Usage of the Internet Resources for Research of the Ionosphere and
the Determination of Radio Wave Propagation Conditions
Olga Maltseva
Institute of Physics Southern Federal University, Stachki, 194, Rostov-on-Don, Russia
mal@ip.rsu.ru
Keywords: GPS. Total electron content TEC. Ionospheric models. Radio wave propagation. Geomagnetic disturbances.
Abstract: In the field of ionospheric research is impossible to obtain the new results, new knowledge without use of
Internet resources. The paper provides examples of the use of these resources on the basis of generalization
and complement to reports made on the three previous ICTRS conferences. Results are presented in four
areas of possible data: (1) vertical sounding, (2) total electron content ТЕС, measured by high orbit
navigation satellites, (3) plasma frequencies measured by low orbit satellites, (4) empirical models of the
ionosphere. The main achievement in the first direction is the creation of GAMBIT, designed to provide
global maps of ionospheric parameters foF2 and hmF2 with delay of 15 minutes in relation to real time. The
estimation of conformity of the IRI model to experimental data of foF2 for high-latitude station of southern
hemisphere is made. Within the second direction the effectiveness coefficient of use of the observational
median of the equivalent slab thickness in comparison with thickness of the IRI model is introduced. It is
shown that this coefficient almost always exceeds 1, reaching values 1.5-2 globally. Behavior features of
deviations of calculated foF2 from observational values during the strongest geomagnetic disturbances of
April 2014 and March 2015 are given. In the framework of the third direction validation of a plasmaspheric
part of a N(h)-profile according to satellite IMAGE data is performed. It was concluded that to disambiguate
N(h)-profiles it is necessary to improve both values of ТЕС, and the shape of the topside part because the
plasmaspheric part is close to existing model RPI. Within the fourth direction the statistics of comparisons
of various models was increased including high-latitude region of the southern hemisphere.
1 INTRODUCTION
Internet affects the lives of every person, providing
huge opportunities in the information sphere. This
paper describes the opportunities associated with
databases for ionospheric research. To solve many
scientific and technological problems need to know
the conditions in the ionosphere, within which there
are the satellites that provide us, in turn, information
about the state of the ionosphere. This article points
out the most important databases and displays the
results of their use. Possibilities of use of Internet
data are illustrated with examples: (1) vertical
sounding (VS), (2) global maps of the total electron
content of the ionosphere TEC, (3) data satellites
CHAMP and DMSP, and (4) the International
Reference Ionosphere model IRI. Each of these
areas corresponds to a separate section (2-5).
2 INTERNET POSSIBILITIES OF
USAGE OF THE VERTICAL
SOUNDING DATA
In the 20th century, the key parameter of the
ionosphere was the peak concentration NmF2 (or
critical frequency foF2), measured by special
receiver-transmitters - ionosondes. By the end of the
20th century, digisondes, being ionosondes with
automatic processing of samples, have been
appeared when the need of the ionospheric data in
real time and on a global scale has required to
automate this process. A lot of programs have been
created. System ARTIST for which now already the
fifth version (Reinisch et al., 2009) was developed is
most widely used. However, as shown by means of
special expert program QualScan estimating quality
of ionograms, usually 1/3 of digitized ionograms can
be rejected (McNamara, 2006). The most promising
is a new principle, which underlies dynasonde
7
Olga M.
Usage of the Internet Resources for Research of the Ionosphere and the Determination of Radio Wave Propagation Conditions.
DOI: 10.5220/0005888700070017
In Proceedings of the Fourth International Conference on Telecommunications and Remote Sensing (ICTRS 2015), pages 7-17
ISBN: 978-989-758-152-6
Copyright
c
2015 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
(Zabotin et al., 2005) and provides the most accurate
determination of the parameters of the ionosphere.
The evaluation of improvements of ionospheric
parameters determination by dynasonde in
comparison with digisonde (ionosonde) was held in
the paper (Maltseva et al., 2007). The most
important are results for foF2 and hmF2. Figure 1
from this paper is an example of difference between
automatic methods of determination of daily
dependence of frequency foF2 by various methods
of sounding: digisonde (method POLAN (P)) and
dynasonde (method NeXtYZ (N) (Zabotin et al.,
2005)) in comparison with the IRI model.
