. weak field, disk vs. limb, eruptive vs. non-eruptive – andare weighted differently helping to remove line-of-sightambiguity. By utilizing testbeds of synthetic data at allwavelengths (from radio to extreme ultraviolet), effectivemeasurement and optimization strategies can be developedwhich will set priorities and for future observationaldevelopment.

Status: DKIST will provide opportunities for polarime-try and testing of observational techniques with high res-olution and sensitivity but in a small (5 arcminute) fieldof view. Proposed new observations include large (1.5 m)ground-based coronagraph(s) with narrow-band filterpolarimeter and spectropolarimeter to observe the fullSun corona at the limb in the visible and infrared (currentlyundergoing engineering design and preliminary designreview as US-China collaboration). Space-based missionswould provide better duty cycle and continuity of


measurements than ground-based, and also would allowmeasurement at short wavelengths otherwise blocked bythe Earth’s atmosphere. Mission concepts have been pro-posed (ESA) with instruments including spectropolarimet-ric coronagraphs in the EUV and IR for off-limbobservations, and spectropolarimeters to observe the solardisk at heights from the corona down into thechromosphere.

F.7. Auroral imaging to map magnetospheric activity and to

study coupling

Rationale: The response of magnetosphere to solar winddriving depends on the previous state of magnetosphere.Similar sequences in energy, momentum and mass transferfrom the solar wind to magnetosphere can lead in somecases to events of sudden explosive energy release whilein other cases the dissipation takes place as a slow semi-steady process. Comprehensive understanding on the fac-tors that control the appearance of the different dissipationmodes is still lacking, but obviously global monitoring ofthe magnetospheric state and system level approach inthe data analysis would be essential to solve this puzzle.Continuous space-based imaging of the auroral oval wouldcontribute to this kind of research in several ways. The sizeof polar cap gives valuable information about the amountof energy stored in the magnetic field of magnetotail lobes.Comparison of the brightness of oval at different UV wave-lengths yields an estimate about the energy flux and aver-age energy of the particles, which precipitate from themagnetosphere to the ionosphere. These estimates are notas accurate as those from particle instruments onboardLEO satellites, but the additional value comes from thecapability to observe all sectors of the oval simultaneouslyand continuously. Such view is useful especially in the caseswhere the magnetosphere is prone to several subsequentactivations in the solar wind. The shape and size of the ovaland intensity variations in its different sectors enablesimultaneous monitoring of, e.g., nightside magnetosphericrecovery from previous activity, while new energy alreadyenters the system from a new event of daysidereconnection.

Objectives: To achieve continuously global UV-imagesto follow the morphology and dynamics of the auroraloval, at least in the Northern hemisphere, but occasionallyalso in the southern hemisphere. Imager data combinedwith ground-based SuperDARN and SuperMAG networksallows solving the ionospheric Ohm’s law globally, whichyields a picture of electric field, auroral currents and con-ductances with good accuracy and sufficient spatial res-olution. This would mean a leap forward in our attemptsto understand M-I coupling, particularly the ways howionospheric conditions control the linkage to the mag-netosphere by, e.g., by field-aligned currents.

Key requirements: An imager which can observe theelectron auroral emissions in the Lyman–Birge–HopfieldNitrogen waveband, discriminating between the LBH-long

and LBH-short bands. This set-up provides informationabout precipitating electrons in the range 1–20 keV. Forsolving the ionospheric electrodynamics (by estimation ofionospheric conductances) also an imager forBremsstrahlung radiation (more energetic electrons, 20–150 keV) would be necessary. For proton precipitationthe mission would need an imager capable to captureDoppler-shifted Lyman-emission from charge-exchangingprecipitating protons. Time resolution of the images shouldbe better than 60 s and spatial resolution should reach�50 km (at perigee).

