COSPAR 2004, Session C4.2, Advances in Specifying Plasma Temperatures and Ion Composition in the Ionosphere.


Report / Dieter Bilitza


This session was organized by the International Reference Ionosphere (IRI) Working Group with the goal of improving the description of the electron and ion temperature and the ion composition in the IRI model. The 62 presentations were divided into sessions entitled: Plasma Temperatures – Measurements and Models I, II, Modeling the Ionospheric Ion composition, Topside Ionosphere and TEC, Electron Density Below the Peak, Description of Disturbed Conditions, and Posters. A wide array of data sources was used in these studies including satellite data from SROSS C2, CHAMP, GPS, SNOE, SORCE, TIMED, COSMOS 1809, Interkosmos 19, Rocsat, Kompsat 1, Akebono, ACTIVE, ISS-b, DE, AE-C,D,E, OGO-6, AEROS, ISIS, Alouette,  Incoherent Scatter Radar (ISR) data from Saint Santin, Millstone Hill, Irkutsk, and EISCAT, MST Radar data, Ionosonde data, and Rocket measurements. The global reach of the IRI effort was documented by presentations from many different countries including Argentina, Austria, Brazil, Bulgaria, China, Czech Republic, France, Germany, Greece, India, Italy, Ivory Coast, Japan, Korea, Poland, Russia, South Africa, Taiwan, UK, and USA.  Financial support was provided by the Committee on Space Research  (COSPAR) and by the International Union of Radioscience (URSI).


Plasma temperatures


Several new models were presented for the electron temperature (Te) and ion temperature (Ti). Bhuyan et al. (India) used data from SROSS C2 to study variation of Te with local time, season and solar activity in the area covered by the satellite (~400 km, low latitudes, 1994-1998) and to develop a model for solar cycle minimum conditions. The climatology of Te and Ti recorded by the French Incoherent Scatter Radar in Saint Santin over more than 2 decades was represented in an empirical model by S.R. Zhang et al. (USA). Data from the first Taiwanese satellite, ROCSAT-1 were used by Chao et al. (Taiwan) to construct a Ti model including seasonal, diurnal, longitudinal, and latitudinal variations at 600 km. Truhlik et al. (Czech Republic) pointed out ways of improving the IRI plasma temperature models in particular including the dependence on solar activity.


Nighttime data from the Korean KOMPSAT-1 satellite at 685 km are 200 to 400 K higher than the IRI predictions, whereas the electron densities are somewhat lower than IRI (Min et al., Korea). Oyama et al. (Japan) succeeded in taking on of the first Te measurements within a sporadic E layer and reported a lowering of Te within the thin layer. Studying polar conditions with Akebono data, Abe et al. (Japan) find a positive gradient for Te increasing with altitude and Bz. Data from a rocket launched in Norway up to 767 km, show elevated electron temperatures associated with energetic particle precipitation in the cusp (Steigies et al., Germany). Murilakrishna’s (Brazil) rocket measurements show larger Te within plasma bubbles than outside. Proelss et al. (Germany) noted the importance of including the Te enhancement in IRI that is observed under the magnetospheric cleft region and they presented first modeling results.

Richards (USA) focused on the nighttime behavior of Te and Ne at mid-latitudes (Millstone Hill) and studied how well his Field Line Interhemispheric Plasma (FLIP) model and the IRI model reproduce the observed features. The main un-modeled feature is a post-midnight Ne enhancement that correlates with a sharp drop in Te. A possible explanation might be a large downward flow of ions facilitated by the decreasing temperatures.


IRI efforts are now focusing on including a full description of solar cycle variations for the plasma temperatures, on developing an independent Ti model, and on coupling with plasmaspheric temperature models.


Ion Composition and Densities


Triskova et al. (Czech Republic) presented models for the absolute ion densities of O+, H+, and He+ based on data from ACTIVE, AE, and OGO-6 and showed discrepancies with the current IRI model. Many of these discrepancies are due to shortcomings of the IRI electron density topside model (see next section), which is used to obtain the individual ion densities from the IRI ion composition percentages. Another source of error is the constant light ion ratio n(He+)/n(H+)=1/9 currently assumed in IRI. This was also pointed out by Sidorova (Russia), who found large differences between the He+ densities measured by ISS-b and predicted by IRI. A better description of the light ion ratio was proposed at an earlier IRI Workshop by Truhlik et al. (Czech Republic) and is now being considered for inclusion in IRI. 


In the F-region which is dominated by O+ ions IRI performs generally quite well. This was also confirmed by Bhuyan (India) who studied the diurnal, seasonal and latitudinal variations of SROSS-C2 ion densities and compared them with IRI. Lathuillere and Kofman (France) discussed the different EISCAT measurement programs, the accuracy of deduced ion densities, and how these data can be used for studying the ion composition behavior in the auroral and polar ionosphere


Electron Density – Topside and TEC


The ISIS/Alouette topside sounder Ne data sets available from the National Space Science Data Center (NSSDC) were used for a variety of studies. Kutiev et al. (Bulgaria) developed a new model for the topside ionosphere scale height (TISH) as a function of month, local time, latitude, longitude and solar flux F10.7. Comparisons by Belehaki et al. (Greece) show that TISH is in general agreement with scale heights deduced from ground ionosondes but the model predicts post-midnight and afternoon maxima whereas the ionosonde data show a noon maximum. Webb et al. (USA) reported on their effort to deduce changes in the plasma temperature and ion composition from changes in the ISIS/Alouette topside profiles. Using the sounder data Coisson et al. (Italy) discussed limitations and possible improvements of the IRI topside model and the NeQuick model.


