The 1995 Workshop of the International Reference Ionosphere (IRI) was held at the National Physical Laboratory in New Dehli, India in the second week of January. This was the first IRI workshop in India and quite fittingly the topic was the modeling of the low-latitude ionosphere. The IRI working group was greatly honored by the fact that the IRI Workshop was made part of the Diamond Jubilee celebrations of the Indian National Science Academy (INSA) which took place during this same time period. The festive inauguration of the IRI workshop on Monday morning was highlighted by the inaugural address of Prof. S.K. Joshi (President of INSA). Prof. E.S.R. Gopal (Director of NPL), Dr. A.P. Mitra (Chair of LOC), Dr. D. Bilitza (IRI Chair), and Dr. D. Anderson (VIM Chair) extended their welcomes to the workshop participants and Dr. K. K. Mahajan (Local Organizer) concluded the inaugural session with his vote of thanks. Support from the local organizers was excellent during all phases of the meeting. A special thank you goes to Dr. K.K. Mahajan and his team who made it all possible through their untiring efforts before, during and after the workshop.
The workshop convened under the sponsorship of COSPAR, URSI, COSTED, TWAS, ISF/ Soros, IAGA, INSA and CSC. COSPAR, COSTED, URSI, TWAS, INSA and ISF provided also direct financial support for participants from developing and post-communist countries. Of the more than 60 participants about half were from India and the other half from Brazil, China, Czech Republic, England, Japan, Poland, Russia, Taiwan and USA. The host country was represented by scientists from ISRO Bangalore, PRL Ahmedabad, VSSC Trivandrum, Saurashtra U. Rajkot, Dehli U., Dibrugarh U., Andhra U. Visakhapatnam, Govt. College Bareli, Barkatullah U. Bhopal, and NPL New Dehli. The more than 70 papers were grouped into the following 10 sessions: Low Latitude Data and Comparisons with IRI; Theoretical Studies; Scintillations, Spread-F and other Irregularities; Verification of Ionospheric Models; F-Region Improvements; Ion Composition; Regional Models and Mappings; TEC and Tomographic Studies; Topside and Plasmasphere; D-Region. A half-day session on Wednesday was reserved for the Validation of Ionospheric Models (VIM) effort of URSI Working Group G.4 (D. Anderson, AFPL, Chair); this activity is closely correlated with the URSI/COSPAR IRI effort. A Final Discussions and Decisions session on Friday afternoon concluded the workshop.
Following the trend of recent IRI meetings, increasing weight is now being put on regional modelling/mapping, on results from computer simulations that could benefit the IRI modelling, and on accurate representation of the total electron content (TEC), which is the ionospheric parameter most important for a wide range of applications/forecasts; this is in addition to the more traditional statistical, comparative IRI studies involving large data volumes. Data sources used in the comparative studies presented at this workshop included: (i) TOPEX and ACTIVE satellite data, (ii) Total Electron Content (TEC) data from equatorial ground stations to the Navy Navigation Satellite System (NNSS), to FLEETSAT, and to ETS, (iii) Indian, Russian and Brazilian rocket data from equatorial rocket flights, (iv) data from low-latitude ionosondes (digisondes and earlier instruments), (v) data from incoherent scatter and HF radars at low latitudes.
Faraday rotation TEC measurements over New Dehli from 1977 tto 1980 show reasonable agreement with IRI; IRI, however, does not reproduce the distinct postsunset maximum that is observed during winter and equinox (L. Singh, J. Gupta, T. Tyagi, NPL, India).
M. Abdu (INPE, Brazil) reviewed IRI studies with data from the Brazilian longitude sector. Overall good agreement was found. However, a number of specific problems are noted that need to be addressed. In particular, it was found that at equatorial latitudes IRI over-estimates TEC at low solar activity and underestimates TEC at high solar activity. Similar results were reported for the high solar activity TEC at the anomaly-crest-stations Rajkot, India (K. Iyer) and Luping, Taiwan (K. Cheng). Using the Luping TEC data K. Cheng confirmed that the equator anomaly is mainly a daytime phenomenon (sunrise to almost midnight) and that the winter crest is usually larger than the summer crest. IRI describes the diurnal occurrence pattern correct but not the winter maximum. Further studies were initiated. It is especially important to combine the TEC data with ionosonde data for the peak to establish the cause for the remaining discrepancy (peak or topside?). [Efforts continue: Abdu, Cheng, Iyer]
TEC values deduced from TOPEX altimeter measurements provide a global mapping of the equator anomaly region in the ocean areas with a 10-day repeat cycle. The data show considerable longitudinal and seasonal differences in the location and magnitude of the anomaly crests reflecting the influence of the neutral wind and the ExB ion drift (Bilitza, USA).
Interesting first results were presented from tomographic studies across the equator anomaly region at American longitudes (J. Klobuchar, D. Anderson et al., AFPL, USA) and at Taiwanese longitudes (C. Huang, K. Yeh et al., NCU, Taiwan). Both teams use the NNSS dual frequency signals and both teams are in the process of setting up more receiving stations to improve the quality of their tomographic reconstruction. The Taiwanese team uses IRI to specify initial conditions.
