The 2001 IRI Task Force Activity took place from May 21 to 25 at the Aeronomy and Radiopropagation Laboratory of the Abdus Salam International Center for Theoretical Physics in Trieste, Italy. Participants included J. Adeniyi (Nigeria), G. Miro (Spain), F. Antonaci (Italy), D. Bilitza (USA), A. Calzadilla (Cuba), L. Ciraolo (Italy), P. Coisson (Italy), B. Forte (Italy), M. Mosert (Argentina), M. Juan (Spain), F. Miquel (Spain), S. Pulinets (Russia), S. Radicella (Italy), and B. Reinisch (USA). The primary focus of this Activity was the development of a specification model for ionospheric variability. Such a model is high on the wish list of users of ionospheric models. Climatological models like IRI provide monthly mean values of ionospheric parameters. Understandably a satellite designer or operator needs to know not only the monthly average conditions but also the expected deviations from these mean values.

The newest version of IRI (IRI-2001) is now ready for distribution. The ICTP Task Force Activity provides a first opportunity for using the new version in comparisons with data. The most important changes were explained and discussed by Bilitza (USA) with special focus on the stormtime updating algorithm for foF2 as developed by Fuller-Rowell et al. The Fortran code was ftped from the NSSDC archive site. The code was cleaned up to comply with the Fortran compiler used at ICTP on their SPARCs and SUN systems.

Recent results from the Jicamarca ionosonde were presented by Reinisch (USA), the bottomside thickness parameter B0 agrees quite well with the IRI-2001 predictions, whereas the shape parameter B1 shows a smaller diurnal variation than assumed in IRI-2001 (which is 2.6 for night and 1.8 for day). Adeniyi (Nigeria) presented D1 measurements from Ougadougou, Burkina Faso. D1 is the shape parameter for the F1 layer. The data (low solar activity) are a factor of 3 higher than the data for high solar activity from Jicamarca.

Variability of ionospheric parameters and how best to represent it was the primary topic of the 2^nd day. A first table was established based on the observation from Ougadougou as interpreted by Adeniyi (Nigeria) and the observation from Tucuman, Argentina as interpreted by Mosert et al. (Argentina).

Ougadougou:

Tucuman and San Juan:

Mosert et al. also studied different parameters to specify variability: standard deviation (in % of mean), difference between from median to lower quartile and to upper quartile (in% of median), difference from lower quartile to upper quartile (in % of median). The first and last values are comparable whereas the other two are about half the value. The median and the inter-quartile difference have the advantage of being less affected by the large dviations that can occur during magnetic storms.

Requirements were discussed for the development of a specification model for ionospheric variability. Latitudinal differences are expected from the magnetic equator to the anomaly crest to middle latitudes and to high latitudes. Besides an altitudinal differentiation a model also should account for the diurnal differences and the changes with solar activity. The plan would be to establish a table of variability values for the different conditions similar to the model for the bottomside thickness parameter B0.

The main topics were techniques for obtaining the ionospheric total electron content (TEC), the variability of TEC, and the ionospheric effects of magnetic storms. Juan (Spain) discussed the global mapping technique used at the Polytechnical University of Catalonia (UPC) to obtain vertical or slant TEC from GPS signals. He specifically showed examples of how the UPC team is evaluating the reliability of their method. Unfortunately, other GPS teams have not yet done similar test. A comparison with a large amount of TOPEX data clearly favors the 2-layer methods (UPC and JPL) over the simple 1-layer methods (ESA, CODE, etc.). Garcia presented results of a study that uses ionosonde data and GPS/MET (low altitude GPS satellite) data to construct more reliable TEC maps from GPS data. Ciraolo explained the method used by their group at the University of Florence and documented the importance of accounting for the full time derivative and not only the partial derivative neglecting the satellites movement during the measurment.

The variability ((Upper quartile — Lower quartile)/median) of TEC was studied for three South American stations (Mosert, Ezquer et al.). Arequipa, Peru close to the magnetic equator, Santiago, Chile close to the Southern crest of the anomaly, and Rio Grande at mid latitudes. The variability is largest at the crest station (about 50% for day and night) and lowest at mid latitudes (30%). Near the equator the variability of TEC changes from 35 % during day to 55% during night with a peak of 60% at sunrise. The other two stations show almost no diurnal differences in variability, except for a dawn peak of about 50% for the mid-latitude station. Comparisons with IRI show that in general IRI underestimates the daytime values and overestimates the nighttime values. The corrections introduced with the new IRI-2001 (B0) are in the right direction and will lead to better agreement with the data especially close to the magnetic equator.