Implementation: The objective of continuous monitoringcan be achieved with a constellation of two identically-in-strumented spacecraft in identical highly-elliptical polarorbits (apogees close to 7 RE above the northern poleand perigees near 2 RE.). The orbits of the two spacecraftshould be phased so that one spacecraft is at perigee whilethe other is at apogee and the imagers onboard should beable to observe the auroras from both positions.

Status: Undergoing more detailed definition withininternational science teams.

Supporting observations: Global ground-based net-works, existing LEO and GEO satellites, existing mid-tailconstellation missions, and – as always – upstream solarwind monitor.

F.8. Observation-based radiation environment modeling

Rationale: The radiation belts are key domains in theEarth’s magnetosphere, which cause spacecraft anomalies.It is essential for satellite operators to know if relativisticelectrons can be expected to show a major increase, whichis related to the high risk of spacecraft anomalies. Thusprocesses of acceleration, transport, and loss of energeticelectrons should be the basis for predicting the dynamicsof trapped energetic particles. However understandingradiation-belt dynamics is a major shortfall in current theo-retical understanding. Currently understanding ofEnergetic Electron and Proton acceleration mechanismsis insufficient to discriminate between the effectiveness ofdifferent energisation, transport and loss processes at differ-ent L-shells and local times.

Trapped energetic particles in the inner magnetosphereform radiation belts. The inner proton belt is stable, butthe outer radiation belt comprises a highly variable pop-ulation of relativistic electrons. Solar wind changes pro-duce changes to trapped particle transport, acceleration,and loss. However the pre-existing magnetospheric stateis also a critical factor. For example efficient productionmechanisms appear to need a seed population of energeticelectrons. Relativistic electrons enhancements are animportant space weather factor with a strong influenceon satellite electronics. Around 50% of magnetic stormsare followed by a corresponding enhancement of relativis-tic electron fluxes.

Storm-time magnetospheric convection intensification,local particle acceleration due to substorm activity,


resonant wave-particle interaction are the main fundamen-tal processes that cause particle fluxes energisation andloss. However, the details of the mechanisms that may beable to contribute to these processes remain a subject foractive research. While progress has been made in the theo-retical understanding of these competing processes, there isas yet no clear consensus on which of these will be signifi-cant in particular situations, and no real predictive meth-ods which can give precise fluxes at different local anduniversal times. There is therefore a pressing need to con-front theoretical models with detailed measurements, inorder to resolve these shortfalls. Quantitative assessmentof the predictive properties of current and emerging modelsis thus essential. Hence detailed testing against fine-grainedobservations is needed.

At the onset of most geomagnetic storms the outer zoneof the radiation belts may be depleted in several hours.Subsequently, the outer zone will some times build up tolevels higher than before the magnetospheric activity began.A variety of processes can apply in different conditions (seeMEP1), with more extreme events some times causingprompt effects. There is no clear correlation between ringcurrent enhancement, as measured by Dst, and the effectsin terms of relativistic electron penetration deeper into theouter zone, slot, and even inner zone. In the past it wasbelieved that the higher the solar wind speed, the morelikely is the event to lead to elevated post-event electronfluxes; however this has recently been called into question.Monitoring such events at GEO is routinely undertaken.Given the very limited understanding of radiation-beltdynamics currently available (see above), the current con-sensus is that excellent monitoring of the real time environ-ment (nowcasting) is the most useful product that can beprovided to satellite operators. This, and well characterizeddescriptions of historical events, can be used to interpretfailures and improve the resilience of spacecraft design.

Objective: In-situ multipoint measurements of relativis-tic and sub-relativistic electrons in the inner mag-netosphere, at least at GEO to control radiation beltparticle populations, at different times, L-shells and localtimes. Continuous control of the geomagnetic activityindices (especially Dst) and solar wind conditions (velocity,field orientation and pressure) to predict the possible rela-tivistic electron fluxes variations. Realistic models capableof predicting the radiation belts dynamics can be con-structed after detailed testing against fine-grained observa-tions. Such a model (or models) will have provided arealistic basis for improved engineering designs and proce-dures, which mitigate potential impacts.