Figure 1: An example of differences between methods of
foF2 determination: digisonde (method POLAN (P)) and
dynasonde (method NeXtYZ (N)).
Table 1 shows the quantitative evaluation of
absolute differences between critical frequencies
obtained by these methods and the values of the IRI
model. Results are shown for stations having
dynasonde and for those days which had
simultaneous measurements. The first column
contains name of stations. The second column shows
the month. The third column points out number of
day when there were measurements. P-IRI column
shows the difference between the method POLAN
and the IRI model, N-IRI shows the difference
between the method NeXtYZ and the IRI model, N-
P column shows the difference between the methods
NeXtYZ and POLAN. Values of P-IRI and N-IRI
are an estimate of improving the determination of
foF2 using experimental data. Values N-P are an
estimate of improving the determination of foF2 by
method NeXtYZ (dynasonde) compared with
method POLAN (digisonde). If you take the average
value of foF2 approximately 5 MHz, the
improvement is 11-14%.
Table 1: Improvement of foF2 determination by new
dynasonde.
station
month
Num-
ber of
days
׀ΔfoF2׀, MHz
P-IRI
N-IRI
N-P
Lycksele
July
16
0.64
0.62
0.16
Tromso
July
16
0.57
0.59
0.19
BearLake
July
5
0.74
0.72
0.18
Tromso
Nove
3
0.6
0.6
0.10
Lycksele
April
5
0.83
0.76
0.14
Tromso
April
6
0.69
0.66
0.17
Figure 2 shows the difference between automatic
methods of determining the values of hmF2,
calculated by different methods of sounding:
digisonde (method POLAN (P)) and dynasonde
(method NeXtYZ (N)). Also values of hmF2 for the
IRI model are given.
Figure 2: An example of differences between values of
hmF2 obtained by various methods of sounding: digisonde
(method POLAN (P)) and dynasonde (method NeXtYZ
(N)).
Table 2 shows the quantitative evaluation of
these differences (Maltseva et al., 2007). P-IRI
column shows the difference between the method
POLAN and the IRI model, N-IRI shows the
difference between the method NeXtYZ and the IRI
model, N-P column shows the difference between
the methods NeXtYZ and POLAN.
As well as in a case with parameter foF2, values
P-IRI and N-IRI characterize improvement in
adapting the model to the experimental values of
hmF2. N-P values have improved the definition of
the maximum height by the new method. These
values are significantly higher than the accuracy of
measurement. Using as the mean value of hmF2 250
km we obtain approximately 15% improvement.
8
Table 2: Improvement of hmF2 determination by new
dynasonde.
station
month
Numb-
er of
days
׀ΔhmF2׀, км
P-IRI
N-IRI
N-P
Lycksele
July
16
40.6
38.5
34.3
Tromso
July
16
48.5
44.7
29.2
Bea
Lake
July
5
47.8
53.2
25.2
Lake
Novem
3
49.2
40.5
33.7
Tromso
April
5
20.3
20.5
24.5
Lycksele
April
6
29.5
19.2
29.7
Unfortunately, the stations equipped by
dynasonde are not a lot. Data of other stations are
collected in the form of multiple databases (GIRO,
SPIDR, DIAS et al.) which are freely available,
include data from several cycles of solar activity,
and are permanently being updated and modified. A
large contribution to this update provide reports
from conference (URSI, 2015), in particular, the
system GIRO was modified in GAMBIT
(https://git.giro.uml.edu), which represents the
global maps of foF2 and hmF2 delayed at least 15
minutes, compared with real time. The European
System of DIAS was supplemented with the data of
such stations as Moscow and Tromso.
