XA9743627

IC/IR/96/14 INTERNAL REPORT (Limited Distribution)

United Nations Educational Scientific and Cultural Organizationand

International Atomic Energy Agency

INTERNATIONAL CENTRE FOR THEORETICAL PHYSICS

PROCEEDINGSOF THE IRI TASK FORCE ACTIVITY 1995

Sandro M. Radicella

MIRAMARE - TRIESTE

May 1996

VOL 28 ife 09

PROCEEDINGS OF THE IRI TASK FORCE ACTIVITY 95

FOREWORD

This ICTP Internal Report contains the programme, conclusions and the write up of a number of presentations delivered during the International Reference Ionosphere (IRI) Task Force Activity 95 that have taken place at the ICTP during November 1995 and particularly centred in the week from November 13 to 17. The 1995 Task Force Activity is the second successful encounter of specialists organized by the URSI-COSPAR IRI Working Group and the Aeronomy and Radiopropagation Laboratory of the International Centre for Theoretical Physics of Trieste, Italy.

Prof. Sandro M. Radicella HeadICTP Aeronomy and Radiopropagation Laboratory

1

TABLE OF CONTENTS

Introduction and Proposed Objectives and Specific Problems.................................................. iiFinal Programme.......................................................................................................................... iiiConclusions and Action Items........................................................................................................ vList of Participants......................................................................................................................... viSTATUS REPORT AND DATA AVAILABILITY

Ionospheric Data Available on CD-Rom and on NDADS, D. Bilitza.................................1FI and Bottomside Region in IRI - Status Report, D. Bilitza........................................... 5

ELECTRON DENSITY PROFILE SHAPE BELOW NmaxElectron Density Profiles for Middle Latitudes, S.-R. Zhang, S.M. Radicella,

X.-Y. Huang and M.-L. Zhang......................................................................................... 13Constraints on the Bottomside Electron Density Profile Shape Parameters,

T.L. Gulyaeva.................................................................................................................... 20Investigation on Equatorial Ionospheric Profiles and IRI Model,

J. O. Adeniyi....................................................................................................................... 25Comparison of the Observed Results of the Electron Density Profiles

with the IRI90, M.-L. Zhang, S.M. Radicella and K.-L. Dai....................................... 32The Shape of the Bottomside N(h) Profile at Two Middle Latitude Stations,

M.M. de Gonzalez and B. Zolesi...................................................................................... 39INTERMEDIATE REGIONS (FI) ELECTRON DENSITY PROFILE

The Presence of the FI Layer Over a Low Latitude Station, M.M. de Gonzalez,R. Ezquer and R. del V. Oviedo........................................................................................46

On the Application of the DU Charme Formula for Predicting the FI LayerCritical Frequency, B. Zolesi and M.M. de Gonzalez..................................................... 52

Variability of the Electron Density at Fixed Heights Near FI Region,M.-L. Zhang and S.M. Radicella......................................................................................55

A Summary of the Theoretical Model Results for the Ionospheric Fl-Ledge,S. -R. Zhang and S.M. Radicella......................................................................................... 63

11

INTRODUCTION

This project continues the IRI Task Force Activities at the International Centre for Theoret­ical Physics (ICTP) in Trieste, Italy. This year’s task force has focus on the model descriptions for the bottomside F-region including the FI layer.

The main discussions and presentations have taken place in the week from November 13 to 17. The format has been similar to last year’s activity with presentations and round-table discussions in the morning and follow-on work in small subgroups in front of computer terminals in the afternoon.

PROPOSED OBJECTIVES AND SPECIFIC PROBLEMS

Topic 1 - To analyze the results of the follow up work on the FI layer in the IRI model basedon the conclusions of the 1994 IRI Task Force Activity.

1.1. - Behaviour of Ne at fixed heights in the FI region and modelling efforts to describe such behaviour.

1.2. - Critical analysis of the present IRI model description of characteristic points in the FI region (foFl, dN/dh min).

1.3. - Use of theoretical models to study the FI layer formation criteria and to improve the IRI model.

Topic 2 - To discuss the temporal and spatial range of validity of the B parameters adopted inthe IRI model to describe the shape of the bottomside profile.

2.1. - Comparison of the IRI model description of the Ne profile shape below NmF2 with ionogram inversion profiles obtained at different locations and times.

2.2. - Results from theoretical models and comparison with data and IRI.

2.3. - Critical analysis of the validity of the bottomside profile shape options adopted by the IRI model.

2.4. - Analysis of the possible differences between monthly mean and instantaneous profile shape below NmF2.

FINAL PROGRAMME

Monday 13 November:

15.00 - 15.45 IRI Status Report D. BILITZA

15.45 - 16.00 Report on the previous Task Force Activity D. BILITZA

16.00 - 16.30 Coffee Break

16.30 - 17.00 Outline of the ProgrammeS.M. RADICELLA

Tuesday 14 November:

09.00 - 10.30 Experimental data: availability and limitations D. BILITZAT. GULYAEVA B. REINISCH

10.30 - 11.00 Coffee Break

11.00 - 12.30 Experimental data: availability and limitations S.M. RADICELLAB. REINISCH R. LEITINGER

14.00 - 16.00 Laboratory work

16.00 - 18.00 Computer demonstration of CD-Rom Data

Wednesday 15 November:

09.00 - 10.30 Electron density profile shape below NmaxS. R. ZHANGT. GULYAEVA B. REINISCH J. O. ADENIYI

10.30 - 11.00 Coffee Break

11.00 - 12.30 Electron density profile shape below Nmax M. L. ZHANGM. MOSERT DE GONZALEZB. ZOLESIS. M. RADICELLA

14.00 - 15.00 Topside ionosphere: present time issues R. LEITINGER

15.00 - 17.00 Laboratory work

Thursday 16 November:

09.00 - 10.30 Intermediate region (FI) electron density profile M. MOSERT DE GONZALEZM.-L. ZHANG B. REINISCH

10.30 - 11.00 Coffee Break

11.00 - 12.30 Intermediate region (FI) electron density profile B. ZOLESIS.-R. ZHANG M. MOORHEAD

14.00 - 17.00 Laboratory work

Friday 17 November:

09.00 - 10.30 General discussion on future activities

10.30 - 11.00 Coffee Break

11.00 - 12.30 Final discussion

V

CONCLUSIONS AND ACTION ITEMS

A. Optimizing the F2 profile shape below the peak

-1. Find better BO and B1 parameters by fitting IRI’s N(x) function to measured profiles, where

N(x) - NmF2. [exp(-eBl)]/coshx; x = (hm-h)/B0.

The profile fitting should be performed between x=0, i.e. the F2 layer peak, and either

a. x=l, i.e. f=0.4883foF2, if foFl < 0.4883foF2, orb. f = foFl if foFl > 0.4883foF2.

If no FI layer exists, case (a) applies. Diurnal, seasonal, latitudinal and sunspot cycle dependence must be established. For this profile fitting process, foFl frequencies should be specified in all cases when the ionogram indicates L conditions. It is recommended to use average representative profiles like ARP (Reinsich and Huang, 1995, Annali di Geofisica) for this study, but individual profiles may also be used.

-2. Investigate the base point (Mosert de Gonzalez and Radicella, 1990, Adv. Space Res.) and the possibility to model it.

B. Improving the FI Region Profile

-1. Determine N and dN/dh at 170 km from measured electron density profiles and satellite measurements. Compare these data with results from physical models.

-2. Model the FI layer characteristic point.

-3. Determine the probability of occurrence of FI (including L conditions) as function of time of day to improve the DuCharme et al. (1973) prediction of the critical (maximum) zenith angle.

-4. Determine the effects of FI profile variations on oblique HF wave propagation.

C. In-house Report of the November MeetingICTP will issue a report publishing the presentations made during the meeting. Manuscripts

must be submitted to ICTP on or before 31 January 1996. The text is limited to 2 pages, and figures/tables to 4 pages. The text should be submitted by email to: rsandro@ictp.trieste.it

LIST OF PARTICIPANTS

Jacob Olusegun AdeniyiUniversity of florinDept, of PhysicsFaculty of ScienceP.M.B. 1515IlorinNIGERIA

Dieter Bilitza (IRI Working Group Chaiperson) NASA, GSFC, NSSDCCode 633Hughes StxMD 20771 GreenbeltU.S.A.

Josef BoskaAcademy of Sciences of the Czech Republic Institute of Physics of the Atmosphere Bocni li/1401141 31 Prague CZECH REPUBLIC

Georgiana De Franceschi Istituto Nazionale di Geohsica Via di Vigna Murata, 605 00143 ROMA

Tamara Gulyaeva (IRI Working Group Vice-Chairperson)Russian Academy of Sciences, IZMIRAN Moscow Region142 092 Troitsk RUSSIA

Reinhart Leitinger Graz UniversityInstitute of Meteorology & Geophysics Halbarthgasse 1 A-8010 Graz AUSTRIA

Michael D.Moorhead Neptune Radar Limited Gardiners Farm GloucesterGL2 9NW Sandhurst ENGLAND

Marta Estela Mosert GonzalezCentro Astronomico El Leoncito (CASLEO)San Juan ARGENTINA

Sandro M. Radicella (Local Organizer)ICTP Aeronomy and Radiopropagation Laboratory 34100 TRIESTE

Bodo W. ReinischUniversity of Massachusetts LowellCenter for Atmospheric Research450 Aiken StreetMA 01854-3602 LowellU.S.A.

Carlo ScottoIstituto Nazionale di Geohsica Via di Vigna Murata, 605 00143 ROMA

Paolo SpallaConsiglio Nazionale delle Ricerche Istituto di Ricerca sulle Onde Elettromagnetiche Via Panciatichi, 64 50127 FIRENZE

John E.Titheridge Graz UniversityInstitute of Meteorology & GeophysicsHalbarthgasse 1A-8010 GrazAUSTRIAPermanent:Dept, of Physics University of Auckland Private Bag 92019 Auckland NEW ZEALAND

Man-Lian ZhangICTP Aeronomy and Radiopropagation Laboratory 34100 TRIESTE

Shun-Rong ZhangAcademia SinicaWuhan Institute of PhysicsP.O. Box 71010Hubei Province430071 WuhanP.R. CHINA

Bruno ZolesiIstituto Nazionale di Geofisica Via di Vigna Murata, 605 00143 ROMA

XA9743628

l

IONOSPHERIC DATA AVAILABLE ON CD-ROM AND ON NDADS

Dieter BilitzaNSSDC, GSFC, Code 633.9, HSTX, Greenbelt, MD 20771, USA

Abstract: Information is provided on two CD-ROMs (for PCs) with ionospheric data: the ionosonde CD issued by NGDC/WDC-A-STP/NOAA/Boulder and the Atmosphere Explorer CD produced by NSSDC/WDC-A-R&S/NASA/Greenbelt. We also briefly describe the iono­spheric /thermospheric data available through NSSDC’s automated mail retrieval system (NDADS) and explain the procedure for obtaining NDADS data.

1. The NSSDC/NASA Atmosphere Explorer CD-ROM

This CD-ROM was generated at the National Space Science Data Center in Greenbelt, Maryland by N. Papitashvili with help from D. Bilitza and J. King. It contains the Unified Abstract (UA) 15-second data from the Atmosphere Explorer C, D, and E satellite missions. These satellites were launched into highly elliptical orbits and were then later in the year maneuvered into circular orbits (300-400 km) with the onboard propulsion system.

Satellite time period height range inclinationAE-C Dec 73 - Dec 78 130-4300 68.1AE-D Oct 75 - Jan 76 150-3800 90.1AE-E Nov 75 - Jan 81 160-3000 19.7

More than a dozen experiments were flown on each satellite with almost identical payloads for all three, providing the following measurements:

electron density, temperatureion densities, temperature, driftion densitiesneutral densitiesneutral densities, windtotal neutral densityairglowUV (nitric oxide)low-energy electrons/ions (0.2-25 keV)photoelectron fluxessolar EUV (40-1850 Angstrom)

The CD-ROM contains the UA data for all Additional files on the CD include:

Langmuir Probe (CEP)Retarding Potential AnalyzerBIMS, MIMS Mass SpectrometersOSS, COSS Mass SpectrometersNATE instrumentMESA accelerometerVAE photometerUVNO, BUV experimentsLEE experimentPES spectrometerEUV spectrometer, photometer

three satellites (9 years) in ASCII format.

