IRI'98 Workshop Report
COSPAR'98 Session C4.1 Lower Ionosphere: Measurements and Models
Session Report / Dieter Bilitza
This session was organized by the COSPAR/URSI interunion Working Group on the
International Reference Ionosphere (IRI) and was held on the first three days
of the 32nd COSPAR Scientific Assembly at the Congress Center in Nagoya, Japan.
This session was co-sponsored by the International Union of Radioscience (URSI).
The main topic of this year's IRI session was the Lower Ionosphere (the D and
E region; altitude range from about 50 to 150 km). A total of 46 papers were
presented during this session including 17 invited, 12 contributed, and 17
poster contributions. The different IRI half-day sessions dealt with the
following topics: D Region Modelling 1 and 2, D Region Data, IRI Improvements
- Middle Ionosphere, New Models and Data for IRI 1 and 2, Lower Ionosphere
- Data, Ion Composition, and Posters.
The first day provided a good overview over the different ongoing modelling
efforts that could benefit an improved IRI model in this region. The next
version of IRI will include results from two of these efforts as new options
for the electron density in the D-region: (1) the neutral density dependent
model of Friedrich et al. (Graz, Austria) based on a large number of rocket
measurements that now also includes a high-latitude part; and (2) the model of
Danilov et al. (Moscow, Russia) that includes representative profiles for
conditions of winter anomaly and of stratospheric warming. Discrepancies
still exist between these different models primarily because of the relative
data sarcity in this region that is difficult to reach by satellite insitu
instruments as well as direct measurements from the ground. Including the
two options in addition to the current IRI model will hopefully stimulate
comparisons with more data and eventually help to select the best model.
Results from theoretical models as presented by Kopp (Bern, Switzerland) and
by Turunen (Sodankylae, Finland) are important to bridge the data gaps and
study global, seasonal and solar cycle trends. Theoretical models are also
an excellent tool for exploring the role of various chemical reactions and
neutral constituents, e.g. the role of the minor constituent nitric oxide.
Kopp finds that switching from the NCAR model to the UARS/HALOE model for NO
results in an oder of magnitude difference in the resulting electron density.
MF differential absorption measurements obtained by ground radar (Igarashi et
al., Japan) and during a rocket flight (Nagano and Okada, Japan) as well as
Omega navigation signals from the Akebono satellite (Miyamura et al., Japan)
were used to deduce D region electron density. Comparisons with IRI highlighted
the strength and weakness of the current IRI model.
The E-peak is, of course, a very important point for lower ionosphere
modeling since it is generally the point of highest ionization in this
region. Presentations during this workshop compared peak data obtained by
ionosonde (Mosert, Argentina) and by incoherent scatter radar (Pandey, India)
with the IRI predictions. Good agreement was in general found for the peak
density and improvements were suggested for the description of the diurnal
variation of the peak height hmE. Based on the excellent predictability of
peak densities Nusinov et al. (Russia) recommend using the long record of E
peak ionosonde observations and a simple model to deduce the solar cycle
variation of solar EUV and soft X-ray fluxes.
The prime data source for the lower ionosphere are rocket and ground
(ionosonde, radar) measurements. Gupta (India) and Chakrabarty (India)
reminded the group that a considerable amount of low-latitude D and E
region electron density data are available from rockets launched from
Thumba over the last decades. A data base of ion composition data obtained
by rockets was presented by Grebowsky and Bilitza (USA) comprising of all
available published sounding rocket observations. EISCAT incoherent scatter
data were used by a number of presenters. Fujii (Japan) used the many
parameters that can be deduced from the EISCAT measurements to study the
energy coupling between the ionosphere thermosphere and magentosphere.
Shibata et al (Japan) and Kofman et al. (France) deduced the ion composition
from the EISCAT data and investigated the variation of the transition height
from oxygen ions to molecular ions. Their model will be helpful in the
redesign of the IRI ion composition model which will be based on the
transition heights as the characteristic points.
