Polarimeter and Lidar evaluation: the ACEPOL field campaign · Polarimeter and Lidar evaluation: the ACEPOL field campaign Kirk Knobelspiesse1, Richard Ferrare2, Otto Hasekamp3, Felix

Polarimeter and Lidar evaluation: the ACEPOL field campaignKirk Knobelspiesse1, Richard Ferrare2, Otto Hasekamp3, Felix Seidel4, Arlindo da Silva5, Hal Maring4

1Ocean Ecology Laboratory, NASA GSFC, 2NASA LaRC, 3SRON, 4NASA HQ, 5GMAO, NASA GSFC

Four multi-angle polarimeters and two lidars were deployed on the ER-2 for the AerosolCharacterization from Polarimeter and Lidar (ACEPOL) field campaign. ACEPOL data will testinstrumentation and algorithms for remote sensing of clouds and aerosols, relevant to plannedmissions such as ACE, PACE and MAIA.

Earth Sciences Division – Hydrosphere, Biosphere, and Geophysics

Name: Kirk Knobelspiesse, Ocean Ecology Laboratory (616), NASA GSFCE-mail: [email protected]: 301-614-6242

References:K. Knobelspiesse, Q. Tan, J. Redemann, B. Cairns, D.J. Diner, R.A. Ferrare, G. van Harten, O.P. Hasekamp, O.V. Kalashnikova, J.V. Martins, J.E. Yorks, F.C. Seidel (2017), A21—2155 Multi-angle polarimeter inter-comparison: The PODEX and ACEPOL field campaigns, presented at 2017 American Geophysical Union Fall Meeting, New Orleans, LA, 11-15, Dec.

Additional research and publications underway. Also see the ACE Polarimeter Working Group website: https://earthscience.arc.nasa.gov/acepwg/content/ACE_Polarimeter_Working_Group_ACEPWG, and the feature at the NASA AFRC website:https://www.nasa.gov/centers/armstrong/features/prototype_space_sensors.html

Data Sources: Each instrument team archives data individually. Calibrated observations will be made available by the ACE team meeting in spring, 2018.

Technical Description of Figures:Left: Pilot’s photo from the ER-2, during an overflight of smoke from controlled burns in Arizona. Photo by Stu Broce. Inset: ACEPOL logo, showing location of the two lidars (CPL, HSRL-2), and four multi-angle polarimeters (AirMSPI, AirHARP, RSP, SPEX-airborne) on the ER-2.Top right: Tracks for the nine ACEPOL science flights. A total of 41 flight hours were flown between 19 Oct. and 9 Nov., 2017. Bottom right: A portion of the ACEPOL science, engineering, aircraft and project team following a successful flight.

Scientific significance, societal relevance, and relationships to future missions: ACEPOL targeted a wide variety of scene types (desert, cropland, urban, ocean) and conditions (cloudy, clear, polluted, smoky, pristine) to provide theideal test, comparison and development dataset for polarimeter and lidar observations of aerosols, clouds, and the ocean. The polarimeters can be considered prototypes for designs that may be used for the Decadal Survey recommended ACE mission, and are similar to polarimeters that will be used for the PACE and MAIA missions. ACEPOL also performed underflights of the CALIPSO instrument for calibration and evaluation. Funding was provided by the ACE mission, by the Netherlands Institute for Space Research (SRON) and the CALIPSO mission.


Land Surface Models Increasing in Skill and SimilaritySujay Kumar1, Shugong Wang1,3, David Mocko1,3, Christa Peters-Lidard2 , Youlong Xia4

1Hydrological Sciences Laboratory, NASA/GSFC, 2Earth Science Division, NASA/GSFC, 3SAIC , 4NOAA EMC


The output averaged over multiple computer model simulations (an ensemble) is often superior to the output from any individual model simulation. This study examines how similar/ dissimilar the outputs from a suite of land surface models are to each other, and quantifies where, when, and why the models are similar or dissimilar.