Key requirements: Multi-point in-situ observations andreal-time analysis of energetic particle fluxes (mostly, 0.1–10 MeV for electrons) in the inner magnetosphere, of thecurrent geomagnetic conditions (mostly, Dst and ALindices) and solar wind/ IMF conditions at L1 point.

Implementation concepts: To realize our objectives,maintaining the current observation facilities related toradiation belt dynamics (particle and electromagnetic field

measurements in the inner magnetosphere is essential). Tofill the gap of observational data, hosting radiation monitorpayloads (and/or cubesat missions) at LEO, MEO, GEO,and constructing new ground-based observatories for spar-sely covered region is also encouraged. Models (empirical,theoretical, numerical) which are based on these observa-tional data will improve our understanding of transport,acceleration, and loss processes in the radiation belt.Establishing methodologies and metrics for validation andcalibration of models and data is another important issue.

Status: Continuing observations of energetic particlesfrom LANL, GOES, ELECTRO-L, POES, Meteor-M.Geomagnetic indices from WDCs. Solar wind parametersfrom L1 (ACE).

Supporting observations: The current constellation ofCluster, Van Allen Probes, THEMIS, with the upcomingMMS, along with data from the geostationary satellites,and ground-based observation networks by magnetometersand HF radars represent a perfect opportunity to achieve amajor step forward. Theoretical studies and modeling isunderway. What is needed is investment in a few minorexpansions and in particular international coordinationto push the program forward.

F.9. Solar energetic particles in the inner heliosphere

Rationale: SEPs present a major hazard to space-basedassets. The strong increases in fluxes of high-energy pro-tons, alpha particles and heavier ions can cause or con-tribute to a number of effects, including tissue damagefor astronauts, increases in radiation doses, single eventeffects (SEEs) in micro-electronics, solar cell degradation,radiation damage in science instruments and interferencewith operations. In addition, large SEP events can alsoaffect aviation, through increases in radiation dose levelsand interference with avionics through SEEs.

The production of SEPs is associated with large flaresand fast CMEs in the low corona, typically originatingfrom complex active regions. The prompt response canarrive at Earth in less than an hour from the onset of erup-tion, and some times in a few minutes after that onset in thecase of a well-connected, relativistic particle event. This isfollowed by a longer lasting, often rising, flux of particlesoriginating at the CME shock as it propagates throughthe heliosphere, and a short sharp very high flux of ener-getic storm particles (ESPs) as the CME shock passes thepoint of measurement. While warning of events in progressis certainly important, many users require significantlylonger warning, e.g., 24 h (e.g., for (e.g., all-clear periodsfor EVAs for astronauts).

At the onset of major (M or X) class flares and fastCMEs, relativistic SEPs (GeV protons) can arrive minutesafter the start of the event in the case of a well-connectedevent. Less energetic protons will arrive minutes to hourslater depending on the energy but also on the propagationin the interplanetary medium. Recent multi-point observa-tions of SEP events off the Sun–Earth line (STEREO


observations) furthermore show that prompt energetic par-ticles have access to a wide longitudinal extent for someevents. Recent studies also show that for a large proportionof SEP events, the prompt energetic particles do not propa-gate along the normal Parker spiral but in the magneticfield of a pre-existing CME. All this represents an addi-tional difficulty for the forecasting of the arrival of SEPsat Earth, since in addition to the conditions of diffusionin the medium, this affects the delay between productionof energetic particles and arrival at the earth. On the otherhand, shocks produced by fast CMEs start to be identifiedin coronagraph images.