Great progress has been made in improving the accuracy of the IRI
electron density predictions from the F peak down to the top of the E-valley.
A special IRI Task Force Activity organized by Radicella (Italy/Argentina)
at the International Center for Theoretical Physics (ICTP) in Trieste, Italy
has produced a number of new model inputs. Bilitza et al. (USA) presented the
results pertaining to the bottomside thickness and shape parameters B0 and B1.
New ionosonde data especially from low and equatorial latitudes have led to
improvements of the IRI model of in some cases more than 30%. During this
meeting new results for B0 and B1 were presented based on parameters deduced
from incoherent scatter measurements from Arecibo (Sethi and Mahajan, India)
and from the Japanese MU radar (Zhang et al., China/Japan). Reinisch (USA)
presented another result of the task force effort, a better scheme for
connecting the bottomside to the valley top avoiding some of the pitfalls of
the approach that is currently used in IRI.
Fuller-Rowell et al. (USA) are continuing their efforts to develop an empirical
model for the F-region strom-time changes in electron density. He reported good
success in describing the negative storm phase and to a lesser extent for the
positive phase. His model is now scheduled to be included with the next version
of IRI. Watanabe et al. (Japan) studied the low-latitude response to magnetic
disturbances using Hinotori satellite data and the Fejer&Scherliess disturbance
plasma drift model. It was again noted that IRI should include a plasma
drift model and that the Fejer&Scherliess model would be the best candidate.
Single-station models and tests were presented based on ionosonde data from
Istambul (Ozguc and Tulunay, Turkey) and from Warsaw (Stanislawska, Poland).
Theoretical models were combined with ground-based data (Balan, Brazil and
Fukao, Japan) and with satellite data (Su and Bailey, U.K., and Oyama, Japan)
to investigate annual and seasonal variations of ionospheric parameters.
Variation of the equinoctial assymetry with altitude and of the winter
summer difference could be partly explained by the variations of neutral
composition and winds. Ionosonde data from Fortaleza and Cachoeira Paulista
(both in Brazil) were compared with the results from the Sheffield model
to determine the equatorial vertical plasma drift and the thermospheric
merdional wind (Souza and Abdu, Brazil, and Bailey, U.K.).
Data from the Intercosmos satellites IK 19, IK 24 and IK 25 were used to
construct empirical models of electron density and temperature highlighting
the discrepancies to IRI (Truhlik et al. Czech Republic). The same team also
established model descriptions for the ion composition measurements made by
the IK 24 ion mass spectrometer from 1989 to 1991. Iwamoto et al. (Japan)
did a statistical analysis of ion mass spectrometer data from ISS-b (1978-
1981) and Akebono (1989-present) and discussed the solar cycle correlation
of the different ion densities and implications for IRI.
All of these empirical models are an important new source for improvements
of the IRI model in the topside. Shortcomings of the present IRI topside
model were noted in comparisons with Hinotori data (Ezquer, Argentina) and
with ISIS sounder data (Bilitza, USA).
IRI Group Specifics:
Several new members were accepted into the IRI Working Group: Peter Dyson,
LaTrobe University, Melbourne, Australia; Iwona Stanislawska, Space Research
Center, Warsaw, Poland, and Tim Fuller-Rowell, SEC, NOAA, Boulder, Colorado.
Accepting an invitation by Bodo Reinisch (UML) the Working Group decided to
hold the 1999 IRI Workshop at the University of Massachusetts in Lowell. The
date was set to August 9 to 12 which is the time period between the Digisonde
Seminar at UML and the URSI General Assembly in Toronto, Canada (August 13
to 21, 1999). The Workshop topic will be quantitative descriptions of
ionospheric variability and analytical models for ray tracing. Selected
papers of the 1996 IRI COSPAR session are now published in Advances in Space
Research, Volume 20, Number 9. The final set of papers from the 1997 IRI
Workshop was submitted to Advances in Space Research in April and publication
is expected soon.
Back to IRI workshops page
Back to IRI homepage