International Reference Ionosphere 1999 Workshop

Summary Report / Dieter Bilitza

The annual workshops are the life blood of the international IRI effort, since they provide an opportunity to review the status of various modeling efforts and to discuss improvements and additions to this widely used ionospheric model. IRI’99 was held at the Center for Atmospheric Research (CAR) of the University of Massachusetts Lowell (UML) from August 9 to 12. Excellent support was provided by the local host Bodo Reinisch (CAR Director) and his team. The meeting was supported through financial contributions from the International Union of Radio Science (URSI), the Committee on Space Research (COSPAR), the National Science Foundation (NSF) and the UML. The four day meeting was attended by 37 participants representing 14 different countries. 47 papers were divided into the following sessions: Ray Tracing, D Region, Measurements and Comparisons, Temperature and Ion composition, Topside and Plasmasphere, Total electron Content, Improvements and New Inputs, Drift Data and Modeling, Evaluation of IRI and Other Models, Applications, and Final Discussion ( As a result of the presentations and discussions the IRI Working Group decided to include several improvements and additions in the next edition of the model, IRI-2000. It is also planned to publish an IRI-2000 report similar to the very popular IRI-1990 report.

Workshop Topics

The special focus of IRI’99 was on ray tracing and on variability. Different ray tracing methods and techniques were reviewed by Dyson and Benett (Australia). Huang and Reinisch (CAR) presented a multi quasi-parabolic description of IRI density profiles that allows for analytical ray tracing computations. The later is now under consideration for inclusion in IRI. A quantitative description of ionospheric variability (= the standard deviation from a monthly average) has long been a goal of the IRI group, since this parameter is of great interest for many Space Weather applications. Future ICTP Task Force Activities will deal with this important modeling aspect using the presentations from the Lowell meeting (e.g. Gulyaeva, Russia and Mahajan, India) as a starting point (Radicella, ICTP).

Electron Density — Bottomside and E and D regions

In the lowest (D) region, Friedrich (Austria) has now assembled all available and reliable rocket measurements (114 profiles) using the IRI selection criteria (in situ radio propagation experiments). He will provide the latest version of his model as a new option for the IRI D-region. Work continues on more realistic description of the nighttime variation of the E peak density and height (Bibl, CAR; Mahajan, India). Recent work by Titheridge (New Zealand) should be also helpful in this context. Inclusion of a statistical model for the occurrence of Sporadic-E was suggested as a new work item and review talks will be invited for the next IRI meeting (Smith, Boulder). The new bottomside model developed by the ICTP Task Force Activity was accepted for IRI-2000 (Adeniyi, Nigeria; Radicella, ICTP; Reinisch, CAR; Bilitza, RITSS/GSFC; Mosert, Argentina; S. Zhang, China; et al.). This includes a new formula for the probability of F1 occurrence, a realistic representation of the bottomside thickness parameter B0 at low and equatorial latitudes and an analytical description of the intermediate region that is free of artificial valleys and discontinuities.

Electron Density — Topside, Plasmasphere and TEC

Bilitza (RITSS) gave a status report about his topside modeling project. Based on all electronically available electron density profiles deduced from Alouette and ISIS topside sounder measurements, his approach plans to use a number of fix points to describe the shape of the topside profile (e.g., the inversion point, the scale height transition point). Comparisons with Hinotori insitu measurements (Ezquer, Argentina), with ATS-6 TEC data (Ezquer, Argentina), with GPS data (Breed and Goodwin, Australia) and with Arecibo incoherent scatter results (Sethi, Pandey, Mahajan, India) all demonstrated the need for an improved IRI topside model. Gallagher (MSFC) presented the status of his plasmaspheric model and linkage to the IRI. Although fairly advanced the model is not yet ready for release. The IRI’2000 Workshop will have the topside and plasmasphere as its special topic and will have the inclusion of a plasmaspheric model as one of its most important goals.

Electron Density - F Peak Parameters

In general IRI describes the monthly average F peak densities (NmF2) reasonably well using either the CCIR (over land) or URSI (oceans) options. In the case of the peak height hmF2 there is, however, a clear need for improvement especially in the low latitude dusk sector. A prerequisite for a new modeling effort is an increased global data base for this parameter or as stated in a resolution submitted to URSI: more topside sounder measurements are needed for a better global representation of the F peak height. This resolution was adopted by Commission G during the 1999 URSI General Assembly (Toronto, Aug 13-21). Richards (U Alabama Huntsville) investigated large enhancements in the nighttime NmF2 for magnetically quiet conditions at Millstone Hill. IRI provides a nighttime enhancement much smaller than the observed values. Richards’s FLIP model can reproduce about half the enhancement if it is adjusted with the measured electron temperatures. For stormtime conditions it is planned to include in IRI-2000 the updating algorithm proposed by Fuller-Rowell (NOAA) during the 1998 IRI Workshop.