Energetic storm particles (ESPs) can also reach>500 MeV and also pose major space weather hazard.They represent a “delayed” radiation hazard and the ESPevent forecast is essentially a shock arrival forecast. A goodindicator of an ESP event is thus a radio-loud shock, i.e. ashock that produces type II radio burst near the Sun andthe interplanetary medium. Shock travel time ranges fromabout 18 h to a few days, so there is a good chance of mak-ing prediction of ESP events. However, the main questionis how the intensity, duration, and arrival time depends onthe SEP event near the Sun, presence of a type II radioburst in the near-Sun interplanetary medium, and thesource location of the CME that drives the shock. ESPevents are generally of low energy, so they are expectedto precipitate in the polar region. Thus, they are hazardousto satellites in polar orbits. They lead to radio fadeouts inthe polar region and may cause communication problemsfor airplanes in polar routes.

Objectives: In addition to the objectives for solar obser-vations of active regions and of the solar global corona andof the inner heliosphere from a perspective off the Sun–Earth line (7.1) obtain multi-point in-situ observations ofSEPs off the Sun–Earth line and possibly closer than theL1 distance.

Key requirements: Multi-point in-situ observations ofSEPs off the Sun–Earth line and possibly closer than theL1 distance.

Implementation concept: Suite of sensors measuring elec-trons, protons, and ions from helium to iron in the keV toover 100 MeV per nucleon range.

Status: Continuing observations of energetic particlesfrom STEREO, ACE, and other platforms. UpcomingSolar Orbiter and Solar Probe Plus missions will providekey measurements of SEPs close to the acceleration region.Different concepts of missions at L5 proposed (NASAsolar and space physics road map, ESA/CAS small missionopportunity, . . .) Development of solar sails would open upnew research opportunities for energetic particle scienceand monitoring.

Supporting observations: Continuing operations ofground-level neutron monitors and near real-time accessto data of these observatories. They provide informationon the arrival of the most energetic protons from flares.Radio observations (ground-based and satellite) of electron

beams and shocks propagating in the interplanetary med-ium. Also: continued analysis of radio-nuclide data in bio-sphere, ice cores, and in terrestrial and lunar rocks, as theseprovide information on pre-historical extreme events thatcannot otherwise be obtained. Use of multiple data sourcesand radio-nuclides helps to provide some information onparticle energy spectra that are needed to better constrainfluences and to specify environmental conditions.

Appendix G. Acronyms

ACE
Advanced Composition Observatory AIA SDO’s Atmospheric Imaging Assembly AMPERE Active Magnetosphere and Planetary
Electrodynamics Response Experiment
AOGS Asia Oceania Geosciences Society ARTEMIS Acceleration, Reconnection, Turbulence
and Electrodynamics of the Moon’sInteraction with the Sun

ASPOC
Active Spacecraft Potential ControlExperiment (on board Cluster)
AU
astronomical unit (Sun–Earth distance) BBF bursty bulk flow CAS Chinese Academy of Sciences CCMC Community Coordinated Modeling Center CEDAR coupling, energetics, and dynamics of
atmospheric regions program
CIP critical infrastructure protection CME coronal mass ejection COPUOS UN Committee on the Peaceful Uses of
Outer Space
COSMIC Constellation Observing System for
Meteorology, Ionosphere, and Climate
COSMO Coronal Solar Magnetism Observatory (in
MLSO)
COSPAR ICSU’s Committee on Space Research CSRH Chinese Spectral Radio Heliograph DHS Department of Homeland Security DKIST Daniel K. Inouye Solar Telescope DMSP defense meteorological satellite program DSCOVR deep-space climate observatory Dst disturbance storm time index EGNOS European Geostationary Navigation
Overlay Service
EGU European Geophysical Union EISCAT European Incoherent Scatter Scientific
Association
ERG Exploration and energization of radiation
in geospace
ESA European Space Agency ESP energetic storm particle(s) ESPAS European strategy and policy analysis
system
EU European Union EUV extreme ultra violet