- experiment/data descriptions provided by the respective Principal Investigator teams;

- satellite/experiment brief descriptions from NSSDC’s Master Catalog;

2

- a 150 page listing of references to scientific papers based on the AE data ordered by experi­ment;

- the software for several international standard models: IRI-94, MSIS-86, CIRA-86, MSISE90,HWM93, IGRF45-95, Tobiska’s EUV model, GEOCGM conversion software (geographic <----> geomagnetic corrected).

Plotting and subsetting software (DOS) was also developed at NSSDC for this data set and can be obtained on floppy disk or can be retrieved from NSSDC’s anonymous ftp site (nssdc.gsfc.nasa.gov, /pub/cdrom/software/dos/aez.exe). The menu-driven system lets a user select the time period and parameters s/he is interested in and then plot the data or store them in a file. The IRI-94 and MSISE90 parameters are part of this menu and by selecting corresponding parameters a user can compare the measurements with the model predictions. Figure la,b shows two plot examples.

The CD-ROM can be ordered from the Request Coordination Office (CRUSO), NSSDC, GSFC, Code 633, Greenbelt, Maryland 20771, USA (request@nssdca.gsfc.nasa.gov).

2. The NGDC/NOAA Ionosonde CD-ROM

The set of two ionosonde CD-ROMs was generated at NOAA’s National Space Science Data Center in Boulder, Colorado by R.O. Conkright, K.F. O’Loughlin, and G.E. Talarski. It contains ionosonde data from about 130 sites (40 000 station years; 1.35 GByte) for the time period from 1957 to 1990. The data are mostly hourly parameters (foF2, h’f2, foFl, etc.) but there are also a few records of recent data with higher time resolution. The data are in ASCII format using the URSI/IIWG data exchange format. A map of all stations represented on the CD is shown in Figure 2.

NGDC also developed DOS software for displaying and selecting data. Data can be selected by time period, area, and parameter. Data for one month can be plotted either individually for each selected station or comparative with all selected stations displayed either in separate panels (see Figure 3) or in a single panel with different colors. All generated graphs can be stored as PCX files.

The “Ionospheric Digital Database - Worldwide Vertical Incidence Parameters” CD-ROM set (including the software on floppy disk) is available from the National Geophysical Data Cen­ter. NOAA Mail code E/GC2, 325 Broadway, Boulder, Colorado 80303. USA (roc@ngdc.noaa.gov, kfo@ngdc.noaa.gov. get@ngdc.noaa.gov; FAX: 303-497-6513).

3. Ionospheric/Thermospheric Data on NDADS

NDADS is an automated retrieval mail system that stages user-requested data from an optical disk jukebox onto NSSDC’s anonymous ftp site. The user can than ftp the data to his home account.

3.1 Available dataThe following ionospheric/thermospheric data are currently available through NSSDC’s

NDADS system:

3

DE-2 Aug 81 - Feb 83 300-1000 km Inclination: 90

electron density, temperature (LANG), ion densities, temperature, and drift (RPA/IDM), neutral densities and wind (NAGS), magnetic field (MAG-B), electric field (VEFI), energetic particles (LAPI), and the 16-second Unified Abstract file with data from LANG, NAGS, WATS, FPI, RPA/IDM. [High resolution WATS and FPI data ingest is in preparation]

San Marco D Mar 88 - Dec 88 260-620 km Inclination: 2.9

ion densities, temperature, drift (IVI), [EUV (ASSI), neutral density (DBI), electric field (EFI) data ingest is planned].

3.2 NDADS Data Request ProcedureTo obtain information about NDADS and its data content, one sends an e-mail message to:

archives@ndadsa.gsfc.nasa.gov with one of the following words in the SUBJECT lineSUBJECT: INFO information file will be sent

MANUAL manual will be sentHOLDINGS general holdings file will be sentHOLDINGS DE DE holdings file will be sentHOLDINGS SANMARCO San Marco Holdings file

To order data one writes

SUBJECT: REQUEST projectname data_type e.g., REQUEST DE LANG..ASCII

The available datatypes are explained in the DE holdings file. In the body of the message the user specifies the particular data file s/he is interested. In many cases data are ordered by year and day of year (yyddd); see holdings file for details. In addition, several of the data sets include documentation, software and inventory files that can be requested with the word DOCUMENT, SOFTWARE, INVENTORY in the body of the message. The message could, for example, look like this:

document, software inventory, 81355, 82001 83068

The requested files will then be staged to NSSDC’s anonymous ftp site and the requester will receive an e-mail message with a list of staged files including their names, location and size.

Figure Captions:

Fig. la,b Example plots made with the AE data and software.

Fig.2 Map of all ionospheric stations represented on the NGDC CD-ROMs

Fig.3 Example of a plot generated with the comperative plot option.

4

Fig. la Fig.lb

•innnv nun:ihi:];! n,:i i: r.n.i'ii v r-> " in , r

WMfalV

Bl/Vief •hi -i:, i- .r- k, h a1 i-; •!'- i-i ■■ !. : "ir it :v i

GlBlI.MiihKfar,:Bl/ISBH

M -l: i- -r- i- h a !■: a > v " i • ■.; v •, i

Fig.2 F'g-3

XA9743629

5

FI AND BOTTOMSIDE REGION IN IRI - STATUS REPORT

Dieter BilitzaNSSDC, GSFC, Code 633/HSTX, Greenbelt, MD 20771, USA

Abstract: In this presentation we discuss the development and current status of the IRI electron density model in the bottomside and FI region. We explain the mathematical formulas used and the underlying data base. Particular focus will be on the various parameters defining the bottomside thickness and on the merging of the F2 and FI layers. This status report was presented as an introduction to the special IRI task force activity on the bottomside/Fl-region layer shape that convened at ICTP, Trieste in November of 1995.

1. Introduction

The IRI model represents the ionospheric electron density from 50 km up to 2000 km. The model profile is divided in several regions, and the profile in a specific region is normalized to the density of the closest peak, e g. the topside and bottomside to the F2 peak and the E and valley region to the E peak. A sketch of the different regions and peaks is shown in Fig. 1. This paper focuses on the region between the F2 peak (hmF2 in Fig.l) and the lower end of the FI region (HZ in Fig.l). A full description of the profile in all regions can be found in the IRI-90 report (Bilitza, 1990).

2. Bottomside

The bottomside is defined here as the region from the F2 peak down to the FI ledge (hmFl in Fig.l). The IRI model description in this region was developed by Ramakrishnan and Rawer (1972) based on composite ionograms from several middle and low latitude sites. They found that the typical shape is best represented by the formula

N(h)/NmF2 = f(x) = exp{—x~Bl}/cosh{z}, x = (hmF2 — h)/B0 (1)

B0 is a thickness parameter that defines the height differences from hmF2 to the point where the profile has dropped down to 0.24*NmF2 (x=l). B1 determines the profile shape (see Fig.2). With their ionosonde database Ramakrishnan and Rawer (1972) established the B0 values given in Table 1 and B 1=3 as best choice.

2.1 The parameter BOUsing the B0 values from Table 1 IRI then applies a linear interpolation in R (12-month­

running mean Zurich sunspot number) and a smooth transition from a constant day to a constant night value. An interpolation scheme is also applied in modified dip latitude using the middle and low latitude values in Table 1 as anchor points. An earlier version of this in­terpolation procedure was flawed in that it did not consider the seasonal differences between the Northern and Southern hemisphere (Peres, 1987). This was corrected in IRI-86. A smooth

6

analytical description of the combined latitudinal and seasonal changes was implemented in January 1993 (Subroutine BOTAB instead of BOPOL).

SHORTCOMING: The original BO data were primarily from mid-latitudes thus the equato­rial and low latitudes are not well represented; in most cases IRI underestimate the actual bottomside thickness.

RECOMMENDATION: Since the largest BOs are observed close to the magnetic equator it would be important to add to Table 1 typical values for the equatorial ionosphere, e.g., BO deduced from Jicamara incoherent scatter profiles.

2.2 The parameter B1Currently the parameter B1 is kept constant at a value of 3 in most cases and is only varied

to facilitate the merging between the F-region and E-region parts of the model. The first step in the merging procedure is to find the height where the bottomside function reaches a value equal to the E peak density NmE. If such a point cannot be found, than B1 is increased in steps of 0.5 (up to 5) until the bottomside function satisfies the merging criteria.

SHORTCOMINGS: Profile shape depends on merging procedure! The steps in B1 can lead to temporal or spatial discontinuities. B1 is not used to describe the variability of the profile shape.

RECOMMENDATION: Establish global database of BO and B1 deduced from bottomside profile obtained by ionosondes and incoherent scatter radars. Use B1 to describe better the profile shape in the region from hmF2 down to hmF2-B0.

2.3 The half-density height h0.5The half-density point h0.5 is the height below the F-peak where the density has fallen off

to half the peak value: N(h0.5) = 0.5*NmF2. This height is closely correlated to hmF2 (e.g. Reinisch, 1991). With ionosonde data Gulyaeva (1987) has established the following formula for the ratio between h0.5 and hmF2 (the so-called G-factor)

G = h0.b/hmF2 = 0.6 + 0.2/(l + exp{-(X - 20s)/15}) (2)

where X is the solar zenith angle in degree and s is a seasonal parameter taking the values 1,2,3 for winter, equinox, and summer (see Fig.3). In IRI s is given in analytical form by s = 2 - cos{ 0.0172*d} where d is the day of the year. For a profile function of type (1) the following relationship holds between G and the parameter BO

BO = hmF2 * (1 — G)/C, (3)

where C is a constant that depends on B1 and takes the values C = 0.755566, 0.778596, 0.797332, 0.812928, 0.826146 for Bl=3, 3.5, 4, 4.5, 5 respectively. Equations (2) and (3) are an alternate way of specifying the bottomside thickness in IRI. Since IRI-90 this is the recommended (standard) profile option.

Mahajan et al. (1995) used data from the incoherent scatter station in Arecibo to validate formula (2) for low latitudes. Different from Gulyaeva they found only very small diurnal and

7

seasonal variations for G (see Fig.4). Their G values vary from about 0.85 during nighttime to about 0.80 during daytime whereas formula (2) varies from 0.8 and 0.6. Mahajan et al. (1995), however, before starting their analysis had sorted their data into two classes depending on whether h0.5 was found above hmFl (class A) or below (class B). The results noted above apply for class A (2/3 of cases). For class B the interference of the FI layer results in lower hO.5 heights and thus lower G-factors (see Fig.5).

In Fig.6 we compare the B0 values predicted by the two options in IRI for noon and mid­night. The solid curve depicts the Gulyaeva (1987) model and the broken curves the model based on Table 1 (the dotted curve is the pre-1993 version with an artificial step at the magnetic equator). Compared to the older model, Gulyaeva’s formula provides larger values during day­time and smaller values during nighttime; the discrepancies are largest at low and equatorial latitudes. The B0 deduced from Gulyaeva’s formula also shows abrupt steps indicating cases where the parameter B1 had to be changed considerably to get the upper and lower profile parts to merge. Fig.7 shows the effect the different B0 values have on the bottomside and F 1-region profile.

SHORTCOMINGS: The formula (2) is based on mid-latitude data and does not seem to de­scribe correctly the variation patterns at low latitudes. It does not differentiate the cases where h0.5 is above or below hmFl.

RECOMMENDATIONS: The Arecibo results of Mahajan et al (1995) should be incorporated into the formula. The mid-latitudes should be re-evaluated after first dividing the data into the two classes.