Data from the incoherent scatter radar at Arecibo, Puerto Rico have contributed largely to the development and improvement of the IRI electron density model in the middle ionosphere. K. Mahajan (NPL, India) reviewed the earlier studies as well as possible future improvements. In a joint contribution with T. Gulyaeva (IZMIRAN, Russia), he proposed an improved low-latitude formula for the half-density point; this will be included in the next version of IRI. [New IRI formula for h0.5: Mahajan, Gulyaeva, Sethi]
B. Reinisch (UML, USA) used data from several digisondes operating at low latitudes to study the electron density profile between the E and F peaks. The dominant IRI parameters in this altitude range are strongly biased towards mid-latitudes because of the global uneveness of the underlying data base. Following a recommendation that was made at last year's IRI task force activity at ICTP (Trieste, Italy), his team investigated the density and gradient at a fixed height (170 km) rather than at the height of the F1 ledge, which typically varies by 10-20% from day to day. It was suggested to also study the density variation at a second fixed height in the 180-190 km range; the difference between the two points would than provide the density gradient. Inclusion of these two points in IRI should lead to a better representation of the electron density in the F1-region. The UML team also presented a better algorithm for obtaining monthly average representative profiles (MARP). [N170 and N185: Reinisch and Digisonde community; Radicella and ICTP team]
The IRI model in the E-F region was also compared with several station-years worth of data from China (X.-Y. Huang and S.-R. Zhang, Wuhan, China), from Argentina (M. Mosert de Gonzalez, Argentina), and from Arecibo, Puerto Rico (V. Pandey and N. Sethi, NPL, India). Although overall IRI represents the most important trends, specific discrepancies were noted and need to be further investigated. N. Sethi (NPL, India) fitted a combination of 4 LAY functions to Arecibo E-F region profiles and proposed a number of improvements for the parameters used in the current implementation of the IRI LAY-option.
Interesting new results obtained with the Sheffield University Plasmasphere Ionosphere Model (SUPIM) were presented by G. Bailey and N. Balan (Sheffield U., U.K.). Studying the effect of the neutral winds in different longitude sectors they were able to explain the longitudinal differences in the interhemispheric asymmetry of the anomaly. During prenoon hours when the upward drift is large and the plasma flows towards the equator from both hemispheres they find an additional peak/layer ("G-layer") a few 100 km above the F layer. Y. Su, K. Oyama et al. (ISAS, Japan) find that SUPIM reproduces quite well the electron densities and temperatures measured by the OHZARA satellite at 600 km altitude, if the ExB drift and the neutral winds are reasonably adjusted.
A discrepancy between IRI and observations, that was often mentioned during this workshop, involves the F peak height hmF2. IRI does not reproduce the characteristic evening peak in hmF2 observed at the magnetic equator. This feature is closely related to the sharp spike in upward ion drift at that time. D. Anderson (AFPL, USA) showed how this aspect of IRI could be improved using parameters from his Parameterized Ionospheric Model (PIM). PIM is based on a theoretical model that includes as input the Jicamarca ion drifts, and thus provides a much better representation of the equatorial hmF2 in the evening. D. Anderson offered to provide analytic representations of the PIM hmF2 and F layer thickness parameters and it was decided to include these as a new PIM option in IRI. Outside the equator anomaly region good agreement is found between hmF2-IRI and observations, as was shown by V. Pandey and N. Sethi (NPL, India) for Arecibo incoherent scatter data. [PIM formulas: Anderson]
First results of their respective modeling efforts were presented by the Wuhan/ICTP modeling team (S.-R. Zhang, Wuhan, China), by K.S. Murthy (Andhra U., India) and by M. Goel (NPL, India). Modeling efforts of the ICTP/Wuhan team are focussed on the lower transition height (atomic to molecular ions). Their theoretical results favor the formula established by W. Oliver (1975). Efforts are currently underway to include the transition height as a seperate parameter in IRI; the Oliver model would be an excellent candidate for this project. [Transition heights in IRI: Bilitza, Danilov, Zhang]
Electric Field, Ion Drift, and Spread-F
B. Krishna Murthy (Trivandrum, India) reviewed studies of the F region electric field in the low-latitude ionosphere with particular emphasis on the work done at his equatorial station stressing the similarities and differences to results from Jicamarca. Data from an All-Indian HF Doppler radar network (dip=0-42N) were used by O. Nagpal (U. Dehli), J. Shastri (Bangalore), and C. Raghava Reddi (Trivandrum) to study the contribution of electric fields and meridional winds to the F-region plasma vertical drift. N. Balan and G. Bailey, (U. Sheffield, U.K) find that vertical drifts deduced from hmF2 measurements at Trivandrum (India) agree favorably with Jicamarca incoherent scatter measurements but differ significantly from Trivandrum HF Doppler data. [Bilitza to contact B. Fejer and his team about participation in the IRI ion drift modeling]