eV
electron-Volt FAC field-aligned current FASR Frequency-Agile Solar Radio telescope FPI Fabry–Perot interferometer(s) FR Faraday rotation GB ground-based GBO ground-based observatory GCR galactic cosmic ray(s) GEM geospace environment modeling program GEO geostatioinary orbit GeV gigaelectron-Volt GIC geomagnetically induced current GLE ground-level enhancement GMD geomagnetic disturbance GNSS global navigation satellite system GOES Geostationary Operations Environmental
Satellite
GOLD Global-scale Observations of the Limb and
Disk
GONG The Global Oscillation Network Group in
NSO
GPS Global Positioning System GTO geostationary transfer orbit H2020 Horizon 2020, funding program of EU for
research and innovation
HELIO heliophysics integrated observatory HF high frequency IAU International Astronomical Union ICON The Ionospheric Connection mission ICSU International Council for Science ICTSW Interprogramme Coordination Team on
Space Weather of WMO
IDL Interactive Data Language IGY International Geophysical Year 1957–58 ILWS International Living With a Star program IPS interplanetary scintillation IR infra-red ISES International Space Environment Service ISR incoherent scatter radar iSWA Integrated Space Weather Analysis System
(NASA)
ITM ionosphere–thermosphere–mesosphere IUGG International Union of Geodesy and
Geophysics
IUGONET Inter-university upper atmosphere global
observation network
JAXA Japan Aerospace eXploration Agency JCSDA Joint Center for Satellite Data Assimilation KAIRA Kilpisjarvi Atmospheric Imaging Receiver
Array
keV kilo electron volt L1 Lagrangian point 1 L5 Lagrangian point 5 LANL Los Alamos National Laboratories
LASCO
Large Angle and SpectrometricCoronagraph
LBH
Lyman–Birge–Hopfield Nitrogenwaveband
LEO
low-Earth orbit LF low frequency LOFAR Low-Frequency Array for Radio
astronomy
LORAN Long Range Navigation LOS line-of-sight LWS NASA/SMD Living With a Star program MeV megaelectron-Volt MEO medium Earth orbit MDI Michelson Doppler Imager (on SoHO) MHD magneto-hydro-dynamic MI magnetosphere–ionosphere MHD magnetohydrodynamic MIT magnetosphere–ionosphere–thermosphere MLSO Mauna Loa Solar Observatory MLTI magnetosphere-lower thermosphere–
ionosphere
MMS Magnetospheric Multi-Scale mission MSAS MTSAT satellite based augmentation
system
MSIS Mass Spectrometer and Incoherent Scatter
Radar (empirical atmosphere model)
MTSAT Multi-functional Transport Satellite MUF maximum usable frequency NASA National Air and Space Administration NERC US National Energy Regulatory
Commission
NRT near real time NOAA US National Oceanographic and
Atmospheric Administration
NSF National Science Foundation NSO US National Solar Observatory OSCAR Observing Systems Capability Analysis and
Review Tool of WMO
OSTP Office of Science and Technology Policy PCW Polar Communications and Weather
satellite system
POES polar operational environmental satellite PFSS potential-field source-surface model PSW ICSU/COSPAR panel on space weather R2O research-to-operations RB radiation belt RBSP Radiation Belt Storms Probes RE Earth radius of 6371 km REP relativistic electron precipitation RTK Real Time Kinematic SAA South Atlantic anomaly SAMA South Atlantic magnetic anomaly SAPS sub-auroral polarization streams SAR synthetic aperture radar