3. Fl-Region

Before adding an Fl-layer IRI first check if a number of criteria are fulfilled. If they are then a parabolic Fl-layer is added to formula (1) as follows

N(h)/NmF2 = f(x) + Cl*sqrt(hmF\-h)/B0) (4)Cl(MD) = 0.09 + 0.11/(1 +exp{-x}), x = (A/D- 30)/10

CD(MD) = CD{ 18) for A/D < 18

MD is the modified dip latitude in degrees. The FI density (NmFl) is determined from a global model and the FI height hmFl is found as the height where the bottomside profile function reaches the value NmFl.

3.1 NmFl and occurrence criteriaNmFl is obtained with the Ducharme et al. (1971,1973) model that was established using

a large amount of ionosonde measurements

NmFl/m — 3 = 1.24D10 * foFl/Mhz * JoFl/Mhz (5)

foFl = fs*cos~n{X}

8

fs = /O * (/100 - /O) * fizl2/100 /O = 4.35 + 0.0058 * |S| - 0.00012 * S * S

/100 = 5.348 + 0.011 * |S| -0.00023 * S* S

n = 0.093 + 0.0046 * |S| - 0.000054 * S * S + 0.0003 * Rz\2

where X is the solar zenith angle in degrees, Rzl2 is the 12-months-running mean solar sunspot number, and S is the geomagnetic latitude in degrees. Their model assumes that an Fl-layer is present if X is lower than a critical value Xs.

Xs = X0*(X100-X0)*Rzl2/100 (6)

X0 = 49.85 +0.35 *|S|X100 = 38.96 + 0.51 * |S|

In consultation with one of the authors (R. Eyfrig) the Ducharme model was modified for IRI to better represent later studies: (i) geomagnetic latitude was replaced by magnetic dip latitude; (ii) FI was omitted at night and during winter. In Fig.8a,b we have plotted the original model against its implementation in IRI. We note the typical camel-back signature of the equator anomaly. In general the differences are small except at high latitudes and also in terms of the limiting latitude for FI occurrence.

SHORTCOMINGS: Several studies have shown that the IRI criteria for FI occurrence are too limiting, especially the “winter” criteria.

RECOMMENDATION: Go back to original Ducharme et al. (1973) model.

3.2 hmFlNo experimental evidence is currently used for describing hmFl in IRI, rather the bottomside

profile function f(x) determines hmFl as the height where f(hmFl)=NmFl. Using the two IRI options for B0 thus results in different values of hmFl, and these differences can be quite large as seen in Fig.Ta. For the mid-latitude summer daytime case in Fig.7a the Gulyaeva-option provides an hmFl of 150 km whereas the Table-1-option gives a hmfl of close to 200 km. The Gulyaeva value is almost in agreement with the widely used simple formula

hmFl/km = 165 + 0.6428 * X (7)

which was used for the dotted profile and the diamond in Fig.7a.

SHORTCOMINGS: hmFl is obtained somewhat arbitrarily through the selection of B0; large discrepancies for different B0 options.

RECOMMENDATION: Using an independent hmFl model (maybe formula (7)) one could then use (hmFl/NmFl) as an additional anchor point in defining the bottomside profile shape; the main anchor point is the point 0.24*NmF2 at hmF2-B0.

9

3.3 FI layer shapeThe current implementation in IRI (formula (4)) is based on just a few ionosonde mea­

surements. The shape function and parameter should be tested and improved with more measurements. Alternate solutions, proposed during last year’s task force meeting, include the use of an additional anchor point (density at a fixed height of 170 or 180 km or both) or of the gradient at the FI point.

References

Bilitza, D., The International Reference Ionosphere 1990, National Space Science Data Center, NSSDC/WDC-A-R&S Report 90-22, Greenbelt, Maryland, November 1990.

Ducharme, E.D., L.E. Petrie, and R. Eyfrig, Radio Sci. 6, 369-378, 1971.

Ducharme, E.D., L.E. Petrie, and R. Eyfrig, Radio Sci. 8, 837-839, 1973.

Gulyaeva, T.L., Adv. Space Res. 7(6), 39-48, 1987.

Mahajan, K.K., R. Kohli, N.K. Sethi, and V.K. Pandey, Adv. Space Res. 15(2). 51, 1995.

Peres, M., Adv. Space Res. 7(6), 73-74, 1987.

Ramakrishnan, S. and K. Rawer, Space Research XII, pp 1253-1259, Akademie-Verlag, Berlin, 1972.

Reinisch, B.W., L.E. McNamara, T. Bullett, and R.R. Gam ache, Adv. Space Res. 11(10), 81-87, 1991.

Table 1. Bottomside Thickness Parameter B0.

B„/ km 1 I Spring Summer Fall WinterLT=12 LT=0 LT=12 LT=0 LT=12 LT=0 LT=12 LT=0

Modip R=10 114 64 134 77 128 66 75 73= 18 fi=100 113 115 150 116 138 123 94 132Modip ff= 10 72 84 83 89 75 85 57 76=45 R=100 102 100 120 110 107 103 76 86

Figure Captions:

Fig.l Buildup of IRI electron density profile.

Fig.2 IRI bottomside function for several B1 values.

Fig.3 Variation of G with solar zenith angel and season in formula (2).

10

Fig.4 Diurnal and seasonal variation of G as deduced from Arecibo data (Mahajan et al., 1995).

Fig.5 Diurnal variation of ratio (hmF2-h0.5)/hmF2 for Arecibo class B cases (Mahajan et al.,1995).

Fig.6a,b BO values for noon and midnight obtained with the Gulyaeva formula (2) (solid curve) and with the values of Table 1 (broken and dotted curves); the dotted curve shows the pre-1993 version.

Fig.7a,b IRI profiles below the F peak for day and night obtained by using the BO values from Table 1 (solid line and asteriks), by using the Gulyaeva formula (2) (broken curve and squares) and by using the analytical LAY representation (dotted line and diamonds). NOTE: The location is given incorrectly as “Boston”, the plots are actually for a geo­magnetic latitude of 30 degrees and a longitude of 0 degrees.

Fig.8a,b Latitudinal profiles of foFl as obtained with the Ducharme et al. (1973) model (solid curve) and with the IRI model (broken curve) for different months.

ZJ<U

M Z ( SO) <

11

lee *

Fig.l

NyNmF2»e*p(-* ')/cosh(*)

AMEC«0. l«y«-77

soior zenith angle

Fig.3Fig.4

ALL SEASONSM O.S ) < timF,

LT Zh

Fig. 5

12

LT=12, Rz = 50, April

modified dip latitude

Fig.6a

s90STCN (72W.AJN) MOOIP ..T =12 JULY Rz 1 2*50

St*nd«rd III: solid lin*

&. option: broken lino

LAI /wetldnd: dotted line

electron density / -n"

Fig.7a

IT = 0, Rz- 100, April

modified dip tolkuda

Fig.6b

.^BOSTON (72W,^3N) mOOIP = 55 7 =0 JULY Rz12 = 50

50 L....................... .i. -.1 . .......... . . .. ..l£-8 l£*9 1£+I0 l C-k 1 1 !£■*■ 12

electron denaitv / m"1

Fig.7b

i

Fig.Sa Fig.Sb

XA9743630

13

ELECTRON DENSITY PROFILES FOR MIDDLE LATITUDES Theoretical Model Results and its Comparisons with Observations for the Bottomside F Region

S.-R. Zhang1,2, S.M. Radicella1, X.-Y. Huang2 and M.-L. Zhang1 1 Atmospheric Physics and Radiopropagation Laboratory,

International Centre for Theoretical Physics, 34100 Trieste, Italy 2 Wuhan Ionospheric Observatory, Wuhan Institute of Physics,

The Chinese Academy of Sciences, 430071 Wuhan, China

Introduction

Variations of the FI traces (the “cusp”) on the ionogram were recognised early in the begin­ning years of the vertical sounding of the ionosphere, while the actual shape of the corresponding electron density profile are still an open problem nowadays, not only from an information ex­tracting point of view but also from its empirical descriptions as have been done within the IRI model. Results from theoretical models can be a useful complement to the experiment data base, considering the availability and reliability of the data, and are particularly essential for the physical understanding of the processes and the characteristics involved.

Some key points on the electron density profile, e.g., the half maximum electron density height, is used in IRI to give a certain representation of the profile shape. Physical discussions on the points are very important because this may provide us a clear idea on the features of our choice and help to judge if this choice is physically reasonable.

Model Highlights

The theoretical model for middle ionospheric profiles over mid-latitudes (Zhang et al, 1993; Zhang and Huang, 1995a,b), is used in this study. Basically, the model solves the continuity and momentum equations for 0+, N0+,0% and N}, as well as electron. Dynamic processes of diffusion and wind, and also photo-chemical processes (with involvement of meta-stable ions of atomic oxygen), are taken into account. The ion productions from photoelectron impact are also considered (Richards and Torr, 1988) The EUV91 model (Tobiska, 1993) is used to specify the solar irradiance, and neutral atmospheric density and temperature are given by the MSIS86 model (Hedin, 1987).

In order to gain a realistic electron density profile, the model adjusts the topside (500 km) density so that the simulated maximum electron density is fitted to the observed one. The model tries to let the height of the F2 peaks close to the observation. This can be done by using a more or less realistic vertical drifts of the ionization (Zhang and Huang 1995), derived from servo model, which are found to be more reasonable than the empirical model values like HWM90 (Hedin et al, 1991) by prior studies. Of course, some modifications on the Servo model coefficients have to be made as suggested by Titheridge (1995).

14

Results and Discussions

Model calculations were carried out for 2 October 1992 over Millstone Hill (42.6°N, 288.5°E). The daily Ap is 10, F107 is 118 and its 81-days-median 122. To have a reasonable reproduce of the Fl-layer shape, the neutral temperatures were increased by 15% from their MSIS86 values, while EUV flux increased by 25% from the EUV91 values over all bands and lines.

• Profile shapeFigures 1-3 give examples of the comparisons among the model profile, the observation, and the IRI profiles with the standard BO and the Gulyaeva BO options.

In general, the theoretical model is able to reproduce the observed profile shape. In some cases, the model underestimates the FI region density.

• Key points below the F2 peak * 1 2The height of the half maximum electron density /i0.s is a useful parameter in the de­scription of the single layer profile, since it approximately lies at the height of the dN/dh maximum (base-point), which can be considered as the base of the layer. It can be proved that, at the base point, 7V& = NmF2j2.03 and p = (hmF2 — hb) / {3.17 H) = 1 where H is the constant scale height.However, the ionosphere above the E-layer cannot be considered as a single layer, even if it is a single one (nighttime) it may not be a standard Chapman layer. For some cases, the ho.s may drop to the E region height. Fig.4 shows the hb and /iq.s variations from the observation and the model, where there exist cases of a single layer (nighttime) and two layers (daytime) in the F region. Differences of the two heights are evident when there is an intermediate layer. In a sense of the F layer morphology, this base point carries information from both F2-layer and Fl-ledge. And the simulated ratio p doesn’t remain at 1 but about 0.3-0.4 during the day and 0.8 during the night, as shown in Fig.5. The F-layer thickness defined by hmF2 — hh is about 40 km by day and almost doubled by night.

• Physical meanings of the hbFig.6 shows the dN/dh profiles obtained with the theoretical model, with the hb and ho.5

given also. The occurrence of the base points is quite evident. In fact, as shown in the 0+ continuity equation, dN/dt = q — (3N — vdN/dh — Ndv/dh, at the height of the base points, changes of the electron density is most sensitive to the transport velocity.

Conclusions

1. The theoretical model is capable to reproduce important observed features in the Fl- layer, and thus can be used for further studies in this area, not only from the physical understanding viewpoint but also from the practical uses in the empirical modelling.

2. The ho.5 parameter has obvious physical meanings(1) the change of Ne at the height is most sensitive to the plasma transport;(2) the height is sensitive to the ionization not only in F2 region but also in the FI region, i.e., it reflects the occurrence of the Fl-ledge. Thus it implies an important use in determining the F2-layer shape as well as the Fl-layer shape.