Spread-F and Sporadic-E occurrence statistics for the Indian sub-continent were presented by R. Saksena (NPL, India) based on data from this longitude sector. K. Iyer et al. (Rajkot, India) compared scintillations obtained near the anomaly crests with the scintillation models of Aarons (AFPL) and Basu (AFPL) and suggested suitable modifications. Scintillation and Spread-F occurrence statistics were also reported for Waltair (P. Sriram et al., Andhra U., India) and China (X.-Y. Huang and S.-R. Zhang, Wuhan, China). A diffracting lenses simulation of scintillation patterns in the ionosphere was discussed by R. Dabas and D. Lakshmi (NPL, India). A special IRI task force was established under the leadership of M. Abdu (INPE, Brazil) to establish/recommend Spread-F/Scintillation models for inclusion in IRI. [Spread-F task force: Abdu, Basu, Saksena, Iyer, Matuura/Maruyama]
Global and Regional Mapping of F-Peak Parameter
S. Shastri, S. Aggerwal and N.K. Sethi (NPL, India) compared the IRI/CCIR maps for the F peak critical frequency foF2 with several years of data from three Indian stations stretching across the anomaly region (Kodaikanal, magnetic equator; Ahemadabad, anomaly crest; Dehli, outside anomaly). They find good agreement at low and medium solar activity but large discrepancies for K and A during high solar activity with IRI overestimating the measured foF2 (contrary to the trend observed in TEC!). This might be attributed to the movement of the location of the anomaly crests with solar activity (due to the increase in neutral winds) and to the insufficient representation of this effect in IRI.
Jiao Peinan and Wu Jian (Xinxiang, China) described the Chinese Reference Ionosphere (CRI) and the solar activity index Ic and the Asia-Oceania Region (AOR) F-peak mapping used in CRI. A regional model for F-peak parameters over the Indian subcontinent was presented by S. Aggerwal (NPL, India). P. Bhuyan (Dibrugarh U., India) used data from four Indian stations to establish a model for the 75ĦE meridian. Comparisons with IRI highlight the advantages of these regional models/maps. Inclusion of some of the regional maps presented during this and the previous IRI workshop are planned for a future edition of the IRI model. [Inclusion of PRIME maps: Mikhailov]
An empirical model for TEC at Luping, Taiwan was presented by S. Jain (Bareli, India). The low-latitude electron density and temperature in the height range 500 to 2500 km was modeled based on data from the ACTIVE satellite (L. Triskova, Praha, Czech Rep.).
Topside and Plasmasphere
V. Pandey and N. Sethi (NPL, India) find that topside data from the Arecibo incoherent scatter radar are consistently below IRI predictions. This is in contrast to the equatorial ionosphere where IRI has been shown to underestimate the topside densities measured by the radar at Jicamarca, Peru. K. Oyama et al. (ISAS, Japan) further pursued the mapping of OHZORA electron temperature and density data at the satellite altitude of 600 km. I. Kimura et al. (Kyoto U., Japan) described ways of connecting their plasmaspheric AKEBONO-based model with IRI. [Continuing efforts: Kimura, Oyama, Bilitza]
S. Gupta (Ahmedabad, India) gave an overview of Indian rocket measurements and comparisons with IRI. Electron density profiles obtained from rockets launched from Thumba and SHAR in India were discussed and interpreted by S. Prakash (Ahmedabad, India). A.P. Mitra (CSIR, India) reviewed Indian measurements and modeling activities in the region below 100 km down to stratospheric and tropospheric altitudes. Rocket data are particularly important for the improvement of IRI in the E- and D-regions. Unfortunately only a rather limited set of measurements exists for low latitudes. A. Danilov (IAG, Russia) compared the IRI ion composition with measurements from rockets launched from Thumba, India and from Russian research vessels and suggested appropriate low-latitude modifications of the IRI model. At last year's IRI meeting it was decided to include in IRI the D-region models of Friedrich (Graz, Austria) and A. Danilov (IAG, Russia); incorporation of these two new options is currently underway. [D-region team: Friedrich, Danilov, Kopp, Singer]
It was suggested (and accepted) to allow input of measured M(3000)F2 values if available for computation of the F2 peak height hmF2.
P. Pasricha, S. Aggerwal and S. Shastri (NPL, India) pointed out that traditional RMS-type descriptions of ionospheric variability underestimate the actual hour-to-hour, day-to-day, or month-to-month variability due the non-random nature of foF2 variations.
A. Mikhailov (IAG, Russia) discussed equatorial F-region strom effects and their dependence on changes in the nuetral gas composition and the ExB vertical ion drift.
T. Gulyaeva (IZMIRAN, Russia) further explained the ionospheric disturbance index developed for the European PRIME project and compared indices obtained from high and mid-latitude stations with those from low-latitude stations.
I. Stanislawska (SRC, Warsaw, Poland) described a Kriging method that can be used for the regional mapping of ionospheric parameters.