The results indicate that recent model development efforts have increased the skill of the land surface models, while pushing them to be more similar to each other. The similarity assessment serves as a benchmark for the assemblage and evaluation of multi-model ensembles.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.90

0.10.20.30.40.50.60.70.80.9

0 0.02 0.04 0.06 0.08 0.1

0 0.2 0.4 0.6 0.8Normalized Similarity (-)

0

0.2

0.4

0.6

0.8

Norm

alize

d Acc

uracy

(-)

0 0 02 0 04 0 06 0 08 0 1


0

0.2

0.4

0.6

0.8

Norm

alize

d Acc

uracy

(-)

0 0 02 0 04 0 06 0 08 0 1


0

0.2

0.4

0.6

0.8

Norm

alize

d Acc

uracy

(-)

0 0 02 0 04 0 06 0 08 0 1


0

0.2

0.4

0.6

0.8

Norm

alize

d Acc

uracy

(-)

0 0 02 0 04 0 06 0 08 0 1

VIC412L

CLSM25Mosaic

VIC403

Current Model Versions New Model Versions


0

0.2

0.4

0.6

0.8

Norm

alize

d Acc

uracy

(-)

0 0 02 0 04 0 06 0 08 0 1

Noah28


0

0.2

0.4

0.6

0.8

Norm

alize

d Acc

uracy

(-)

0 0 02 0 04 0 06 0 08 0 1

Noah36


0

0.2

0.4

0.6

0.8

Norm

alize

d Acc

uracy

(-)0 0 02 0 04 0 06 0 08 0 1

NoahMP

Name: Sujay V. Kumar, Hydrological Sciences Laboratory, NASA GSFCE-mail: [email protected]: 301-286-8663

References:Sujay V. Kumar, Shugong Wang, David M. Mocko, and Christa D. Peters-Lidard, Y. Xia, 2017. Similarity Assessment of Land Surface Model Outputs in the North American Land Data Assimilation System (NLDAS), Water Resources Research, 53, 8941-8965, doi:10.1002/2017WR020635.

Data Sources: NLDAS-2 model outputs available from the NASA GES DISC; NLDAS Science Testbed developmental model outputs

Technical Description of Figures:Graphic 1: Probability density of latent heat flux estimates from land surface models (LSMs) as a function of the similarity (x-axis) and accuracy (y-axis). The operational versions of the models (Noah version 2.8, Mosaic and VIC version 4.0.3) are shown in the left column and the ”new” models (Noah version 3.6, Catchment version Fortuna 2.5, VIC version 4.1.2.l and Noah-MP) are shown on the right column. The figure indicates that the similarity level has increased in the new models, which raises questions about their true utility to a multi-model ensemble.

Scientific significance, societal relevance, and relationships to future missions: Multi-model ensembles are often used to produce ensemble mean estimates with increased simulation skill of model outputs. If multi-model outputs are too similar, they add little additional information to the multi-model ensemble, whereas if the models are too dissimilar, it may be indicative ofsystematic errors in their formulations or configurations. The study presents a formal similarity assessment of multi-model ensemble outputs to assess their fidelity and utility to the ensemble, using a confirmatory factor analysis. The results indicate that both the level of similarity and accuracy has increased in the new versions of the models for estimating key terrestrial water budget terms.

This study will be used in the selection of the models for the next phase of the North American Land Data Assimilation System(NLDAS), which supports a variety of applications including drought assessment, flood risk estimation and weather/climate initialization. NLDAS is run operationally at NOAA, and is a collaboration project led by NOAA and by NASA. NASA’s Land Information System (LIS) software framework, which is the software platform that runs these models, is being transitioned into operations at NOAA for the development of the next phase of NLDAS. LIS will be used to incorporate several land relevant NASA remote sensing measurements (e.g. SMAP soil moisture, GRACE terrestrial water storage etc.) into these LSMs in the operational environments at NOAA.