15

Acknowledgements: We are grateful to Prof. W. B. Reinisch for providing the digisonde data to the APRL of ICTP, and to NSSDCA for providing us the IRI90, MSIS86 and EUV91 models.

References

Hedin, A. E., MSIS-86 thermospheric model, J. Geophys Res., 92, 4649-4662, 1987.

Hedin, A. E., M. A. Biondi, R. G. Burnside, G. Hernandez, R. M. Johnson, T. L. Kileen, C. Mazaudier, J. W. Meriwether, J. E. Salah, R. J. Sica, R. W. Smith, N. W. Spencer, V. B. Wickwar and T. S. Virdi, Revised globe model of thermospheric winds using satellite and ground-based observations, J. Geophys. Res., 96, 7657-7688, 1991.

Titheridge, J. E., The calculation of neutral winds from ionospheric data, J. Atmos. Terr Phys., 57, 1015-1036, 1995.

Tobiska, K., Revised solar extreme ultraviolet flux model, J. Atmos. Terr Phys., 55, 1637- 1659, 1993.

Zhang S.-R. and X.-Y. Huang, A Numerical Study of Ionospheric Profiles for Mid-Latitudes, Ann Geophysicae, 13, 551-557, 1995a.

Zhang S.-R. and X.-Y. Huang, An Ionospheric Numerical Model and Some Results for the Electron Density Structure below the F2 Peak, Adv. Space Res., 16, (16)119-( 16)120, 1995b.

Zhang S.-R., X.-Y. Huang, Y.-Z. Su and S. M. Radicella, A Physical Model for One-dimensionand Time-dependent Ionosphere----Part I: Description of the Model, Annali di Geofisica,36(5-6), 105-110, 1993.

Richards P.G. and D.G. Torr, Ratios of photoelectron to EUV ionization rates for aeronomic studies, J. Geophys. Res. 93, 4060-4066, 1988.

16

Figure Captions:

Fig.l Comparisons of the simulated, the observed and the IRI Profiles for 01LT.

Fig.2 Comparisons of the simulated, the observed and the IRI Profiles for 12LT.

Fig.3 Comparisons of the simulated, the observed and the IRI Profiles for 14LT.

Fig.4 Diurnal variations of the base point height (Hbs) and the half maximum density height (Hp5) obtained from the profiles of the theoretical model and the observation.

Fig.5 Diurnal variation of the ratio p = (hmF2 — /i(,)/(3.17H) obtained with the theoretical model.

Fig.6 Height profiles of the dN/dh values obtained with the theoretical model. Note the corresponding heights of base point and the half maximum density heights.

h (km

) h (

km)

17

Millstone Hill, 2 October 1992 (01LT)

simulated observed

IRI options

fN (MHz)

Fig.l

Millstone Hill, 2 October 1992 (12LT)

190 •

170 -simulated — observed

IRI options - -150 -

fN (MHz)

Fig.2

h (km

) h (

km)

18

Millstone Hill, 2 October 1992 (14LT)

simulated — observed -*—

IRI options - -150 ■

fN (MHz)

Fig.3

Millstone Hill, 2 October 1992

Hbs:observed — Hp5:observed - - Hbsrsimulated « Hp5:simulated +

♦ + •

. - +

Fig.4

h (km

)

19

Millstone Hill, 2 October 1995

LT (h)

Fig.5

Millstone Hill, 2 October 1992

22LT

O.OeO 2.0e6 4.0e6 6.0e6 8.0e6 1.0e7 1.2e7 1.4e7dN/dh (el/mA4)

Fig.6

XA9743631

20

CONSTRAINTS ON THE BOTTOMSIDE ELECTRON DENSITY PROFILE SHAPE PARAMETERS

T.L. GulyaevaIZMIRAN, 142092 Troitsk, Moscow Region, Russia

The IRI-1990 system includes a new option for calculating the bottomside profile shape (Bilitza, 1990) based on strong dependence of the F2 layer half-density height h0.5 on the F2 layer peak height hmF2. Recently it was pointed out that the formulae should be distinct for A class cases when h0.5 belong to the F2 layer and B class when h0.5 occurs below hmFl (Mahajan et al., 1995). Class B means 0.5*NmF2 is less than NmFl. In such cases an influence of the FI layer significantly distorts the profile shape. Fig.l (Bilitza and Rawer, 1990) shows a sketch of A and B cases for daytime electron density - height profile. Constraint for the plasma frequency fr < 0.7*foF2 corresponds to the F2 layer half-density point of 0.5*NmF2.

Diagrams of occurrence of A and B cases at 5 latitudes of North Eastern Hemisphere (Table 1) show that class B is not observed during equinoxes while it is prevailing at high latitudes for low and high solar activity. It has been shown elsewhere that the B class cases neither occur at geographic equator nor at geomagnetic equator. Appearance of the B cases could be detected both with observed data and relevant ionospheric models. Examples of A and B cases are found differrent with IRI (Bilitza, 1990) and Russian regional model (Dvinskikh and Naidenova, 1989) for the same conditions. Hour-to-hour variability of the ionosphere show even more alternates of A and B cases, in particular, at low solar activity.

In this context correct definitions of the F2 and FI layers peak and bottomside profile shape parameters become of great importance. At these heights the electron production rate reach maximum. The empirical constraints could be posed on the said parameters to be used for validation of their consistency. Fig.2 shows frequencies (left) and heights (right) at Ahmedabad (lat=23 N, long=73 E) for June 1993, low solar activity (sunspot number Rz=56). Circles show CCIR prediction of foF2 (left) and hmF2 and hmFl (right) used in IRI (Bilitza, 1990). Solid lines - observed critical frequencies foF2, foFl and foE (left), and relevant peak and half-density heights (right) calculated from M(3000)F2 with IRI option (Bilitza, 1990). Observed hp values are given by points (top right).

At low latitudes class B occurs during the first half of day. It is observed at the bottom of vertical bars connected to foF2 showing half-density portion of profile above 0.7*foF2 which are less or equal to foFl (solid line) at 08 and 09 hr LT. Similar bars below hmF2 (right) show half-density heights of profile after correction of bottomside profile shape parameter h0.5 for low latitudes (Gulyaeva et al., 1995). It can be compared with earlier option for h0.5 in IRI (Bilitza, 1990) (circles above hmFl) which are typical for A cases.

While constraint 0.5*NmF2 compared with NmFl defines A or B class occurrence, we need the F2 and FI layers peak heights to be properly determined. If then the shape parameter h0.5 is compared with hmFl one can check the consistency of all the above parameters. This is illustrated in Fig.2 (right) where two other models for hmFl are presented. The hmFl as a function of the solar zenith angle is shown by crosses decreasing towards the mid-day

21

(Bilitza and Rawer, 1990). These values are related by linear segments with another hmFl model (Mosert de Gonzales and S. Radicella, 1987) defined from the h’F2 observations. The latter values show maximum hmFl at noon while the IRI FI peak heights are growing towards the afternoon hours. When newly produced h0.5 altitudes are compared with different hmFl approximations one cannot draw the definite conclusion which case A or B is occurring. So we are facing the problem of modelling the FI layer peak height independent of the F2 layer heights and shape parameters.

References

Bilitza, D. (ed.), NSSDC/WDC-A-RPS90-22, URSI/COSPAR (1990).

Bilitza, D and Rawer, K., Adv. Space Res.10, 11, 7-16 (1990).

Dvinskikh, N.I. and Naidenova, N.Ya., Regional empirical models of the ionosphere. Preprint 1-89, SibIZMIR, Irkutsk, 1989.

Gulyaeva, T.L., Maliajan K.K. and Sethi, N.K., Modification of IRI half-density option for low latitudes, Adv. Space Res., in press (1995).

Mahajan, K.K., Kohli, R., Sethi N.K. and Pandey, V.K., Adv. Space Res. 15, 2, 51-60 (1995).

M. Mosert de Gonzales, M.and Radicella. S., Adv. Space Res. 7, 6, 65-68 (1987).

Figure Captions:

Fig.l Specification of class A and B cases using sketch of IRI electron density profile.

Fig.2 Specification of class A and B cases using routine ionosonde observations of the FI and F2 layers critical frequencies and heights at Ahmedabad for June 1993.

22

TABLE 1. Diagrams of Occurrence of A or B Class Cases for Low and High Solar Activity at Latitudes 40-80 N, Longitude 90 E; March (Equinox) and June (Summer).

A , B - IRI model; a , b - Regional h’f model.

A , a : foFl < 0.7*foF2 B , b : foFl > 0.7*-foF2

Rz=50 MARCH JUNE

LT:08 10 12

i*-* i 4* 1

16 LT:1 8

102 04 06 08 10 12 14 16 18 20 22 :

BON 80 B B B B B B B B B B B B B B B B B B B B B B B b:80 a a a 80 a a b b b b b b b b b a70 A A A A A A 70 B B B B B B B B B B B B B B B70 a a a 70 a b b b a a a60 A A A A A A A 60 B B B B B B B B B B B B A60 a a a 60 a b b b a b a50 A A A A A A A 50 B B B B B B B B B B B50 a a a 50 a b b b b b a40 A A A A A A A 40 A A A A A A A A A40 a a a 40 a b b b a b a

o7?c

MARCH JUNE

LT:08 10 12 14 16: LT: 00 02 04 06 08 10 12 14 16 18 20 22 :

BON A A BON A A B B B B B B B B B B B B B B B B B B B B B b:80 a a a 80 a a a a b b a a b b a a70 A A A A A A 70 B B B B B B B B B B B B B A A70 a a a 70 a a a a a a a60 A A A A A A A 60 A A A B B B B B B B B A A60 a a a 60 a b b a a a a50 A A A A A A 50 A A A A A A A A A A A50 a a a 50 a b b b a a a40 A A A A A 40 A A A A A A A A A40 a a a 40 a b b b b a a

Topside M)

hmF2

F 2 (21

hmFlFI (31 -

— — Interned. (61

---- E-Valley (5 1 —HABR

E/DI6)

NmF2NmFl

log N

Fig. 1. Buildup of IRI electron density profile.

24

■-s e15 LI

Fig. 2

XA9743632

25

INVESTIGATION ON EQUATORIAL IONOSPHERIC PROFILESAND IRI MODEL

J O. AdeniyiPhysics Department University of florin, P.M.B. 1515, florin Nigeria.

Abstract: Ionospheric profiles below the F2 peak ionisation density are compared with those of the International Reference Ionosphere (IRI). The data used are those of Ibadan (Lat. 7.4° N, Long. 3.9° E). The IRI model gives a much thinner bottomside F region ionisation density than what is observed experimentally, in winter; both at high and low solar activity. Similar departures are observed in the summer of both solar epoch but on a reduced scale. The closet agreement occurs during the march equinox of high solar activity.

1. Introduction and Data Analysis

The aim of the study is to investigate how the profiles generated by the International Ref­erence Ionosphere (IRI) model agree with observed profiles, of an equatorial station. Samples of good ionograms were selected for the period of high and low solar activity; R12 = 10 and R12 = 110 respectively. The data are those of Ibadan (Lat. 7.4° N, Long. 3.9° E, Dip 6.3° S). The months of January, April, July and October were used to represent the Winter (December Solstice), March equinox, Summer (June Solstice) and September equinox respectively. Iono- gram inversion by the POLAN method (Titheridge, 1985), was used to obtain the experimental values. Both the new and old options of bottomside thickness parameter (Bilitza, 1990), in the IRI model were used in generating the model profiles in order to confirm the better option for the equatorial station. Experimental values of the critical frequency of the F2 layer and the corresponding height (F2 peak density height) were use to normalise the IRI model in each case for the purpose of a fair comparison.

2. Results

Generally, the new option (Gulyaeva, 1987) gives profiles closer to experimental observa­tions for this station than the old option. Some discrepancies are however still present when the new option is used.