Temperature, Ion Composition and Drift

Potekhin (Russia) found good agreement between IRI and plasma temperatures measured by the Irkutsk incoherent scatter radar during summer and discrepancies of several 100 degrees during winter. AKEBONO data could be a good source for modeling the plasma temperatures in the transition region from ionosphere to plasmasphere (Oyama, Japan). Data from several Russian satellites were used to establish global models of the electron temperature at several fixed heights. Comparisons with DE and Hinotori data show good agreement with the model values (Truhlik, Czech Republic). These models will be used to improve the topside electron temperature model for IRI. Mass spectrometer measurements from the Russian ACTIVE satellite and the AE-C,-E satellites have been used to study the global, seasonal, diurnal and solar cycle variations of the upper transition height in the ion composition (Triskova, Czech Republic). It is planned (although not for IRI-2000) to replace the current ion composition model in IRI with a model that is anchored by the transition heights (Bilitza, Grebowsky, GSFC). An important new addition to the parameters provided by IRI will be the vertical ion drift. For IRI-2000 it was decided to include the model for equatorial latitudes developed by Scherliess and Fejer (USU). Scherliess reported about the global model that he is working on now and that might become a good candidate for a later update of the IRI model. Drift data from the Japanese MU Radar were presented by S. Zhang (China).


Total Electron Content (TEC) is the ionospheric parameter that is most important for many applications. Reinisch (CAR) showed how TEC can be deduced from groundbased DIGISONDE measurements and Radicella (ICTP) used his DGR model to obtain TEC from IK19 topside sounder measurements. Bust et al. (U Texas Austin) and Cornely (UML) described how IRI and similar models can be of help in tomographic reconstruction of ionospheric contour maps. IRI-related applications and web systems were reviewed during a special half-day session including: telecommunications (Bradley, U.K.), evaluation of TEC computations (Wilkinson, Australia), ionospheric corrections for microwave remote sensing from space (Abraham and LeVine, GSFC), WWW interfaces for data/model computations, retrieval and plotting (Conkright, NGDC; Bilitza, GSFC), an IRI Windows interface (Huang, UML), and the use of IRI for FMCW signal simulations (Potekhin et al., Russia). Bradley (UK) pointed out that the International Telecommunication Union (ITU) requires specific radio propagation parameters that are in general different from the standard IRI output parameters. He will coordinate an effort to use IRI for the computation of ITU parameters. The WDC C2 Kyoto maintains Web pages that let users compute ionospheric conductivities using the IRI and CIRA models. This service was recently discontinued. But because of several requests from users and from the IRI team they were put back online (Araki, Japan). A number of these IRI-related systems were also described during the URSI session G5 on Ionospheric Data and Models Available on the WWW; agenda and links are available at

New Developments

A neural network approach to representing F peak and profile parameters from South Africa was presented by Poole and McKinnell (South Africa). Rawer and Eyfrig (Germany) studied longterm trends in M(3000)F2 data and found non-negligible secular changes. IK-19 topside sounder data were used to demonstrate the ionospheric effects induced by seismic activity and an attempt was made to explain the coupling mechanism (Pulinets, Russia). An evaluation of ionospheric models is being undertaken by Decker et al. (AFRL). First results for two TEC stations (Hamilton and a Taiwanese station) show that the three models perform equally well and that the predictions are about a STD away from the measured mean. Rich and Sultan (AFRL) investigated the shortcomings of the IRI, PIM, and RIBG topside models with DMSP insitu measurements at 840 km. The model predictions are again fairly close to each other and up to a factor 2 from the data. Results were also presented on a comparison of IRI with a station-specific model for Millstone Hill (Buonsanto, Millstone Hill). It showed small discrepancies in representing the solar cycle variation and indicated the importance of including in IRI the motion of the auroral trough.

Members, Meetings, Publications, ISO, URSI Resolution

A number of new members were elected into the IRI Working Group: J. O. Adeniyi (Nigeria), A. Poole (South Africa), S.P. Gupta (India), R. Ezquer (Argentina), S. Pulinets (Russia), X. Huang (UML, USA). Four members are no longer participating in the IRI effort: B.C. N. Rao (India), L. McNamara (Australia), W. Hoegy (USA), K. Champion (USA). The next IRI meeting will be held as session C4.1 during the COSPAR Scientific Assembly in Warsaw, Poland (July 16-23). Special emphasis will be given to topside and plasmasphere modeling. The editing process for the papers from the 1998 IRI Workshop was completed in July 1999 and the package of 32 papers on 226 pages was submitted to Advances in Space Research (Editors: K. Rawer, D. Bilitza, K. Oyama and W. Singer). Currently there are two proposals before the International Standardization Organization (ISO) for an international standard ionosphere involving the SMI and the IRI models, respectively. The Lowell Workshop provided a good opportunity to discuss the consolidation of these two efforts and resulted in plans for a joint project (Gulyaeva, Chasovitin, Russia; Reinisch, UML; Bilitza, RITSS). During the URSI General Assembly in Toronto (Aug 99) Commission G adopted the following IRI-related resolution:

Recognizing the need for an international standard for the specification of the ionospheric environment, and

Recognizing that the Presidents of URSI and COSPAR have written to international organizations in support of the International Reference I onosphere as an ionospheric standard,

URSI Commission G resolves

That the International Reference Ionosphere (IRI), as developed by the URSI/COSPAR Inter-Union IRI Working Group, be internationally recognized as the standard for the ionosphere.

* Back to IRI workshops page
* Back to IRI homepage