Temperature and snowfall in western Queen Maud Landincreasing faster than climate model projections

Brooke Medley1, Joseph McConnell2, Thomas Neumann1, Carleen Reijmer3, Nathan Chellman2, Michael Sigl4, and Sepp Kipfstuhl5

1Cryospheric Sciences Laboratory, NASA GSFC, 2Desert Research Institute, 3Utrecht University, 4Laboratory of Environmental Chemistry, Paul Scherrer Institute, 5Alfred Wegener Institute


An ice core record from Queen MaudLand indicates modern day snowfall issignificantly larger than the prior 2,000years. A 19-year temperature recordshows the region is warming by over1°C decade-1. These observedincreases are more rapid than anensemble of climate model projections,suggesting modern conditions are notrepresentative of the long-term mean.

Figure 1

Figure 3

Figure 2

(Ice Core)

(Climate Model)

Snow

Name: Brooke Medley, Cryospheric Sciences Lab, NASA GSFCE-mail: [email protected]: 301-614-6705

References:Medley, B., McConnell, J. R., Neumann, T. A., Reijmer, C. H., Chellman, N., Sigl, M. & Kipfstuhl, S. (2017). Temperature and snowfall in western Queen Maud Land increasing faster than climate model projections. Geophysical Research Letters, 44. https://doi.org/10.1002/2017GL075992.

Data Sources: The NASA MERRA-2 reanalysis total precipitation and evaporation products were obtained from NASA’s Earthdatawebsite. The ERA-Interim fields were downloaded from the ECMWF website, CFSR and CESM-LE fields from the NCAR website.

Technical Description of Figures:

Figure 1: Location of the study site encircled in yellow where both the ice core and AWS are located. The ice core record is referred to as “B40” and is labeled as such. It is located in western Queen Maud Land, East Antarctica and the inset map puts the map in a larger scale context. Colored shading is surface velocity.

Figure 2: Ice core (B40) and model-derived (CESM) annual time series (and a bold 20-year running mean) of snowfall at the B40 site in mm of water equivalent. The B40 record is truncated (1500-2010 CE) but is over 2,000 years long. Dashed lines represent the 2-sigma bounds based on preindustrial values (prior to 1850). The CESM data represent twenty-nine 1920-2100 CE ensemble runs (grey shading), the red line is one ensemble (1850-2100 CE), and the green represents a preindustrial run (~1850 CE). The black vertical line shows where the B40 record surpasses the +2-sigma bound, which occurs much later in the CESM data.

Figure 3: The 1998-2016 AWS temperature record, which shows significant annual warming of ~1.1°C per decade.

Scientific significance, societal relevance, and relationships to future missions: Antarctic snowfall is direct control on sea level. Climate model projections suggest that increased snowfall over Antarctica, largely due to atmospheric warming, will partly offset dynamic ice losses by the end of the century. There are very few observations, however, that can be used to differentiate between the varying abilities of the climate models to accurately reproduce Antarctic precipitation and temperature. Here, we use a 2,000-year record of snowfall and a 19-year record of air temperature to evaluate whether an ensemble of climate model simulations is able to accurately reproduce the observed change at a site in western Queen Maud Land in East Antarctica, where we find both snowfall and air temperature are significantly increasing. The rates of increase in both snowfall and temperature within the simulations are too low, which suggests a subdued mitigating impact of enhanced snowfall into the future. Observations, especially ones covering a long time period, are few in number but of critical importance in understanding and interpreting climate model projections over the Antarctic Ice Sheet.


Progress towards an Atom Interferometer Gravity Gradiometer A. Sugarbaker1, S.B. Luthcke2, B. Saif2

1AOSense Inc., 2 NASA Goddard Space Flight Center

Significant progress has been achieved in the development of an Atom Interferometer Gravity Gradiometer (AIGG) for earth sciences. A major milestone has recently been achieved, the first fringe measurements by the prototype instrument. Once completed the prototype instrument will pave the way for a spaceborne instrument with ~10 micro-Eötvössingle shot precision, enabling observations of Earth time variable gravity signals with two orders of magnitude improvement in RSS accuracy across a spatial band down to 200 km resolution.