(a) WinterFig.la shows an example of profile for the winter season of low solar activity. IRI does not

predict the presence of the FI layer in Winter, but experimental data show a well formed FI layer most of the time at low solar activity. The IRI model gives a thinner bottomside thickness of the F region than that observed experimentally. This is true for both the old and new options for the bottomside representation, although the new option gives a thicker bottomside than the old. For the winter of high solar activity, a well defined FI layer is not observed. This agrees with IRI prediction. Fig.lb shows an example for the high solar activity. The IRI models still give thinner bottomside thickness of the F region.

26

(b) SummerIn agreement with IRI prediction, FI layer is present during the summer of high and low

solar activity. Fig.2a show an example for the summer of low solar activity. A thinner bot- tomside ionisation density is given by the IRI model. The deficiency is however lower for the new option of bottomside parameter representation. The result for the summer of high solar activity is quite similar to that of the low solar activity (Fig.2b).

(c) March equinoxA well defined FI layer is present at low solar activity. This agrees with the IRI prediction.

At high solar activity, the FI layer is not observed during this season, although IRI predicts its presence. At low solar activity, the IRI gives a lower ionisation density, particularly around the FI region (Fig.3a). The model is closer to the observed profile at high solar activity than at low solar activity (Fig.3b).

(d) September equinoxThe FI layer is present during this season at low solar activity but absent at high solar

activity. Like for the March equinox, IRI predict its presence during this season for both solar epoch. The result for low solar activity during this season is similar to that of Winter. Generally IRI gives a thinner bottomside ionisation than the experimental profile (Fig.4a). At high solar activity, the agreement between the model and observation is good except around the FI region where the model gives a lower electron density (Fig.4b).

3. Conclusion

The departure of the IRI model profiles from the experimental profiles occurs below the F2 peak; especially around the region of the FI. With the exception of Summer season, the departure is greater at low solar activity than at high solar activity. The best agreement between the model and experimental observation, occurs during the March equinox of high solar activity. A better definition of the bottomside thickness parameter of the F region is needed in order to improve the IRI model. A study of the FI region characteristics, taking into account seasonal and solar cycle effects will assist in establishing helpful parameters for improving the model.

References

Bilitza, D. , International Reference Ionosphere URSI/COSPAR 51-64 1990.

Gulyaeva, T.L., Progress in ionospheric informatics based on electron density profile analysis of ionograms, Adv. Space Res. 7(6)37-48, 1987.

Titheridge, J.E., Ionogram analysis with the generalised program POLAN, Report UAG-93 World Data Centre A for STP, NOAA, Boulder, USA, 1985.

27

Figure Captions:

Fig.l Typical profile for the Winter season at (a) low solar activity (b) high solar activity.

Fig.2 Typical profile for the Summer season at (a) low solar activity (b) high solar activity.

Fig.3 Typical profile for the March equinox season at (a) low solar activity (b) high solar activity.

Fig.4 Typical profile for the September equinox season at (a) low solar activity (b) high solar activity.

He l

g h t

km H

e i g

h t600 -]

500-

E 400

300-

2200 -

* OBS+ — — IRI NEW o o-o o-o IRI OLD

(190164)

( 1 200 hr LT)

4.0 6.0s l ty 1 0* *(10) Zm* * ( 3)El e c t n o n De

500 m450 -

300 E250 E200 E OBS

+ iri OLD o o-o o-o IRI NEW

(090160)(1200 hr LT)

1 50 -

. 0 1 0.0 220. (2)

El ec Iron De ns l L y 1 0* *(10) Zm* * ( 3)Fig. 1

Hei

ght km

H

e Ig h

t29

550-

450 -

.....OBS+ 4-+ +-+ IRI NEW O G-O G-0 IRI OLD

(130764)

(1300 hr

1 50 -

Electron Density 10**( 10) /m* * ( 3)

450 -

250 -

.....OBS+ — — IRI NEWO G-O G-G IRI OLD

(250760)( 1 200 In r LT)

Electron Density 10**( 10) /m* *( 3)Fig.2

He l g h t k

m He l g h t

550 -

350 -

—+OBS+ +-f IRI NEW o o-o o-o IRI OLD

( 1 60464)( 1 200 h r LT)

1 50 —

. 0 2.0 4.0 6.0El e c Iron De ns 1 t y 1 0* *(10) Zm* * ( 3)

450 -

250-*-» OBS + IRI NEW

O 0-0 O-G IRI OLD (190460)( 1 200 h r LT)

1 50 —

20.01 5.01 0.0El e c Iron De ns L t y 1 0* * ( 1 0) Zm* * ( 3)

Fig.3

He Ig h t k

m He Ig h t 350-

* * * * * OBS+ +-+ IRI NEWo o-o o-o IRI OLD

( 231 064)(1200 hr

1 50 -

. 0 2.0 4.0 6. 0 8.0

El ecIron Dens Ity 10* * ( 10) Zm* *( 3)

— — OBS + +~+ IRI NEW O O-Q O-G IRI OLD

( 1 -41 060)

( 1 200 h r1 50 -

1 0.0El ec Iron De ns Ity 1 0* *(10) Zm* * ( 3)

Fig.4

11111111111XA9743633

32

COMPARISON OF THE OBSERVED RESULTS OF THE ELECTRON DENSITY PROFILES WITH THE IRI90

Man-Lian Zhang1, Sandro M. Radicella1 and Kai-Liang Dai2 1 International Center for Theoretical Physics, Trieste, Italy.

2 China Research Institute of Radiowave Propagation, XinXiang, China.

Abstract: The daily and composite profiles and the thickness parameter of the electron density profiles are compared with the results of IRI90 for the stations of Ramey (15.8°N, 292.9°E) , Wuchang (30.6°N, 114.3°E), Chongqing(29.5°N, 106.4°E) and Wrumchi(43.8°N, 87.6°E). It is found that the electron density profiles produced by IRI90, both with the old Standard BO and the new Gulyaeva-BO thickness parameter, are too thick below F2-peak compared with the observed results. It is also shown that the IRI90 results show a very poor agreement with the observed results for the intermediate (FI) layer of the ionosphere.

Data Used

The data used for the present comparison are (1) the daily electron density profiles inverted from the digisonde ionograms by the ARTIST method at noon time for the Ramey Station. The seasons chosen are October of 1992 and January, March and June of 1993. For each month 10 or more than 10 days’ data are selected for the comparison. (2) The composite ionogram data of the typical months of March, June, September and December for the year 1968 for the three Chinese stations: Wuchang, Chongqing, Wrunichi.

Analysis and Results

Figs.l and 2 are the electron density profiles obtained by the inversion of the digisonde ionograms using the ARTIST method for the Ramey station for the seasons of January, March, June and October of low solar activity years 1992-1993. Also shown are the results calculated by the International Reference Ionosphere (IRI90). For the IRI90, both the profiles produced bv using the standard BO and the new Gulyaeva-BO parameters are used to make the comparison. It can be seen that, except for the cases for January where the IRI90 results show reasonable good agreement with the observational results, the electron density profiles below the F2-peak calculated by IRI90 are too thick compared with observed ones. This is true especially for the cases when the Gulyaeva-BO thickness parameter are chosen in IRI90. As for the results for the intermediate (FI) layer, The IRI90 results show very poor agreement with the inverted results from the digisonde ionograms.

Fig.3 is the result for the Chinese Wuchang station, where the composite ionogram data are used when inverted into the real-height profiles. The same conclusion is obtained, i.e., the electron density profiles calculated by IRI90 are too fat compared with the observed ones for the daytime.

In Fig.4, we made the comparison between the observed and the IRI90 results in another way using the composite data from the Chinese stations, where the half-density height h0.5

33

parameter is used to compare. The half-density height h0.5 is the height at which the electron density reaches half of the F2-peak density NmF2:

N(h0.5) = 0.5NmF2.

The half density height h0.5 can be considered as a parameter describing the thickness of the electron density profile. In fact, in IRI90, when choosing the Gulyaeva-BO parameter to calculate the profile, the BO is related to h0.5 by a formula (Bilitza, 1990).

It can be seen from Fig.4 that, for the daytime of the seasons of March, June and October, the hO. 5 calculated from IRI90 is always lower than that obtained from the profiles inverted from the composite ionograms for all the three Chinese stations, while for the nighttime or the daytime of the winter season (December), the agreement between IRI90 results and the observed ones is reasonably good. These results are consistent with the conclusion obtained above.

Conclusions

From the present study, the following conclusions are obtained:

(1) The electron density profiles produced by IRI90, both using the standard BO and the newGulyaeva-BO thickness parameter, are too thick compared with the observational results under the F2-peak for the equinoctial (March and October) and the summer (June) seasons for the daytime. This conclusion is especially true for the IRI90 when using new Gulyaeva-BO thickness parameter. For the winter season, in general, the profiles produced by IRI90, both using the standard BO and the new Gulyaeva-BO are reasonable compared with the ARTIST results.

(2) Compared with the observational results, the IRI90, both with the standard BO and thenew Gulyaeva-BO, show a poor agreement with the observed intermediate (FI) layer for all the four seasons.

Acknowledgments: The authors would like to thank Prof. B. Reinisch for providing the digisonde ionogram data of the Ramey station for the present study.

Reference

Bilitza, D., International Reference Ionosphere 1990, National Space Science Data Center, NSSDC/WDC-A-R&S 90-20, Greenbelt, Maryland, 1990.

34

Figure Captions:

Fig.l Electron density profiles obtained by inversion of the digisonde data and by IRI90.

Fig.2 Electron density profiles obtained by inversion of the digisonde data and by IRI90.

Fig.3 Electron density profiles obtained by inversion of the composite ionograms and by IRI90.

Fig.4 Half density height h0.5 obtained from the inverted composite real-height profiles and from IRI90.

ARTIST

U Ov**N) lOv"-'1'

turn (Oi<j boi ♦

IRIWiNew BOi 6 ;

i

..(k'Hjj K.ue^is i.Oe+06 I.2C+I*. 1.4«+IX1

Net h)

:(>e+05 4.UC+05 O.Oe+05 8()e+05 I.Uc+<X> l.2c+<>6 I 4c+0(,

Net 111

Fig. 1

V

AR J IS 1

1RI‘Mi iQltl BUI ♦

IRIWU iNv.v MO- a

;iih.--nK) 5.t)c+us I Dc-MXi l.5e+4K) 2.0c+(W' ?

Ramey IW3 Day=82 ILT=I2:00)

I5H —

250 !•

ARTIST

IRI90 i Old BO)

IRI90(New BO) a

0 Oc+OO 5 Oc+05 I Oe+Oh I 5e+<K> 2 Oe+OA 2.5e+(X>

Nei h i

i... i/m: r'u

/ ■

ARTIS I

IRl“OlUKI HO) ♦

IR!90(Nc.v HOi e

‘"vom :.Oc4()‘! -l.Oc+os Oc+05 8.0c+05 |.0e+06 I >+0(>

Nei h i

Ramey IW3 Day=l69 (LT=|2:00)

ARTIST

IRI9<)(0ld B0>

IRIVOiNew BO)

OOc+nii :.0c+i>5 4 0C-HI5 6.0c+05 S.0c+<)5 I.0c+o6 I :c+06

Ncth)

I 4e+o<>

Fig.2

1 • i i ^•

0

,0(1 $

MiHk'+OO S.IKk'+(lA

XRIIST —

IRI'-" ■< M«l Vu>. *

IRI'X) i V» HU. a

l*>e+o<i !..'Ov+i ih imv+oh ;.'oc*oo

Neon

R.miey i 992 Duy=283 (LT=l2:Um

ARTIST

IRI90 (Old BO)

IRl<X) (New BO) s

0 00c0*1 V00V+O5 I .(K)l-+06 I 50c+06 Z.OOeXKi 2.5<k-+00

NcOi)

%

37

Wuchang (March 1968 LT=11:00) Wuchang (March 1968 LT=12:00>

£ 200-

Nelli) {1.0E+5 cm-3)

Z 200

- 150

Nelli) ll.OE+5 cm-3)

------IRI-90(LI0-Cul) -------Composite ionograrn 1R1-90(D0-Gul) ------Composite ionograrn

Wuchang (March 1968 LT=13:00)

- 150

Ne(h) (1.0E+5 cm-3)

------IRI-90(B0-Gul) ------- Com|xisite ionograrn

Wuchang (March 1968 LT=14:00) Wuchang (March 1968 LT=15:00)

.2 zoo-- 150

Ne(h) (1.0E+5 cm-31 Ne(h) (1.0E+5 cm-3)

------1RI-90(B0-Gul) ------Composite ionograrn ------1RI-90(B0-Gul) ------Composite ionograrn

Fig.3

hO.5

(km)

h0.5

(km)

Chongqing (1968)

Mar. June Sept. Dec.