Reinisch and Huang (USA) showed how the topside scale height for IRI can be obtained from the bottomside profile using the technique developed for their Digisonde measurements. Stankov (Bulgaria) and Jakowski (Germany) used CHAMP radio occultation data to deduce topside scale heights and plan to develop a model based on these data.


M.L. Zhang et al., (PR China) compared TEC derived from Digisonde measurements at a low latitude station in China with GPS maps (IGS) and with IRI predictions. The Digisonde TEC shows a somewhat better agreement with the IGS values. Of course, the IGS TEC includes also a plasmaspheric contribution that is not included with the other two methods. The properties of the TEC in the equator anomaly region were studied with GPS data by Wan et al. (PR China).


The main focus of future IRI activities is on improvements of the topside electron density profile using a Chapman approach with variable scale height and on merging the topside profile with a plasmaspheric model.


Electron Density – F peak


Abdu et al. (Brazil) compared the evening F layer peak height and vertical drift from several Brazilian ionosonde stations with the IRI predictions and find that the ion drift peak is often underestimated by IRI whereas the F peak density is overestimated. 

Comparisons with Japanese ionosonde data show good agreement with IRI except for small discrepancies of the electron density at sunset (Tsujita et al., Japan).  


Real-time updating of IRI bottomside profiles for the European sector with ionosonde data was discussed by Buresova et al. (Czech Republic) and Stanislawska et al. (Poland). An improved algorithm for IRI-TEC updating was presented by Pelevin et al. (Russia)


Studies of ionospheric variability continued with data from ionosondes in Africa (Obrou et al., Ivory Coast), Europe (Mosert et al., Argentina) and Japan (Ezquer et al., Argentina). A first order variability model for IRI is currently under development based on inputs from the ground and topside ionosondes.

"A first order variability model for IRI is currently under development based

on input from the ground ionosondes (Fotiadis and Kouris) and topside


I have presented a paper on : relation between TEC and

peak density which I submitted for consideration and publication in Adv.Space Res. In this work there was reported the results on

the diurnal and seasonal variations of slab thickness and the variability of TEC , using mainly hourly daily  GPS data from Matera

( Italy ) and Hailsham ( England) . Specifications on these variations are also given in relation with those in foF2.


Electron Density – Bottomside and Below


New models for the electron density in the lower ionosphere at auroral latitudes were presented by Danilov (Russia) and McKinnell (South-Africa). Danilov describes variations in the D-region in terms of solar zenith angle and daily sum of Kp based on rocket data from the Northern and Southern hemisphere. McKinnell applied a Neural Network approach to a combined data base of rocket and EISCAT ISR data. Friedrich et al. (Austria) have combined EISCAT-Svalbard data with polar-latitude rocket data to investigate the “true quiet” electron density profile in the D, E and F regions; this is the minimum envelope of all data.


Solomon (USA) investigated how well theoretical models represent the E-region ionosphere when fed with accurate solar irradiances. He found that the EUVAC solar irradiance model is in better agreement with SNOE, TIMED/SEE and SORCE measurements than the Hinteregger EUV model. But even with the EUVAC model the theoretical model still underestimates IRI and ionosonde E peak densities particularly at solar minimum.


IRI predictions agree overall quite well with data from the Irkutsk ISR (1997 - 2003) (Potekhin et al., Russia) and the Irkutsk ionosonde (Ratovsky et al., Russia), some overestimation is found during summer and underestimation during winter. Lei et al. (PR China) compared the bottomside parameters B0, B1 deduced from Millstone Hill and EISCAT measurements with IRI and pointed out needed improvements.


IRI Publications, Workshops, and New Members


Papers from the 2002 IRI session during the World Space Congress in Houston, USA have been published in Advances in Space Research (ASR), Volume 33, Number 6, 2004. The ASR issue with papers from the 2003 IRI Workshop in Grahamstown, South Africa has just come out as Volume 34, Number 9, 2004. A special session was held during the 2003 German URSI meeting in honor of Karl Rawer’s 90th Birthday. The papers from this session are now published as the 2nd Volume of the Advances in Radioscience (formerly “Kleinheubacher Berichte”).


The 2005 IRI Workshop is now planned for the week of June 27 to July 1 at the Ebro Observatory in Roquetes, Spain ( and its theme will be “New Satellite and Ground Data for IRI, and Comparisons with Regional Models”. For the 2006 COSPAR General Assembly in Beijing, PR China the IRI team has proposed a session on the “Solar Activity Variations of Ionospheric Parameters”.


Three new members were welcome into the IRI working Group: Dr. O. Obrou from the University of Cocody in Abidjan, Ivory Coast; Dr. P.K. Bhuyan of the University of Dibrugarh in Dibrugarh, India; Dr. M.L. Zhang, Center for Space Science and Applied Research, Chinese Academy of Sciences, Beijing, PR China.

* Back to IRI workshops page
* Back to IRI homepage