Gravity GradiometerFirst Fringe Measurements

Projected Gravity Model Improvement

Name: Scott B. Luthcke, Geodesy and Geophysics Lab, NASA GSFC E-mail: [email protected]: 301-614-6112

References:Rakholia et al. Development of an Atom Interferometer Gravity Gradiometer for Earth Sciences. Poster presented at: The 48th Annual Meeting of the APS Division of Atomic, Molecular and Optical Physics (DAMOP); June 7, 2017; Sacramento, California.

Luthcke et al. Development and Performance of an Atomic Interferometer Gravity Gradiometer for Earth Science, Oral presentation (invited), American Geophysical Union, Fall Assembly 2016, G53A-06.

Data Sources: The fringe measurements are from the prototype gradiometer instrument lab tests. The GRACE gravity error spectrum is from GFZ calibrated error spectrum. The hydrology spectrum is from the Global Land Data Assimilation System (GLDAS)/Noah Land Surface Model (Noah), while the AIGG gravity error spectrum is from a rigorous simulation.

Technical Description of Figures: The images document recent progress in the development of the Atomic Interferometer Gravity Gradiometer (AIGG) prototype instrument as well as the projected performance of a spaceborne instrument in recovering the Earth’s monthly gravity signal.

Gravity Gradiometer: The image shows the AIGG instrument in the laboratory with its two meter baseline and two gravimeter sensor heads. The large white structure is not part of the instrument, but a mechanical structure to hold the instrument in the laboratory. While the instrument has a two meter baseline its other dimensions are relatively modest enabling the instrument to be housed in a common spacecraft bus.

First Fringe Measurements: The figure presents the first fringe measurements observed by the lower gravimeter sensor head of the gradiometer. The lower graph in the figure shows the fractional atoms in one state which is equal to the probability (y-axis) of finding the atom in that state. The sinusoidal curve comes from modulating the laser phase (sweeping the phase shown in the x-axis) and extracting the induced phase due to gravity from fitting the curve and finding where the biased phase is zero. The top graph shows the residuals of the observed phase from the sinusoidal phase sweep. The phase measurements are the instrument’s fundamental observations that enable gravity observations.

Projected Gravity Model Improvement: The figure shows a gravity spectrum of the earth’s monthly (April 2009) hydrology from the GLDAS model, the error spectrum for a monthly un-regularized spherical harmonic gravity field solution, and an estimate of the dramatic reduction in monthly un-regularized spherical harmonic gravity field error from a spaceborne AIGG in a ~330 km polar orbit estimated from a detailed simulation.

Scientific significance, societal relevance, and relationships to future missions: Observing the Earth’s surface mass variation from spaceborne gravimetry (e.g. the GRACE mission) has revolutionized our understanding of the Earth’s water cycle including hydrology, land ice, and inland seas and oceans, and its response and impact on climate and the impact on humans. The recent National Academies earth science decadal survey has called out as a top priority the continuation of mass change observations, and to significantly improve the spatial resolution of the mass change observations down to hydrological basin scales of ~50 km. This represents nearly an order of magnitude improvement and will require new technologies and approaches such as the AIGG.


GEDI and TanDEM-X Fusion for Enhanced Height/Biomass 1T. E. Fatoyinbo, 1,2 S.-K Lee, 2S. Hancock, 2R. Dubayah

1Biospheric Sciences Laboratory, NASA GSFC, 2Department of Geographical Sciences, UMD

A new fusion technique that will combine GEDI (Lidar) and TanDEM-X (SAR) data to generate higher resolution, wall-to-wall forest height and biomass maps has been developed using simulated GEDI metrics over dense tropical forests in Central Africa. With this method, Forest canopy height can be estimated using the Pol-InSAR technique combined with the fused GEDI/TanDEM-X Digital Terrain Model at a finer resolution (20 m) than the planned GEDI product 1km) and with a height estimation error of ~15 %.