3 INTERNET POSSIBILITIES OF
USAGE OF THE TOTAL
ELECTRON CONTENT OF THE
IONOSPHERE
Despite the fact that, apparently, the method of the
VS can be considered the best method for measuring
the ionospheric parameters, its main deficiencies are
rare network and quite often data gaps on some
stations. In the 21st century, with the development
of network of navigation satellites has been a shift to
the identification of the total electron content (TEC)
of the ionosphere as the main parameter, and even
proposed to replace the parameter NmF2. The
advantages of this option is its continuous
monitoring, the availability of online databases of
global maps for a period longer than the cycle of
solar activity, information about the N(h)-profile. In
(Maltseva et al., 2012; 2013; Maltseva, 2014) the
TEC parameter was used to determine foF2 with
experimental median of ionospheric equivalent slab
thickness τ(med). Much more conformity of the
calculated values of foF2 with measurements in
comparison with conventionally used value τ(IRI)
(McNamara, 1985; Houminer, Soicher, 1996;
Gulyaeva, 2011) was obtained. The improvement is
achieved by taking into account a big difference
between what was (τ(IRI) = TEC(IRI)/NmF2(IRI),
NmF2(calc) = TEC(obs)/τ(IRI)), and what became
(τ(med) = med(TEC(obs)/NmF2(obs)), NmF2(calc)
= TEC(obs)/τ(med)). Illustration of this difference
for the reference station Juliusruh is shown in Fig. 3.
In this paper we introduce the efficiency factor Keff
of using τ(med) compared with τ(IRI), and give the
results of its determination for the individual stations
in the various regions of the world and globally.
Keff factor is defined as the quotient of the absolute
deviations |ΔfoF2| of calculated foF2 from the
experimental values using these two parameters
τ(IRI) and τ(med). Fig. 4 shows the coefficients for
the 6 stations: the European mid latitude Juliusruh
station, the American mid latitude Goosebay station,
the American high-latitude station Thule in the
northern hemisphere, the equatorial Ascension
Island station, the low-latitude and high-latitude
Grahamstown and Mawson stations of the Southern
hemisphere. For comparison, the value of K=1 is
displayed. Integrated characteristics are given in
Table 3.
Figure 3: Differences between used values τ(IRI) and
τ(med) on an example of the long-year data for July and
December and the Juliusruh station.
.
Usage of the Internet Resources for Research of the Ionosphere and the Determination of Radio Wave
Propagation Conditions
9
Figure 4: Coefficients Keff in various regions of the globe.
Table 3: Conformity averaging for all years between the calculated and observed values of foF2: coefficients Keff and the
relative deviations σ (foF2), %
σ(foF2), %
July
December
station
Keff
med
ins
τ
Keff
med
ins
τ
Juliusruh
2.52
9.82
15.03
6.91
2.20
13.30
17.68
8.45
Goosebay
2.03
9.41
15.46
9.36
1.72
13.54
17.61
11.36
Thule
2.03
11.46
15.43
8.46
1.34
10.41
18.21
13.91
Grahams
1.70
9.53
15.84
10.09
1.99
12.44
17.26
9.40
Ascension
1.98
23.40
27.21
14.89
1.57
13.04
17.52
12.45
Mawson
2.26
44.26
48.83
25.7
1.36
15.28
24.82
20.75
It can be seen that the efficiency factor is close to
2 for July, i.e. results are in 2 times better, and near
1.5 for December. The same can be said about the
relative deviation σ(foF2),%. With the exception of
the Antarctic station Mawson deviations for τ(med)
lie within 10%. Keff has been calculated globally for
months with disturbances. Figure 5 shows the
average monthly variations depending on the
latitude for τ(IRI) (green triangles). For τ(med) red
dots show results for the global map JPL. Figure 5
also includes the efficiency factors. Blue circles in
all Figures show results of comparison of model IRI
values with experimental medians for an illustration
of accuracy of the model in relation to measurement.
10
Figure 5: Behavior of Keff on a global scale.
Keff is close to 2 around the globe, and σ is less
than 10% with the exception of the stations of the
Southern hemisphere. This indicates that the
simulation situation for the southern hemisphere has
problems. For these months, the details of the
comparison of deviations have been identified in
disturbed periods. Fig. 6 shows the Dst-index for
these months (http://wdc.kugi.kyoto-u.ac.jp/dstdir/).
Figure 6: Behavior of Dst-indexes in months with
disturbances
The behavior of Dst-index specifies which days
had the greatest disturbances. In April 2014 this day
is 12. In March 2015 these are 17-19 because
disturbance was stronger. Fig. 7 presents the global
results for specific days. In both cases values are
also given for the quiet days preceding the
disturbances. Two top figures show the results for
τ(IRI), the lower - for τ(med), and these charts
include in addition the values for τ(IRI), averaged
over the 5 days considered.