The stormtime variation of topside parameters was studied by Pulinets with InterKosmos 19 (IK-19) topside sounder data. His data indicate the importance of considering the UT/LT effect and he shows changes not only in peak height and density but also in the shape of the topside profile. Calzadilla highlighted longitudinal differences by comparing the storm response as seen in IK-19 data received at Havana, Cuba with IK-19 data from the European sector. Gloria investigated the main differences between the quiet and storm time foF2 and hmF2 at middle latitudes with ionosonde data from Spain and Italy. As a followup the predicitions of the new IRI strom model are being computed for these specific stroms and will be compared with the measurements.

The main topic on this day was the improvement or replacement of the IRI topside model. Bilitza summarized the main problems of the current model: (1) Overestimation of measurements at high altitudes and underestimation at F region heights; (2) The total electron content (TEC) computed with IRI underestimates the measurements during high solar activity and overestimates the measurements during low solar activity especially at low latitudes. Comparing Alouette and ISIS topside sounder data with IRI, Bilitza found a systematic height-dependent misrepresentation of the data. From this analysis it might be possible to deduce a height-dependent correction factor that would provide a first order improvement of the current IRI topside model. Pulinets reported about progress with his Epstein representation of IK-19 data. An extension to low solar activities is planned with Cosmos 1809 data. Coisson has used the IK-19 data to deduce a correction factor for the topside model developed by Radicella and his team at ICTP.

An interesting effort was undertaken by Ezquer et al. in the search for more topside data. They have taken the insitu RPA data from the Japanese TAIYO satellite as published in an ISAS report and have put them into computer accessible form. The TAIYO data are from the solar cycle minimum period 1974-75 from a low-inclination orbit (30 degrees) in the altitude range 300 to 700 km. The comparison with IRI shows in general very good agreement.

ITEC obtained from Digisondes was discussed in presentations by Reinisch, by Adeniyi, and by Mosert. Bilitza discussed characteriastic points in the topside ionosphere that could be used to anchor the topside profile: the inflection point (hmin/Nmin) close above the F peak, (2) the height where the plasma scale height changes.

Comparison of slant TEC and IRI at Palehua (24N, 202E) show underestimation by the model during daytime and overestimation during nighttime (Cabera, Ezquer et al.).

On the last day of the activity plans for future work were discussed. The main focus will be on establishing a table of variability values as a first attempt towards a variability model. It was decided to take the median and inter-quartile range as the primary indicators of variability. Mean values and standard deviations can be also used but should be based on data sets from which magnetically disturbed periods are removed. Statistical studies will be undertaken with data from stations at mid-latitudes, stations near the crest, and stations close to the magnetic equator.

The parameters primarily studied will be the F peak plasma frequency foF2 and height hmF2, the bottomside shape parameters B0, B1, and D1. The monthly variability will be investigated for four time periods (Local Time: 10-14, 18-20, 22-2, 5-7), two levels of solar activity, and the four seasons. To get an indication of the changes in variability with height, the monthly variability will be also studied at 50, 100, and 150 km below the F peak and at 50, 100, 200, and 300 km above the peak. For heights above the peak the investigations will be based on Interkosmos 19 and Cosmos 1809 data (S. Pulinets) and on Alouette 1, 2 and ISIS 1, 2 data (D. Bilitza).

A number of these stations have also TEC measuring capabilities through either the Digisondes (ITEC) or collocated GPS ground receivers. These data will help to establish representative variability values for TEC. Another important source for TEC variability studies will be the global TEC maps established by various groups from the international GPS measurements (M. Hernandez-Pajares).

Regarding topside modeling for IRI, an effort will be undertaken to develop a height-dependent correction term for the current IRI model based on the topside sounder data from Alouette and ISIS (D. Bilitza) and from Interkosmos 19 (P. Coisson, ICTP). This should help to correct the overestimation at high altitudes and the underestimation at low altitudes. Efforts will continue to establish characteristic points in the topside profile and to study their global and temporal variations (e.g., the inflection point above the F peak and the transition height from O⁺ to light ions). As a TEC specific issue the influence of the topside profile function (close to the peak) on the TEC will be investigate in an effort to explain very high measured TEC values that are difficult to reproduce with topside models. Topside sounder data will be used to investigate the large TEC values found near the crests of the equator anomaly.

The date for next year’s ICTP IRI Task Force Activity was tentatively set to June 24-28, 2002. A website will be established at ICTP with the papers from the IRI Task Force Activities.

Back to IRI workshops page

Back to IRI homepage