Pol-InSARInversionusingFusionDTM

0m

55m

LVISRH95

Pol-InSAR

Inversion

R2=0.76RMSE=3.72m

LVISRH95

A A’

B B’

Figure 1

Figure 2

Figure 3

Groundlayer

A A’

B B’

0m

100m

Elevation

Name: SeungKuk Lee, Biospheric Sciences Lab., NASA GSFC E-mail: [email protected]: 301-614-7047

References:Lee, S.K., Fatoyinbo, T., Osmanoglu B., Lagomasino, D., and Feliciano, E.,”3D forest structure parameter retrieval : Polarimetric SAR interferometry and waveform lidar airborne data”, in proc. Int. Geosci. Remote Sens. Symp. (IGARSS), Fort Worth, USA, Jul. 2017.

Data Sources: GEDI data was simulated from LVIS airborne waveform lidar data. The LVIS data was acquired during NASA AfriSARcampaign in 2016. TanDEM-X data from German Aerospace Center (DLR).

Technical Description of Figures: The images show 3D forest structure information fusing the (simulation) GEDI and TanDEM-X data. The forest height was estimated using polarimetric SAR interferometry (Pol-InSAR) technique. The height inversion uses the fusion DTM generated from low-resolution interpolated GEDI DTM and high-resolution TanDEM-X DEM by means of wavelet transformation.

Figure 1: Forest height maps at 20 m resolution over Mondah forest in Gabon, scaled from 0 m to 50 m. Upper image represents LVIS rh95 height and bottom image shows Pol-InSAR inversion result from fusion of GEDI simulation data and single-baseline TanDEM-X data at HH polarization.

Figure 2: The validation result between airborne lidar and GEDI + TanDEM-X fusion product showed a good correlation (R2: 0.76 and RMSE: 3.72 m).

Figure 3: Forest vertical structure profiles from the GEDI and TanDEM-X fusion product.

Scientific significance, societal relevance, and relationships to future missions: The GEDI instrument will operate on the International Space Station (ISS) and laser ranging measurements for 3D structure of the Earth from 2019. The GEDI instrument will not providecontinuously imaging on the ground resulting in gaps between ground tracks and adjacent swath at the footprint level. In order to provide wall-to-wall mapping of 3D forest structure and biomass while maintaining the fine resolution measurement of each footprint, GEDI data will be combined with single-pass interferometric SAR data from the TanDEM-X mission. The approach indicates a great possibility for generating global-scale forest map and biomass maps with unprecedented spatial resolution, which will help advance studies of global carbon stocks, biodiversity and water resources.


Midnight on Pine Island Glacier – Ice Thicknesses from Landsat 8Christopher A. Shuman, Cryospheric Sciences Laboratory, NASA GSFC, UMBC JCET

Two views of the remnants of Pine Island Glacier’s Iceberg B-44 from a Landsat 8 ascending scene acquired near local midnight on Dec. 15, 2017. The left panel (above) shows a strong thermal contrast between the ice front of the glacier, the iceberg fragments, and the sea ice cover farther out in Pine Island Bay relative to a bright, white, warm polynya. The right panel (above) shows the deeply shadowed rifts due to the low sun angle at this time (8.05°) as well as the location of the inset (at far right).

The length of the shadows at the solar azimuth give the height above the sea (~49 m) from the sun angle and the thickness of the iceberg (~315 m) can then be estimated (after Scambos et al. 2009; Bindschadler et al., 2011; Lambrecht et al., 2007).

Pine Island Bay

EvansKnoll

sea ice

Pine Island Bay

EvansKnoll

sea ice

Thermal Infrared Sensor Band 10 Operational Land Imager Band 8

Operational Land Imager Band 8


Name: Christopher A. Shuman, Cryospheric Sciences, NASA GSFC and UMBC JCETE-mail: [email protected], [email protected]: 301-614-5706