—'— 11(1-00(00-Gul) —•— From Real-profile

Urumchi (1968)

IRI-OO(OO-Gul) From Real-profile Fig.4

hO.5

(km)

hO.5

(km)

Wuchang(1968)

Mar. Jun Sep. Dec. Mar. JunSep. Dec Mar. Jun Sep. Dec.

—'— IRI-OO(OO-Gul) —'— From Real-profile

Wuchang (1968 March)

280- ,

—r l l i i i—i—i----1—i----- 1—i—l—l l i—i----- 1---- 1—l---- 1—i----- r00:00 0200 0400 0600 0800 1000 1200 1400 1600 1800 2000 2200

IRHO(BO-Gul)

IIIIIIIIIIIIXA9743634

39

THE SHAPE OF THE BOTTOMSIDE N(h) PROFILE AT TWO MIDDLE LATITUDE STATIONS

Marta Mosert de Gonzalez CASLEO, CONICET, San Juan, Argentina

and

Bruno ZolesiIstituto Nazionali di Geofisica, Rome, Italy.

Abstract: Noon bottomside N(h) profiles derived from ionograms of two middle latitude stations: ROME and USHUAIA are compared with those given by the two IRI-90 model options for the bottomside thickness parameter Bo. The study shows that there is no clear indication for an univocal choice of Bo and that this parameter is also important for the correct evaluation of the intermediate region ionization.

Introduction

The International Reference Ionosphere 1990 , IRI-90 model (Bilitza, 1990) offers two choices for the bottomside thickness parameter Bo: (1) a table of values based on reduced ionosonde profiles as in previous editions of IRI(OLD) and (2) Gulyaeva’s model (Gulyaeva,1987) for the half-density height h0.5 (NEW). The purpose of the present paper is to compare the profiles given by these two options with observed profiles at two middle latitude stations.

Data Base

Noon bottomside profiles derived from ionograms of ROME (41.9N; 12.5E) and USHUAIA (54.8S; 291.7E) were used for the analysis. The data were recorded during the representative months of summer, winter and equinox and for years of different level of solar activity (1965, R=16; 1986, R=14; 1990, R=145 and 1991, R=144).

Profiles were obtained by the inversion of ionograms using the POLAN computer code (Titheridge, 1985) and the IRI-90 model profiles were normalised with the experimental values of foF2 and hmF2.

Results of the Comparison

Examples of the comparison between observed profiles (OBS) and those given by the two IRI-90 model options for the parameter Bo: OLD and NEW, are illustrated in figures 1 to 5 for different seasons and solar activity conditions.The analysis indicates:(1) For winter and low solar activity, LSA (Fig. 1), a good agreement between experimental values and the two options is observed in the F2 layer for both stations, but the intermediate

40

region is not well reproduced by either of the two options. During high solar activity, HSA (Fig. 2), the results are contradictory: for Ushuaia the NEW option reproduces better the upper F region although overestimating the thickness, while for Rome the OLD option is the best one.

(2) For equinox during LSA and HSA (Figs. 3 and 4) the NEW option shows a better agree­ment between observations and the model predicted values, at least in the upper F region. In general, in the intermediate region large discrepancies are found.

(3) For summer (Fig. 5) under different solar activity conditions the OLD option is closer the experimental values for the station Ushuaia although overestimating the thickness of the F2 layer during HSA. In the intermediate region the agreement is not good.

Conclusions

The present test study shows:

(1) There is not a clear indication for an univocal choice of the bottomside thickness parameter Bo.

(2) The choice of Bo is important for a correct evaluation of the intermediate region ionization.

Acknowledgments: The authors undertook this work with the support of the “ICTP Pro­gramme for Training and Research in Italian Laboratories", Trieste, Italy.

References

Bilitza, D., International Reference Ionosphere 1990, NSSDC/WDC-A-R&S 90-22, World Data Center A for Solar Terrestrial Physics, CO, USA, 1990.

Gulyaeva, T.L., Progress in Ionospheric Informatics based on Electron Density Profiles Anal­ysis of Ionograms, Adv. Space Res., Vol.7, No.6, 1987.

Titheridge, J.E., lonogram Analysis with the Generalized Program POLAN, Rep.UAG-93, WDC-A-for Solar Terrestrial Physics, Boulder, CO, USA, 1985.

41

21,01 -8:;L:': IZ.U0

BBO. Or'

j

290. m ■

; OT. 99f/

H30, oe-i

rV

BO. 00*0 • TTTTriV: s

0, fee 1 rrx-x <-vvrr-,--p rrvvxvvrj'rv r “"v trrnr, rn-. rr-T'vvv'rv. T-v, % 2 W 5,00 M,0V

h L SB" .)

200 . Otp3

250. 00 j i:{

200,A

USHUMA B5/W I..T, : 12,00

«

-iItO.OO-i

<00, 00 '! ....■'r\

60, m-fr-v 6, ►>!-

r .1 i y. 1

OB:.,i-i["i.. r.. ..u...-:, » 1 ’ ! I

TT >"T ! -r ,-TT r : 'nr, TTT r r :-;-- , : , r H"''

T.vO r. )N ■ l;}$i W! ;, | ,;, |-,i |

NINTI P.. Luo Vu: AP A<. '. VI ! ,

vT|4 vC

i

42

‘m ROME: 0B, 01 , c'0d L T 12.. 00

:Ak> Pfc'-ri

UX.' kX>-;

26".0":

-:w, 0" j

:i Y yV-*'i f y, 00d

/100.00^

. OG j'rr r-T .T-p-T T -r~j-r-T-r-T-T-T...,-r-T j-T-Y r-r- -f-.y- T"r"r'!'T~"i'"r-v -v v-r"\r'i

0.00 5.00 10.00 15.00 20.00

;« so-, USHUA1A 3U/07/9IL,T: 12,00

si?, m !

A

:00.PX?:;]

I t*.'. b’lv -j /*'1 J-i <4.. ■4 #

•• rf- i‘ 4 —* ; /

■■■•'* ...... * NLUAil D

v40. " hi ", t' r-rri" V T -;—V v •,• r y r~r >• j- v-ry t rC,% 6. W lC.%

! '/;0( "6 !>,„■■• J]' "'I 9 !' ! : !>.•': ' l-! r; 1' 6 H2GR AALAR *' '• i V.!' f 1

43

490. 903 ROMP 1 5 -A9. -A6 3 LT. 12.00

3t2e.ee-!

ZW, 09 J

100. m

0. 00-4-1 • V 0 to '\.to 2. to 3. to 4.00

hi [ k m

300. 00-q

|250. 00 j

I209. tozj

s1901. OO-i

ii to, 00-g

3

USHUAIA 25/04/65L.T. : 12.00

J.< >

■ to-

iY

Q(3«?I-* ('.151.1

-J|-. *. .vt OLDJ, 00 4n60. i--ig -h iiiTm r-iTrnTrnTr(TTrrrri rnTTTTiTXTiiTnTT-nrrri rTnTrn')

0,00 1,00 2,00 3.00 4.00 6.00 6.00M " I 0R H ER -jrn- ■ ? ]

no, 3 .[ equinox,low solar act ivm j

44

h'O'ME 25/03/90 ^ L . T . : ! Z, 00

4W, (X'i1

3300. 00-11

200.004 /,*' ^1 //'./

V

100.00^

0. 0£> -j-T'rrrrr.-rrj-i-rrTTi v i-T'pnr rr*r-in- iT-)"rvit'ITTTTT r|'TT1" !TrrTT“:

0i i® 5,00 10^00 15:00 20.00 25:00

h [ K rn ]

uLI; 12.00

400 00 3 USHUAJA ©6/10/91

360, 00-3

1300, 00-j

32m. mi

J

l

150. 00-

100, 00

-V ,rr- ”1

:

I ^ (i£-}5 %NEW-» OLD

60, 00 Tn'i-mTH"f i tt1'-rrmrrrr rmr 'fTrfTTrrrrTrn0.00 6.00 10.0© 16.00 20.00

H/10HHB 3]

FIG, 4 . [ EQUINOX. HIGH SOLAR ACTJvm .1

'.mu.i.ov yvios KviH gnv i-io i '/Buwns i 1 s 'ou[>: '"JO ? ;

•K'i • > tit! ‘J W • t- ®'£ %' • :: titi ‘ l titi 'ti1.UJU.U.LU..L yfi ' ;+fj

f j |( : '-y -fc —fc- -h •-*

i'1 'IN - "h h_•••/}• Jj j -p-«

A.

J/

Em ml

^00

£-00 'fitit:

00 3 L : Tl

f:k*j 'PQrk

39/10/31 VlVnHSn t 00 mi­

ll IJJ:i ] m

0.^1 1J11(1OT........... 00 O

f:0

A****-"' h&Q '001

,y*-/V-V1

03 '03t:

Em '©i

C0 '31 : 11 uii v IViT k;n 0(?'03k!

IIUIIIIHIIIIIXA9743635

46

THE PRESENCE OF THE FI LAYER OVER A LOW LATITUDE STATION

Marta Mosert de Gonzalez1, Rodolfo Ezquer2 and R. del V. Oviedo 1 CASLEO, CONICET, San Juan, Argentina

2 UNT, CONICET, Tucuman, Argentina 3 LIIF, UNT, Tucuman, Argentina.

Abstract: Hourly median values of the ionospheric parameter foFl observed at a low latitude station , TUCUMAN (26.9 S; 294.6 E) have been compared with those given by the IRI-90 model for years of different solar acitivity. It is found that, in general, the agreement between the observed and predicted values of foFl is good when IRI predicts a value for it. Discrepancies are found in the occurrence of the FI layer, in particular, in winter during low solar activity.

Introduction

IRI-90 model uses the expressions introduced by DuCharme et al. (1973) to derive the monthly median values of the critical frequency of the FI region, foFl. These expressions predict this parameter as a function of solar activity, the magnetic dip latitude and the solar zenith angle. The model also provides the maximum solar zenith angle for which foFl is observed: the parameter is only assumed when the solar zenith angle is smaller than this limit angle. IRI-90 model does not predict the FI layer at night and in winter.

In this paper experimental median values of the parameter foFl obtained at TUCUMAN for years of different solar activity and seasonal conditions are used to check the validity of the IRI-90 model to predict foFl and the occurrence of the Fl layer.

Data Used and Analysis

Hourly monthly values of foFl measured at TUCUMAN (26.9 S; 294.6 E) are compared with those given by the IRI-90 model (Bilitza, 1990), for different seasonal conditions and for years of low (1965, 1966, 1973, 1974), middle (1971, 1972) and high (1968, 1969, 1970) solar activity. The “median count”, the number of numerical values that are used to determine the median, that is the most important indication of the reliability of the median, has been taken into account in the analysis. Values of foFl replaced by “L” are omitted from the median count. The qualifying letter “U” has been used to indicate that the median count is less than 5.

Figs.l to 4 show some examples of the comparisons for different seasons of the years: 1973(R=38), 1966(R=47), 1971(R=67) and 1968(R=107) respectively. Asterisks are plotted for the experimental values of foFl and points for the IRI-90 model predictions. The “median count” is also indicated in each value and the qualifying letter “U” when it corresponds.