We see that deviations for τ(IRI) are maximum
in disturbed days. In the southern hemisphere still
get big deviations also for τ (med). To conclude this
Section it is necessary to say a few words about the
alternatives. In recent years, assimilation techniques
become widespread (e.g. Khattatov et al., 2005;
Fuller-Rowell et al., 2006), in which N(h) -profile of
the ionosphere is fitted by the very powerful
methods so that it satisfies the measured TEC(obs),
but emphasis on the determination of foF2 was not
done. And when compared with measurements by
ionosondes large discrepancies are obtained, which,
for example, are shown in Fig. 8 from (Yao et al.,
2014). Differences are seen not only in critical
frequencies, but also in a profile of the topside part
although in general density distribution throughout
the profile provides conformity with the TEC(obs)
for the account of the plasmaspheric part.
Usage of the Internet Resources for Research of the Ionosphere and the Determination of Radio Wave
Propagation Conditions
11
Figure 7: Details of deviation comparison in the disturbed
periods.
Figure 8: Conformity of assimilation profiles with the data
of vertical sounding (from Yao et al., 2014).
4 INTERNET POSSIBILITIES OF
USAGE OF THE LOW ORBIT
SATELLITES DATA
By the end of the 20th century, data from different
satellites began to play a larger role in the study of
the ionosphere and radio propagation in it. One of
the achievements of the 21st century is a creation of
database of satellite CHAMP (http://isdc.gfz-
potsdam.de/), which includes the value of the
electron density at altitudes near the maximum
height hmF2. Currently, this array also covers
virtually one cycle of solar activity (2000-2011).
Even earlier, similar data were provided by a series
of satellites DMSP (http:// cindispace. utdallas.edu/
DMSP/ dmsp_data_at_utdallas.html). Since 2006,
the huge amount of data was obtained by satellite
COSMIC (http://tacc.cwb.gov.tw/en/). One of the
goals of the simulation is to coordinate the model
N(h)-profile and experimental TEC(obs). The IRI-
Plas model (Gulyaeva, 2011) provides adaptation to
TEC(obs), but the obtained profiles do not always
ensure the coincidence with the plasma frequency,
measured at the low orbit satellites. As shown in
(Maltseva, 2014), adapting the model to the plasma
frequency of one or two satellites gives values
TEC(sat), not equal to TEC(obs). In this paper, the
12
difference Δ(TEC(sat)) = TEC(obs)-TEC(sat) was
attributed to a plasmaspheric part of the profile and
introduced coefficient K(PL), modifying this part of
the profile so that the TEC of the modified profile is
equal to TEC(obs). This factor is a multiplier, which
the density of the profile at the upper limit (20,000
km) is multiplied by. Examples of profiles of the
station Juliusruh in April 2001 have been shown that
if the difference Δ(TEC(sat)) is positive then the
coefficient K(PL) may be greater or less than 1.
When K(PL)> 1 the profile may have nonphysical
shape. When K(PL)<0 it is necessary to suppose
density on the upper limit equal to zero, since it
cannot be negative. TEC for such a profile may
differ from the TEC(obs). This indicates a need for
testing a plasmaspheric part of the profile. To do
this, in this paper the model of (Ozhogin et al.,
2012), based on the experimental data RPI (Radio
Plasma Imager) of satellite IMAGE, was used. This
model was developed in a range of L-shells from 1.6
to 4, so this test cannot be carried out for the station
Juliusruh, as its L-shell does not fall in this range.
Example of test is given for the station Ascension
Island.
Figure 9: Comparison of values of ТЕС for various
options of calculation and corresponding coefficients
K(PL).
This example makes it possible to show the
uncertainty of profiles associated with both the
shape of the profile, and with the uncertainty of
value of TEC(obs), given by various global maps.
Comparison of TEC for that station for the
calculation of various options is given in Fig. 9
during moments of flight of the satellite CHAMP
over the station. The horizontal axis represents the
days of the month. Time of observation is shown at
the top of the graph. Red dots on the top panels
show the observed values, the black dots are the
values for the original (without adaptation) model
IRI. Green triangles belong to profiles of models
adapted to foF2(obs) and fne(sat2), violet asterisks
indicate the TEC for the profile, which is the starting
point for modifications. Orange circles show the
TEC for profile modified by the coefficient K(PL).
