ICTP'98 IRI Task Force Activity

The IRI Task Force Activity was held for the fifth time at the
International Center for Theoretical Physics in Trieste, Italy from June
22 to 26, 1998. The participants were: J. Adeniyi (U. Ilorin, Nigeria), 
D. Bilitza (NSSDC/RSTX, USA), D. Buresova (Prague, Czech Rep.), R. Ezquer
(Tucuman, Argentina), R. Hanbaba (CNET, France), M. Hernandez-Pajares (U.
Barcelona, Spain), A. Kumluca (U. Ankara, Turkey) B. Lazo (Havana, Cuba),
R. Leitinger (U. Graz, Austria), M. Mosert de Gonzalez (San Juan,
Argentina), O. Obrou (U. Cocody, Ivory Coast), S. Pulinets (IZMIRAN,
Russia), S. Radicella (ICTP), B. Reinisch (UML, USA).

As was done during previous meetings the team focussed on a specific
modelling problem each day with presentations and discussions in the 
morning and with smaller subgroups trying to resolve specific aspects of
the model problem with the help of computers and networks. 

1. The IRI Model in the Intermediate Region

The height of the F1 point is found in IRI as the point where the F2
bottomside profile reaches the F1 density predicted by the DuCharme et
al. model. The IRI model adds a special F1-layer function if the
existence of an F1-layer is predicted. A parabolic merging procedure is
than applied to pull the end of the F1-layer function down to the top of
the E-valley.

	B. Reinisch presented a new redefined IRI model for the F1 region.
The mathematical approach is similar to the current IRI model but the new
method avoids two important shortcomings of the earlier IRI model:

(A) Users had found that under certain circumstances the IRI electron
density profile will exhibit a small valley right below the F1 point. A
feature not seen in measurements. This is an artifact of the current IRI
F1 region formula. This is a particularly anoying problem for IRI users
who want to use IRI for ray tracing studies. The new formula proposed
during this TFA does not produce such a valley. 

(B) The merging of the bottomside function with the E-valley function
can only be accomplished if the F1-region function reaches down to the
E-layer densities. Quite often this is not the case and special fitting 
procedures have to be applied, e.g. changing the B1 and B0 bottomside
parameters. The newly proposed F1-layer function was chosen such
that it will always reach down to E-peak densities and thus avoids the
merging problems of the earlier function.

	Introduction of the new formalism into IRI will be straight
forward but will require a model for the new F1-region parameter D1. D1
describes the shape of the F1-region profile. B. Reinisch presented first
diurnal plots of D1. The variation of D1 with latitude and solar activity
need to be studied further.

	Discussions focussed on the question on how to introduce
additional anchor points, like the N170 (electron density at 170 km) point
and the F1 region inflection point. The current IRI model and the new
proposed F1-region function do not allow for the inclusion of such an
anchor point. The team will continue studying ways to introduce more
experimetal evidence into the description of the intermediate region
(this is the region that merges the F1-region with the E-valley region.
	A good test of the new F1-region model will be the comparison with
the N170 and inflection point values and models that have been established 
during earlier IRI/ICTP Task Force Activities. 

2. Topside Profile and Electron Content

A number of groups are in the process of developing new models for the
topside lectron density profiles. All of these efforts are of great
interest for IRI. S. Pulinets tested Chapman and Epstein functions to
represent Intercosmos-19 topside sounder profiles. The Epstein approach
uses the Epstein layer function in conjunction with a variable scale
height; this approach was first introduced by Radicella et al. (?). 
Pulinets reported best results with the Epstein profile function.

	An interesting data set from Havana, Cuba was provided by B. Lazo. 
It includes ground ionosonde data together with Intercosmos-19 topside
data and total electron content (TEC) data all for the same location. This
data set should help to resolve issues of wheather IRI shortcomings are
related to the F2 peak parameters or the topside and bottomside profile
shape parameters.

	R. Leitinger reported about the status of the magnetic field
aligned topside and plasmasphere model developed by his group at the
University of Graz together with Titheridge (New Zealand). Testing with
insitu measurements was suggested as the next step in evaluating this
modelling approach. A good data source for testing would be the Japanese
Hinotori data. R. Ezquer showed comparisons of the Hinotori data at 600 km
with IRI that indicated large discrepancies (underestimation by a factor
of 2 and more) during daytime in the equatorial zone.