SBAS
space-based augmentation system SCW substorm current wedge SDO Solar Dynamics Observatory SEE single event effect(s) SEP solar energetic particle(s) SEU single-event upset SMD NASA’s science mission directorate SHINE solar, heliosphere, and interplanetary
environment program
SKA Square Kilometer Array SMEX NASA SMall EXplorer SOAP Simple Object Access Protocol SoHO Solar and Heliospheric Observatory SOLIS Synoptic Optical Long-term Investigations
of the Sun (NSO)
SPASE space physics archive search and extract SPE solar particle event SPP Solar Probe Plus (NASA mission) SSA space situational awareness SSI solar spectral irradiance STEREO Solar-Terrestrial Relations Observatory SuperDARN Super Dual Auroral Radar Network SWPC US/NOAA Space Weather Prediction
Service
SWx space weather TEC total electron content THEMIS time history of events and macroscale
interactions during substorms
TID traveling ionospheric disturbance(s) UCAR university corporation for atmospheric
research
UN United Nations WAAS wide area augmentation system WDC World Data Center of ICSU WDS World Data System of ICSU WMO World Meteorological Organization
References

American Meteorological Society, 2011. Satellite Navigation and SpaceWeather: Understanding the Vulnerability and Building Resilience.AMS.

Bk-Ste�slicka, U., Gibson, S.E., Fan, Y., Bethge, C., Forland, B.,Rachmeler, L.A., 2013. The magnetic structure of solar prominencecavities: new observational signature revealed by coronal magnetome-try. ApJL 770, L28. http://dx.doi.org/10.1088/2041-8205/770/2/L28,<arXiv:1304.7388>.

Barbieri, L.P., Mahmot, R.E., 2004. October–November 2003’s spaceweather and operations lessons learned. Space Weather 2, 9002. http://dx.doi.org/10.1029/2004SW000064.

Beer, J., 2010. Astrophysical influences on planetary climate systems. In:Schrijver, C.J., Siscoe, G.L. (Eds.), Heliophysics III: Evolving SolarActivity and the Climates of Space and Earth. Cambridge UniversityPress.

Bisi, M.M., Jackson, B.V., Hick, P.P., Buffington, A., Clover, J.M.,Tokumaru, M., Fujiki, K., 2010. Three-dimensional reconstructionsand mass determination of the 2008 June 2 LASCO coronal massejection using STELab interplanetary scintillation observations.

ApJL 715, L104–L108. http://dx.doi.org/10.1088/2041-8205/715/2/L104.

Boteler, D.H., Pirjola, R.J., Nevanlinna, H., 1998. The effects ofgeomagnetic disturbances on electrical systems at Earth’s surface.Adv. Space Res. 22, 17–27.

Burke, W.J., Weimer D.R., Maynard N.C., 1999. JGR 104, 9989–9994.Committee on a Decadal Strategy for Solar and Space Physics, 2012. Solar

and Space Physics: A Science for a Technological Society. NationalResearch Council, Washington, DC.

Committee on the Peaceful Uses of Outer Space, 2013. Report on theUnited Nations/Austria symposium on space weather data, instru-ments and models: looking beyond the International Space WeatherInitiative. Report of the Third United Nations Conference on theExploration and Peaceful Uses of Outer Space, Vienna, 19–30 July1999 (United Nations publication, Sales No. E.00.I.3) (chap. I,resolution 1, sect. I, para. 1 (e)(ii), and chap. II, para. 409 (d)(i)).

Committee on the Peaceful Uses of Outer Space, 2014. Working report ofexpert group C: space weather. United Nations. A/AC.105/C.1/2014/CRP.15.

Committee on Progress and Priorities of U.S. Weather Research andResearch-to-Operations Activities, 2010. When Weather Matters:Science and Service to Meet Critical Societal Needs. NationalResearch Council, Washington, DC.

DeRosa, M., Schrijver, C., Barnes, G., Leka, K., Lites, B., Aschwanden,M., Amari, T., Canou, A., McTiernan, J., Regnier, S., Thalmann, J.,Valori, G., Wheatland, M., Wiegelmann, T., Cheung, M., Conlon, P.,Fuhrmann, M., Inhester, B., Tadesse, T., 2009. A critical assessment ofthe feasibility of nonlinear force-free field modeling of the solarcorona. ApJ 696, 1780–1791.

DHS Office of Risk Management and Analysis, 2011. Geomagneticstorms: an evaluation of risks and risk assessments. US Department ofHomeland Security, Washington, DC.