The results of the comparison show:

(i) During Winter IRI-90 model does not predict values for the parameter foFl. However,

47

experimental values are often observed during daytime when solar activity is low. The occur­rence of the FI layer becomes less frequent when solar activity increases and, for high solar activity, experimental values are rarely observed in agreement with the model predictions.

(ii) During equinox, in general, the agreement between observed and predicted values of foFl is good when both values are present and seems to be better during low solar activity than during high solar activity. Discrepancies between observations and model predictions are found in the time of occurrence of the FI layer.

(iii) During summer IRI-90 model generally predicts well the values of the parameter foFl although some discrepancies between measured and predicted values are also found in the time of occurrence of this parameter.

Conclusions

The results of this test study indicate the need of an extension of this analysis to more stations to confirm the necessity to introduce improvements in the IRI-90 model that give a better representation of the observed values of the parameter foFl.

Acknowledgments: The author undertook this work with the support of the “ICTP Pro­gramme for Training and Research in Italian Laboratories”, Trieste, Italy.

References

Bilitza, D., International Reference Ionosphere 1990, NSSDC/WDC-A-R&S 90-92, World Data Center A for Solar Terrestrial Physics, Boulder, U.S.A., 1990.

DuCharme, E.D., Petrie, L.E. and Eyfrig R., A method for Predicting FI Layer Critical Frequency Based on Zurich Smoothed Sunspot Number, Radio Sci., Vol.8, 837-839, 1973.

Figure Captions:

Fig.l Comparison between IRI-90 predicted values of foFl(*) and observations (*) at Tucuman for different seasons of the year 1973 {R = 38).

Fig.2 Comparison between IRI-90 predicted values of foFl(*) and observations (*) at Tucuman for different seasons of the year 1966 (R = 47).

Fig.3 Comparison between IRI-90 predicted values of foFl(*) and observations (*) at Tucuman for different seasons of the year 1971 (R = 67).

Fig.4 Comparison between IRI-90 predicted values of foFl(*) and observations (*) at Tucuman for different seasons of the year 1968 (R = 107).

MH

z M

Hz

48

PREDICTED AND MEASURED F0F1TUCUMAN - MARCH 1673

PREDICTED AND MEASURED FCF1TUCUMAN - JUNE 1973 (NO IRI PREDICT.)

PREDICTED AND MEASURED F0F1TUCUMAN - SEPTEMBER 1973

PREDICTED AND MEASURED F0F1TUCUMAN - DECEMBER 1973

10 11 16 17

5 I} /•? l<i

10 11 13 14 16 16 17 18

Fig.l

MH

z M

Hz

49

PREDICTED AND MEASURED F0F1TUCUMAN - MARCH 19GS

PREDICTED AND MEASURED F0F1TUCUMAN - JUNE 1966 (NO IRI PREDICT.)

% 3

10 11 12 13 14 1S 16 17 1612 13 14 15 16 17 1810 11

PREDICTED AND MEASURED F0F1TUCUMAN - SEPTEMBER 1866

PREDICTED AND MEASURED F0F1TUCUMAN • DECEMBER 1966

10 11 12 13 14 16 1 17 18

Fig.2

MH

z M

Hz

50

PREDICTED AND MEASURED FCF1TUCUMAN - MARCH 1971

PREDICTED AND MEASURED F0F1 TUCUMAN - JUNE 1971 (NO IRI PREDICT.)

I..T. L.T.

PREDICTED AND MEASURED F0F1TUCUMAN - SEPTEMBER 1071

PREDICTED AND MEASURED F0F1TUCUMAN - DECEMBER 1971

10 11 13 14 1S 16 17 18 10 11 12 13 14 16 1 17 18

Fig. 3

MH

z M

Hz

51

PREDICTED AND MEASURED F0F1TUCUMAN - MARCH 1968

PREDICTED AND MEASURED F0F1TUCUMAN - JUNE 1608 (NO IRI PREDICT.)

10 11 17 1812 1

PREDICTED AND MEASURED F0F1TUCUMAN • SEPTEMBER 1668

PREDICTED AND MEASURED F0F1TUCUMAN - DECEMBER 1868

10 11 12 13 14 16 18 17 1816 17 18

Fig. 4

iniiiminiXA9743636

52

ON THE APPLICATION OF THE DU CHARME FORMULA FOR PREDICTING THE FI LAYER CRITICAL FREQUENCY

Bruno Zolesi1 and Marta Mosert de Gonzalez2 1 Istituto Nazionale di Geofisica, Rome, Italy

2 Casleo, Conicet, San Juan, Argentina.

Abstract: A short analysis on the median monthly hourly values of the critical frequency of the FI layer, taken in five european stations during more than two solar cycles, gives some indications on the applicability of the Du Charme et ah, formula (Du Charme et ah, 1973). Comments and considerations are presented on the results of the formula for twilight hours and on the prediction of the occurrence of the FI layer including L conditions.

1. Introduction

A correct evaluation of the characteristics of the FI layer is an important part of many models for the electron density profiles. The FI layer as well the E layer are often considered as simple Chapman layers without taking into account their complex behaviour especially in the transition periods as for high values of solar zenith angles. The Du Charme formula (Du Charme at ah, 1973), a method for predicting the FI layer critical frequency as a function of the smoothest sunspot number, gives very good performances and it is still used in many electron density profile procedures as IRI (Bilitza, 1990) or DGR (Radicella and Zhang, 1995) etc.

2. Comments and Discussions

Values calculated using the Du Charme formula have been compared with the median monthly data observed in five European ionospheric stations. (Table 1.) taken during differ­ent periods and solar cycles. Values predicted by the Du Charme formula are very close to the observed data in every station considered, including that at high latitude. No important differences exist using geomagnetic or dip latitudes coordinates. On the other hand problems occur for high values of solar zenith angle when the function cosx in the formula goes quickly to 0. This inconvenient may be bypassed applying a function that defines the occurrence of the FI layer when the solar zenith angle is below a given maximum value or, as it is in the IRI, omitting the FI feature at night and in the winter months. These solution seems to be too drastic for different reasons:

a) For strumental reasons, infact in the statistical analysis of the median observed values of the critical frequency there is the reasonable certainty of a correct scaling of the ionograms; instead for what concerns the analysis of the L events, the results are strongly dependent by the efficiency of the ionosondes. The data Bank used by Du Charme refers to years from 1958 through 1969 when the quality of the ionograms were not so good like in the digisonde era.

53

b) For geophysical reasons, infact considering the importance of the solar activity in the occur­rence of the FI layer we should take into account also the special trend of the last two solar cycles.

A rough analysis on the 5 European stations, made comparing observed and predicted values leads to the following comments also briefly summarised in Table 1. In low solar activity (R12 < 50), during summer there are many values observed in the twilight hours and not predicted by the Du Charme formula. During the winter months there are still many observed values for every hour of the day and not predicted. The situation is a little different in high solar activity (R12 > 80). In summer there are still some values observed and not predicted in twilight hours. The number is lower than in low solar activity and decreases with the latitude. During winter the prediction is good, sometimes there are predicted values when there are L events observed.

3. Conclusions

The Du Charme formula gives already excellent results predicting the critical frequency of the FI layer, however a new and recent data set should be applied to improve the formula taking into account the strong restrictions for high values of solar zenith angle. Special attention should be given to correctly predict the values of the critical frequency during twilight hours and to evaluate the probability of occurrence of the layer including and considering L events. A model to predict a “virtual value” of electron density, that is in those situations where the layer is not really observed but is still existing, should be considered too, for applications in the electron density profiles models.

References

Du Charme E.D., Petrie, L.E. and Eyfrig, R, A method for predicting the FI layer critical frequency based on the Zurich smoothest sunspot number. Radio Science, Vol. 8, number 10, pages 837-839, 1973.

Bilitza D., International Reference Ionosphere, NSSDC90-22, WORD DATA CENTER A, Greenbelt, USA - page 155, 1990.

Radicella S. and Zhang, M.L., The improved DGR analytical model of electron density height profile and total electron content in the ionosphere. Annali di Geofisica, 38 (I), 34-41, 1995.

54

Table 1

HIGH SOL. ACT. LOW SOL. ACT.

Station lat. wint. summ. wint. summ. Years

Rome 41.8 o o o * 0 49-89

Beograd 44.8 o 0 o o o 0 64-84

Poitiers 46.6 o * o o o o 64-84

Lannion 48.8 * 0 o 71 -84

Uppsala 59.8 * o o 67-76

t tdiurnal hours twilight hours

O o Many values observed and not predicted

O Values observed and not predicted

* Values predicted and not observed

XA9743637

55

VARIABILITY OF THE ELECTRON DENSITY AT FIXED HEIGHTS NEAR FI REGION

Man-Lian Zhang and Sandro M. Radical la International Center for Theoretical Physics, Trieste, Italy.

Abstract: The relative variability of the electron density at fixed heights below F2-peak of every 10-20km starting from 130km are calculated from the profiles inverted from digisonde ionograms by the ARTIST method for the stations of Jicamarca, Ramey, Millstone Hill and Puerto Madryn for the low solar activity years 1992-1993. The minimum variability and its corresponding height, the minimum variability height, are identified for the local time from 10:00 to 16:00. The behaviour of the minimum variability height and its relationship with the geographical latitude as well as the average Fl-peak height are studied.

Introduction

It was found that, in terms of the percentage ratio of the standard deviation to the mean value, the variability of the electron density at fixed heights reaches a minimum value at some certain height (Mosert de Gonzales and Radicella, 1994; Radicella et al., 1994; Zhang et ah, 1994). However, the relationship of the minimum variability height with the Fl-peak height as well as the geographical latitudes is still not clear. The purpose of the study in the present paper are (1) to confirm the conclusion obtained by previous studies that there exists a height at which the variability of the electron density at fixed heights reaches a minimum value; (2) to obtain the relationship between the minimum variability and its corresponding height with the geographical latitude; (3) to study the relationship between the minimum variability height and the Fl-peak height.

Data Used

The digisonde data of four stations with geographical distributions are used for the present study. The electron density profiles are converted from the digisonde ionograms with the ARTIST method. The four stations and the years used are shown as follows

Stations Year(s) Number of daysJicamarca ( -11.9N, 283.IE) Ramey ( 18.5N, 292.0E)Millstone Hill ( 42.6N, 288.5E) Puerto Madryn ( -42.7N, 294.7E)

19931992-19931992-1993

1993

91897248

Analysis and Results

Figs.l(a)-l(d) show the electron density profiles converted from the digisonde ionograms for the samples of the local time 12:00. From the obtained electron density profiles, the electron densities at fixed heights of every 10-20 km are extracted. For each hour from 10:00 to 16:00LT,

56

the standard deviation and the mean values are calculated for each fixed height. The relative variability V(%) is calculated as follows.

V(%)= Standard Deviation/Mean *100%

The results for the four stations are shown in Figs.2(a)-2(d). It is clearly seen from Figs.2(a)- 2(d) that there exists always a minimum variability at some heights for all the four stations and for all the local times from 10:00 to 16:00LT. This minimum variability is about 10-25%, while its corresponding height is generally located between 150-190km. This can be seen more clearly in Table I, where the minimum variability (Vmin) and its corresponding height, the minimum variability height (h-vmin), are listed. From the Table I, it is shown that, in general, the minimum variability of the electron density at fixed heights and its corresponding height have a tendency to increase with increasing geographical latitude.

To study the relationship of the minimum variability height (h-vmin) with the Fl-peak height hmF 1, we also compare the minimum variability height with the average Fl-peak height. The results are shown in Table II. It can be seen from Table II that the minimum variability height is very close to the average Fl-peak height. The difference between these two heights (dh in Table II) is less than 20-25km in most of the cases. This result implies that the Fl-peak height is located in a relatively stable region of the ionosphere. This may provide a base to the possibility of the modelling the FI region ionosphere.