References:EOSDIS Worldview - https://go.nasa.gov/2CxDje4 (note ‘sea smoke’ blowing off the polyna; flicker 9/22 and the day before to see berg calving motion)Rift on Pine Island Glacier - https://earthobservatory.nasa.gov/NaturalHazards/view.php?id=91022Pine Island Glacier lost 267 km2 https://twitter.com/StefLhermitte/status/911664879135854593US NIC Iceberg B-44 - http://www.natice.noaa.gov/doc/USNIC%20-%20Iceberg%20B-44%20Press%20Release%20-%2025SEP17.pdf and US US NIC Tracking Criteria -http://www.natice.noaa.gov/doc/Notice_Iceberg_Tracking_Criteria.pdfB-44 Disintegrating - https://twitter.com/StefLhermitte/status/920784272692215808The Quick Demise of B-44 - https://earthobservatory.nasa.gov/IOTD/view.php?id=91203Bindschadler et al. (2011), Variability of basal melt beneath the Pine Island Glacier ice shelf - https://doi.org/10.3189/002214311797409802Scambos et al. (2009), Ice shelf disintegration by plate bending and hydro-fracture: Satellite observations and model results of the 2008 Wilkins ice shelf break-ups -https://doi.org/10.1016/j.epsl.2008.12.027Lambrecht et al. (2007), New ice thickness maps of Filchner–Ronne Ice Shelf, Antarctica, with specific focus on grounding lines and marine ice -https://doi.org/10.1017/S0954102007000661Dutrieux et al. (2013), Pine Island glacier ice shelf melt distributed at kilometre scales - https://doi.org/10.5194/tc-7-1543-2013

Data Sources: The Landsat 8 images were identified in GloVis (https://glovis.usgs.gov/), processed, and then annotated by Chris Shuman. Image processing was done with PCI Geomatica with additional image manipulation in Adobe Photoshop and annotations added in Powerpoint.

Technical Description of Figures:Figure 1 (left panel): Shows the fragments of Iceberg B-44 using Landsat 8’s Thermal Infrared Sensor (TIRS) relative to a strong polynya at the front of Pine Island Glacier (PIG). PIG is a major outlet glacier, ~35 km wide at its terminus, and flows in a northwestly direction into Pine Island Bay, part of Antarctica’s Amundsen Sea embayment. A large polynya or area of open sea water lies between PIG’s front and the remnants of B-44 and indicates the presence of upwelling warm water (bright white) relative to the darker greys of the glacier, iceberg fragments, and nearby sea ice.

Figure 2 (right panels): Shows the front of Pine Island glacier and the fragments of Iceberg B-44 using Landsat 8’s Operational Land Imager (OLI) with the large polynya now indicated by black, open water relative to the sun-lit shades of grey on the ice surfaces. Because this clear imagery was acquired on an ascending pass (Path 157 Row 131) at 7:03 UTC (Figure 2), the acquisition was near midnight local time at 100°W longitude with a solar azimuth of 174.5° (north is up in all images). With the austral summer solstice near on Dec. 15, the sun elevation was 8.05° at this time of day. As a result, the topography in the area, including Evans Knoll to the northeast (right) of Pine Island Glacier, casts very distinct shadows and the ice edges facing south are strongly illuminated. Further, the length of the shadows from the iceberg fragments allows the freeboard of the fragments to be accurately measured where irregularities in the ice edges indicate the solar azimuth (see inset). An average iceberg freeboard of 48.7 m was found at ten locations. Using a formula adapted from Scambos et al. (2009) and Lambrecht et al. (2007), a thickness of 315 m (~1033 ft) for the iceberg and adjacent ice front can be calculated. This value means that the high basal melt rates found by Dutrieux et al. (2013) and Bindschadler et al. (2011) have reduced the grounded ~1.3 km (0.8 mi) thick Pine Island Glacier by 75% as it has extended into Pine Island Bay.

Scientific significance, societal relevance, and relationships to future missions: The continuing observations from Landsat 8 in both thermal and visible channels enable insights on ongoing changes to icebergs large and small even during periods of total darkness. This imagery also provides compelling visuals for the public, colleagues, and policymakers. Other observations, past and present, will be used to assess changes and front positions for Pine Island Glacier and other major ice sheet outlets. See also: Pine Island Glacier Retreat, Antarctica - https://svs.gsfc.nasa.gov/30914


Documents

Polarimeter and Lidar evaluation: the ACEPOL field campaign · Polarimeter and Lidar evaluation: the ACEPOL field campaign Kirk Knobelspiesse1, Richard Ferrare2, Otto Hasekamp3, Felix