FERC, 2013. Order No.779: Reliability standards for geomagneticdisturbances. US Federal Energy Regulatory Committee.

G8 Communique, 2013. Charter on open data signed by G8 leaders topromote transparency, innovation and accountability. UK CabinetOffice, UK presidency of the G8, 2013. <https://www.gov.uk/govern-ment/publications/open-data-charter>.

Gaunt, C.T., 2013. Reducing uncertainty – an electricity utility response tosevere solar storms. J. Space Weather Space Clim. 4, A01. http://dx.doi.org/10.1051/swsc/2013058.

Hapgood, M., 2011a. Lloyd’s 360� Risk Insight: Space Weather: ItsImpacts on Earth and the Implications for Business. Lloyd’s, London,UK.

Hapgood, M.A., 2011b. Towards a scientific understanding of the riskfrom extreme space weather. Adv. Space Res. 47, 2059–2072. http://dx.doi.org/10.1016/j.asr.2010.02.007.

Hewish, A., 1955. The irregular structure of the outer regions of the solarcorona. R. Soc. London Proc. Ser. A 228, 238–251. http://dx.doi.org/10.1098/rspa.1955.0046.

Hewish, A., Scott, P.F., Wills, D., 1964. Interplanetary scintillation ofsmall diameter radio sources. Nature 203, 1214–1217. http://dx.doi.org/10.1038/2031214a0.

International Energy Agency, 2013. Key World Energy Statistics. IEA.JASON, 2011. Impacts of Severe Space Weather on the Electric Grid

(JSR-11-320). The MITRE Corporation, McLean, VA.Kovaltsov, G.A., Usoskin, I.G., 2014. Occurrence probability of large

solar energetic particle events: assessment from data on cosmogenicradionuclides in lunar rocks. Solar Phys. 289 (1), 211–220.

Krausmann, E., 2011. The Space-Weather Awareness Dialogue: Findingsand Outlook, JRC Scientific and Technical Reports, EuropeanCommission, Joint Research Center Ispra, Italy.

Krauss-Varban, D., 2010. Particle acceleration in shocks. In: Schrijver,C.J., Siscoe, G.L. (Eds.), Heliophysics II: Space Storms and Radiation:Causes and Effects. Cambridge University Press.

Langhoff, S., Straume, T., 2012. Workshop report on space weather risksand society. NASA/CP-2012-216003, NASA Center for AeroSpaceInformation, Hanover, MD.

http://refhub.elsevier.com/S0273-1177(15)00225-2/h0005



http://dx.doi.org/10.1088/2041-8205/770/2/L28

http://dx.doi.org/10.1029/2004SW000064

http://dx.doi.org/10.1029/2004SW000064





http://dx.doi.org/10.1088/2041-8205/715/2/L104

http://dx.doi.org/10.1088/2041-8205/715/2/L104


















http://https://www.gov.uk/government/publications/open-data-charter

http://https://www.gov.uk/government/publications/open-data-charter

http://dx.doi.org/10.1051/swsc/2013058







http://dx.doi.org/10.1098/rspa.1955.0046

http://dx.doi.org/10.1098/rspa.1955.0046

http://dx.doi.org/10.1038/2031214a0

http://dx.doi.org/10.1038/2031214a0











Lloyd’s, 2013. Solar Storm Risk to the North American Electric Grid.Lloyd’s, London, UK.

Malanushenko, A., Schrijver, C.J., DeRosa, M.L., Wheatland, M.S.,2014. Using coronal loops to reconstruct the magnetic field of an activeregion before and after a major flare. ApJ 783, 102. http://dx.doi.org/10.1088/0004-637X/783/2/102, <arxiv:1312.5389>.

National Intelligence Council, 2013. Global Trends 2013: AlternativeWorlds. NIC.