Conclusion

From the present study, the following conclusions can be obtained: (1). The results for all the four stations show that there is always a height at which the variability V(%) is minimum for all the local times from 10:00 to 16:00. In most of cases, the minimum variability height is between 150 and 190km (as shown in Table I). (2). Both the minimum variability of the electron density at fixed heights {Vmin) and its corresponding height, the minimum variability height (h-vmin), have a tendency to increase with increasing latitude. Generally, the minimum variability is about 10% to 15%. (3). The minimum variability height (h-vmin) is very near the average hmF 1 height. In most of the cases, the difference dh between the two heights is less than 20-25km and this difference dh has the tendency to decrease with latitude. This result may give a base to the possibility to model the FI layer.

Acknowledgments: The authors would like to thank Prof. B. Reinisch and the colleagues in University of Massachusetts Lowell, Center for Atmospheric Research for providing the digisonde data and the related software for processing the digisonde data for the present study.

References

Mosert de Gonzalez, M. and Radicella, S.M., Study of Ionospheric Variability at Fixed Heights Using Data from South America. Adv. Space Res., 1994.

Radicella, S.M., Zhang, M.-L. and Jodegne, J.C., Variability of electron density of fixed heights using digisonde data. COST238/PRIME Workshop on “Numerical Mapping And Mod­elling and Their Applications to PRIME”, pp.121, Eindhoven, The Netherlands, 1994.

57

Zhang, M.-L., Zhang, S.-R. and Radicella, S.M., Theoretical Simulation of Ionospheric Vari­ability Below and Above the F2 Peak. Proceedings of the International Beacon Satellite Symposium, 11-15 July, 1994, pp294-297, University of Wales, Aberystwyth, UK.

TABLE I. Minimum Variability of Electron Density at Fixed Height (V_min) and the Corresponding MinimumVariavility Height (h_vmin)

Jicamarca (11.9°N, 283.1°E)

Ramey(18.5°N,292.9°E)

Millstone Hill (42.6°N, 288.5°E)

Purto Madryn (-42.7°N, 294.7°E)

LTh_vmin

(km)Vmin h_vmin

(km)Vmin h_vmin

(km)Vmin h_vmin

(km)VminW

10:00 150 8.66 150 12.61 170 17.19 180 16.1211:00 150 9.87 160 9.33 170 14.87 170 13.2212:00 150 9.60 150 11.21 170 14.08 170 13.7313:00 150 8.97 140 11.59 170 11.25 170 9.9214:00 150 10.82 130 12.28 180 12.87 180 14.1515:00 150 13.20 130 14.71 190 15.24 180 14.4316:00 150 19.29 180 17.49 190 22.63 190 13.98

TABLE II. Mean of the Electron Densities at the Minimum Variability height and the average FI-peak height.

58

( dh=hmF 1 -hvmin )

Jicamarca (-11.9°N, 283.1°E)

LT hvmin dhhmFl NmFl N( hvmin)10:00 173. .278E+06 150. .258E+06 2311:00 170. .284E+06 150. .275E+06 2012:00 172. .289E+06 150. .281E+06 2213:00 173. .288E+06 150. .276E+06 2314:00 172. .274E+06 150. .261E+06 2215:00 177. .268E+06 150. .234E+06 2716:00 183. .249E+06 150. .192E+06 33

Millstone Hill (42.6°N, 288.5°E)

LT hvmin dhhmFl NmFl N( hvmin)10:00 168. .247E+06 170. .265E+06 -211:00 170. .269E+06 170. .281E+06 012:00 167. .264E+06 170. .286E+06 -313:00 168. .263E+06 170. .275E+06 -214:00 171. .259E+06 180. .311E+06 -915:00 171. .225E+06 190. .341E+06 -1916:00 176. . 188E+06 190. .268E+06 -14

Ramey (18.5°N, 292.9°E)

LT hvmin dhhmFl NmFl N( hvmin)10:00 168. .308E+06 150. .254E+06 1811:00 167. .317E+06 160. .302E+06 712:00 165. .308E+06 150. .276E+06 1513:00 170. 313E+06 140. .231E+06 3014:00 171. .301E+06 130. . 188E+06 4115:00 176. .286E+06 130. 155E+06 4616:00 178. .237E+06 180. 249E+06 -2

Purto madryn (-42.7°N, 294.7°E)

LT hvmin dhhmFl NmFl N(hvmin)10:00 186. .300E+06 180. .308E+06 611:00 190. .315E+06 170. .286E+06 2012:00 186. .306E+06 170. .287E+06 1613:00 185. .302E+06 170. .283E+06 1514:00 182. .283E+06 180. .298E+06 215:00 181. .278E+06 180. .292E+06 116:00 187. .265E+06 190. .286E+06 -3

59

.lic;inuMc;i 199> (LT = I 2 >

M.OOI--KX) 5 (KIE+U5 i.00E+()6 1.50E+06 2.O0F.+O6 2.50E+06

Nelli) (a)

MiiKiom- m2 <<; l‘M.l ||;r=12i

n.(Xk'-KM» 5.0Ue+(i5 l.(K>c+(Xi l.50e+06 2 (X>e+0* I50e+06

(b)

Fig. 1

60

Ramey (1992 & 1993 LT=I2)

Mix'’ Ul - ■ylUH.'+US 1 0<)c+or. 1 50c+06 2.00c+00 2.50e+06Nudil (c)

PMADRYN (1993 LT= 12:00)

ii.iki. h« m xmv+ns 00c+06 l.50e+00 : (Mk+06

Nedn (d)

50e+<>6

Fig.l

61

Fig.2

62

R.mivx ,V ,|.T=I(). II i:. I?. \A. 15. IfVi

-JOSul ,-Mv.in MKf , (c)

5U

PMADKYN \W (LT=1U. I I. 12. 13. N. 15. 16)

-h i5 :o :> >o

\ x‘‘' i - Su!.- \Kmii I lx) .

ii -15 50

(d)

Fig.2

XA9743638

63

A SUMMARY OF THE THEORETICAL MODEL RESULTS FOR THE IONOSPHERIC Fl-LEDGE

S.-R. Zhang1 and S M. Radicella2 1 Wuhan Institute of Physics, The Chinese Academy of Sciences,

P.O. Box 71010, 430071 Wuhan, China 2 Atmospheric Physics and Radiopropagation Laboratory, International Centre for Theoretical Physics, Trieste, Italy

A theoretical model for the middle ionospheric structure over mid-latitudes was constructed at Wuhan Institute of Physics (WIP) in 1992 and is now in further development. The model solves the continuity and the momentum equations for the atomic oxygen ions and the continuity equations for molecular NO+, 02+ and N2+. Major dynamic and chemical processes occurring in the height region over the area are taken into account. The model can thus give a proper description of the ionospheric structure particularly below the F2-peak. Research work done during the ICTP/WIP cooperation was focussed on the study of the behaviour of the ionospheric Fl-ledge shape: its occurrence and evolution, the characteristic points on the F1-F2 layers profile, the dynamic response of the ionospheric F region to the atmospheric changes (including the most recent investigations on TIDs). Most of the work is reflected in the Zhang’s PhD thesis. This paper gives a short review of the results relevant to the Fl-ledge.

1. Occurrence and Evolution of the Fl-ledge and the Physical Mechanisms

Variations of the Fl-ledge with local time, season, solar activity and geomagnetic activity are modelled with using the theoretical model. The evolution of the shape is in agreement with observations in such a way that the ledge is well defined near noon, summer and low- solar activity. Key factors responsible for the ledge formation appears to be a small neutral atom/molecule concentrations ratio at the ledge and a high altitude of the atom-molecule transition level, as well as a high temperature and a small gradient of the temperature at the ledge. These parameters are not equally important. The model is also used to study the relative importance of different processes that control the ledge evolution.

2. Characteristic Points

The characteristic points on the profile is defined according to the height gradient of the electron density. Results for the height of the base point and of half the maximum density are present in this issue of publication. It seems to us that the height of the base point not only has strong physical meanings but also carries important information from both the F2-layer and the Fl-layer. It can be expected that ionospheric density profile can be more approximately represented by empirical models like IRI when there exists an intermediate layer, provided this parameter can be properly introduced. (Plots are shown in the Zhang et al.’ paper in this same proceedings.)

64

3. Variability

Calculations were made for some periods (half a month or a full month). The critical frequency of the F2-layer was compared to the observations. This suggests a good consistency in a general trend. In addition, the solar-geophysical index dependence of the foF2, hmF2 and the other parameters can also be found, implying the possible physical controls. Statistical analysis on the output results from the model were carried out by M.-L. Zhang et al. The variability along the height are found to reach a minimum around 170- 190 km, which agrees with the observational data.

4. Comparisons with Observed Profile and IRI Profiles

Several comparisons with IRI and observations have been made. It was found that at Wuchang the daytime profile obtained by using Gulyaeva BO and the nighttime profile using standard BO are thicker in the bottomside F2 region. Experimental results show a similar trend as the theoretical model. After considering the photoelectron impact responsible for additional ionizations in the theoretical model, the observed Fl-ledge shape can be reasonably reproduced.

5. Lower Transition Height (LTH)

The LTH, defined as the prevailing height from molecular ions dominant to the atomic ion, reflects a transition of the relative importance on the ionospheric behaviour from photochemical process to the dynamic process. When the model results were compared to two of the IRI ion composition options [Danilov and Yaichinikov (D&Y) option and the standard option] and the empirical Oliver model, it can be found that three models agree rather well by day except in the summer of lower solar activity, while large differences exist by night. The D&Y option produces a very small day-night difference, while the standard one gives a smaller one in summer and in winter while a larger one in equinox in lower solar activity. The model shows a solar activity dependence and local time variation of the LTHs which is different from the IRI LTHs.

Figure Captions

Fig.l. Evolution of Fl-ledge with local time(a), season(b, c) and solar activity (b, c).

Fig.2. Variations of (dN/dh)min for different local time (a), season (b) and solar activity (b).

Fig.3. Profile varaibilities for non time obtained with the theoretical model (a, b).

Fig.4. Lower transition heights obtained with the theoretical model, the IRI and the empirical Oliver model.

350

300

250

200

150

100(

350

300

250

200

150

100

350

300

250

200

150

100

June 1986, Wuchang 65

March 1986 June 1986

September 1986 December 1986

9 10 11

March 1982 June 1982

September 1982 December 1982

plasma frequency of electron (MHz)14

Fig.l

dN/dh Min. (m-4)

T1OOto

o IV) COCD CD CD CD+ + + +o CD O) CD

CD+CD

Q)

C 3

' CDO3

COCD

"O

oCDO

-

--- 1----

+

------ 1---- ------ 1--

o -

- + O -

- + ❖ -

-__ I__

+____ I__ 1

1----1982

o 1986

+

Wuchang (12LT)

local time (hour)

dN/dh Min. (m-4)PO CD CO —L —L —io o O o O ho bCD CD CD CD CD CD CD CD+ + + + + + + +CD CD CD CD CD CD CD CD

CDCD

June 1986, Wuchang

altit

ude (

km)

* w

al

titud

e (km

)

67

(b)

January 1-15, 1982, Roma

100 -------------------- '-------------------- '-------------------- '--------------------5.00e+10 5.50e+11 1.05e+12 1.55e+12 2.05e+12

electron density (m-3)

August 1-31, 1982, Roma

5.00e+10 5.5064-11 1.0564-12 1.5564-12 2.05e4-12electron density (m-3)

Fig.3

Wuchang, March 1982

simulation — standard IRI — D&Y Model - -

Oliver Model —

a 200180 - -

0 3 6 9 12 15 18 21 24local time / hour

Wuchang, September 1982

simulation — standard IRI — D&Y Model - -

Oliver Model —<» 220a 200

0 3 6 9 12 15 18 21 24local time / hour

altit

ude

/ km

al

titud

e / k

m

Wuchang, June 1982

Fig.

simulation-----standard IRI ----D&Y Model - -

Oliver Model —

0 3 6 9 12 15 18 21 24local time / hour

Wuchang, December 1982

simulation----standard IRI----D&Y Model • -

Oliver Model —

0 3 6 9 12 15 18 21 24local time / hour »