	The potential of deducing ionospheric information from GPS
electron content measurements were discussed by M. Hernandez-Pajares.
Using IRI as pseudo ground truth he tested the reliability of his global
mapping procedure. This is clearly a good way to test the ability of the
various methods that have been established to reconstruct global snapshots
of the ionospheric electron content from GPS slant measurements. A
recommendation will be send to the IGS Iono Working Group. The method
showed that quite often the limited ground station coverage at low
latitudes did not allow to resolve the equator anomaly crests. Adding a LEO
GPS transmitter (e.g. GPS/MET) resulted in a much better spatial
3. Bottomside Parameters B0, B1

New bottomside parameters were provided by M. Mosert for Tucuman, San
Juan, and Buonos Aires, by J. Adeniyi for Ougadougou and Ibadan, by 
D. Burosova for Pruhonice and by O. Obrou for Korhogo. Based on these data
representative B0 and B1 values were established for three latitude zones,
for day and night, for high and low solar activity and for all seasons.

	This new "Table-option" will be presented at the upcoming IRI 
meeting in Nagoya, Japan with the proposal of replacing the old
"Table-option" with this new one. This will result in a major improvement
of the IRI bottomside profile at low and equatorial latitudes, since the
current IRI does not include values for the equatorial region. 

	The bottomside shape parameter did not show large variations with
either season nor solar activity. It was decided to use constant values
until more detailled studies can be made. In almost all cases were the
experimetal B1 values below the constant value (=3) used in the current
IRI bottomside function.

	More data are needed from stations at the magnetic equator
especially during high solar activity. An effort will be made to obtain B0
and B1 values from Jicamarca ionosonde as well as incoherent scatter data.

4. General

The papers presented during the 1998 TFA will be published as an ICTP
report similar to the reports that have been prepared after the 1995, 1996
and 1997 TFAs. These reports provide a good opportunity to publish plots
and tables of data relevant to the TFA work. All participants were again
strongly encouraged to take advandage of this opportunity, which is also a
good way of documenting our progress in improving the IRI bottomside
prfofile. The final version of the papers should be sent to Sandro
Radcilla by November 1, 1998.

	The date for the next TFA was set for June 21 to 25, 1999. As a 
new topic we will include studies to better represent the equator anomaly
region in the IRI model. For this purpose we will try to have some review
talks about the current state of our understanding and modelling of the
equator anomaly region. Of special interest will be the data and results
from the International Equatorial Electrojet Year (IEEY) 1992-94, an
initiative of IAGA's Inter-Commission on Developing Countries (ICDC)
headed by Luiz Barreto (Argentina). Dr. Abdu and Paul Vila were mentioned
as collaegues who are knowledgable about IEEY. A coming issue of Annales
Geophysicae will be dedicated to the IEEY results.

	The UML team has developed a special analytical version of the IRI
electron density model for ray tracing studies. Such a version is 
important for many propagation application of IRI, e.g. OHT radars. Bodo
Reinisch will explain the buildup of this special version during the next

5. Work to be Done in Preparation for the next TFA

Bodo Reinisch will provided the program with the new F1-region
representation. This program will fit profiles with the new formulas and
thus obtain the D1 parameter in addition to B0 and B1. D1 is the parameter
that descibes the F1-layer shape. A data base of D1 values is the first
step towards a description of the variations of this new parameter with
local time, latitude, and solar activity. 

	Studies will continue concerning the global mapping of the
parameters N170 (electron density at 170 km) and hF1 (inflection point in
the F1 region). These parameters will be excellent for testing the new IRI
model in the F1-region. 

	More data are needed for B0 and B1 from equatorial stations. 
Jicamarca ionosonde and incoherent scatter data offer the most promise and
will be pursuit. B0 and B1 tables based on this year's TFA will be
presented at the upcoming IRI session during the COSPAR meeting in Nagoya,
Japan in mid-July. The IRI program has been already modified to accept
the new parameters for the equatorial zone.   

	Statistical studies of the variation trends and patterns of D1, B0
and B1 should also include an indication of the standard deviation (either
as scatter plots or as sd bars) since it is important to verify that a
certain trend exceeds the sd limits; e.g, it was found that the diurnal
variation of B0 observed at Prague exhibits a maximum at night and a
minimum during daytime, contrary to the diurnal variaiton pattern at most
other stations.   

	The next phase in topside modelling will be the comparison with
in-situ measurements. Several of the new approaches have now matured far
enough to be considered for inclusion in a new version of IRI. The testing
with experimental evidences is essential for selecting the best new model.

* Back to IRI workshops page
* Back to IRI homepage