NERC, 2012a. NERC report on March 13, 1989 geomagnetic disturbance.<http://www.nerc.com/files/1989-Quebec-Disturbance.pdf>.

NERC, 2012b. Special Reliability Assessment Interim Report: Effects ofGeomagnetic Disturbances on the Bulk Power System. NorthAmerican Electric Reliability Corporation.

NERC, 2013a. Application Guide: Computing Geomagnetically-InducedCurrent in the Bulk-Power System. North American ElectricReliability Corporation.

NERC, 2013b. Geomagnetic Disturbance Planning Guide. NorthAmerican Electric Reliability Corporation.

Newell, P., Sotirelis, T., Liou, K., Meng, C.I., Rich, F.J., 2007. A nearlyuniversal solar wind-magnetosphere coupling function inferred from10 magnetospheric state variables. JGR 112, 10:1029/2006JA012015.

Organization for the Economic Cooperation and Development, 2011.Geomagnetic Storms. OECD/iFP futures project on ‘Future globalshocks’.

Pham, N., 2011. The economic benefits of commercial GPS use in the USand the costs of potential disruption. NDP consulting.

Population Reference Bureau, 2013. 2013 World Population Data Sheet.PRB.

Rodgers, D., Coates, A., Johnstone, A., Daly, E., 1998. Correlation ofMeteosat-3 anomalies with data from the space environment monitor.

In: Space Weather (ESA Workshop). European Space Agency, pp.301–306.

Royal Academy of Engineering, 2013. Extreme Space Weather: Impactson Engineered Systems and Infrastructure. RAE.

Satellite Industry AssociationTauri group, 2013. State of the SatelliteIndustry Report. SIA.

Schrijver, C.J., Beer, J., 2014. Space weather from explosions on the Sun:how bad could it be?. EOS Trans. 95 201–202. http://dx.doi.org/10.1002/2014EO240001.

Schrijver, C.J., Mitchell, S.D., 2013. Disturbances in the US electric gridassociated with geomagnetic activity. J. Space Weather Space Clim. 3,A19. http://dx.doi.org/10.1051/swsc/2013041, <arXiv:1304.5489>.

Schrijver, C., Rabanal, J.P., 2013. A survey of customers of space weatherinformation. Space Weather J. 11, 529–541.

Schrijver, C.J., Dobbins, R., Murtagh, W., Petrinec, S.M., 2014. Assessingthe impact of space weather on the electric power grid based oninsurance claims for industrial electrical equipment. Space Weather 12,487–498. http://dx.doi.org/10.1002/2014SW001066, arXiv:1406.7024.

Schulte in den Baumen, H., Moran, D., Lenzen, M., Cairns, I., Steenge,A., 2014. How severe space weather can disrupt global supply chains.Nat. Hazards Earth Syst. Sci. 14, 2749–2759.

Space Studies Board, 2008. Severe Space Weather Events – UnderstandingSocietal and Economic Impacts. National Academy Press,Washington, DC, USA.

Trichtchenko, L., Kalugin, G., Armitage, J., Boudjemline, K., Waller, D.,2013. Feasibility study of using Muon observations for extreme spaceweather early warning (final report). Geological Survey of Canada 7451.

Weimer, D.R., King, J.H., 2008. JGR (Space Physics) 113 (A7), Cite IDA07106.

World Economic Forum, 2013. Global Risks 2013, eighth ed. WEF.



http://dx.doi.org/10.1088/0004-637X/783/2/102

http://dx.doi.org/10.1088/0004-637X/783/2/102



http://www.nerc.com/files/1989-Quebec-Disturbance.pdf






















http://dx.doi.org/10.1002/2014EO240001

http://dx.doi.org/10.1002/2014EO240001




http://dx.doi.org/10.1002/2014SW001066










Documents

Understanding space weather to shield society: A global ... · Earth’s magnetic ﬁeld and upper atmosphere. This results in a variety of manifestations, including geomagnetic vari-ability,