The lower panel shows the corresponding
coefficients K(PL).
This situation is typical for all the stations, only
values of the difference between the TEC may vary
depending on the latitude. Interestingly, the N(h)-
profiles passing through fne(sat), give more lower
values TEC(sat) than TEC(obs). The modification
results in full compliance with the TEC(obs) using
K(PL) from the lower panel. Fig. 10 shows the N(h)-
profiles for one of the cases (April 12, UT = 3).
Profile is represented for clarity in two parts: from
the beginning of the ionosphere to the height of
2,000 km and from an altitude 2,000 km to an
altitude of 20,000 km.
Fig. 10 includes the profile of the original model
(black dots), for which the TEC(IRI) exceeds the
TEC(obs), the profile calculated for a model adapted
to foF2(obs) and TEC(obs) and shown by purple
diamonds (mark "All"), profiles passing through
fne(CHAMP) and foF2(obs). These profiles cannot
always build. They are shown with blue crosses, and
there are only in two panels. Profiles going through
foF2(obs) and fne(DMSP) are shown by green
triangles. They are built for all cases. Large red dots
show fne(sat) on the top panels. Small red dots show
the values of the model RPI on the bottom panels. A
plasmaspheric part of all profiles is close to the
model of RPI. Profile s12+PL, providing
TEC(JPL)=44 TECU, in the upper panel, does not
pass through the plasma frequency. The second
panel shows the profiles for UPC map with
TEC(UPC) = 31.8 TECU. In the profile "All", the
topside part was decreased, but the plasmaspheric
part was increased. It can be seen that the profile s12
+ PL passes through fne(DMSP), but again there is
no passage through fne(CHAMP).
Usage of the Internet Resources for Research of the Ionosphere and the Determination of Radio Wave
Propagation Conditions
13
Figure 10: An illustration of conformity between model N(h)-profiles and observed ТЕС.
We attempted to modify the profile in two
directions: to increase foF2(obs) (third panel) or
hmF2 (fourth panel) in order to pass the profile
through fne(CHAMP). In both cases, the profile has
gone through fne(CHAMP), but the TEC of these
profiles are greatly increased. In case of increase of
foF2(obs), difference Δ(TEC(sat)) was negative,
resulting in a zero value at the upper limit of profile.
In the second case, the topside part was not changed
much. This led to a positive difference and an
increase of the plasmaspheric density. These results
lead to the important conclusion that to disambiguate
N(h)-profile it is necessary to specify more precisely
the values of TEC and shape of the topside profile
because the plasmaspheric part is close to the
existing model.
.
5 INTERNET POSSIBILITIES OF
USAGE OF EMPIRICAL
MODELS OF THE
IONOSPHERE
Presence of huge array of data and approximation
methods has led to possibility of creating of
empirical models of the ionosphere which with
success are used for forecasting of its state. Among
set of empirical models which are disposed on the
Internet, the IRI model is one of the most demanded
and most dynamically developing
(http://irimodel.org). In this century it has undergone
some updates: IRI2001, IRI2007, 2010, 2012
(Bilitza, 2001; Bilitza, Reinisch, 2008; Bilitza et al.,
2014).
Main attention focuses on the correspondence of
model parameters to experimental values. In this
study, a comparison of the model with observed data
14
was held including data of stations which have
appeared on the Internet only recently (since 2013).
It allowed us to obtain results in a global scale (Fig.
5, 7).
Figure 11: An estimate of accuracy of the model IRI
according to data of the Antarctic station Mawson.
Of great interest is evaluation of the IRI model
possibilities in the high latitudes of the southern
hemisphere in connection with the project GRAPE
(De Franceschi, Candidi, 2012). In this work, the
first results were obtained according to data of the
station Mawson. Fig. 11 shows the absolute and
relative deviations of July and December for all the
years with available data. Black dots show
deviations from day to day, blue dots deviations of
medians and red dots deviations of foF2 values,
calculated with usage of the TEC and τ(med).
It is evident that in July (winter) deviations of
curves with marks “med” and “ins” are significantly
exceed the values typical for the northern
hemisphere. Using the TEC and τ(med) improves
the situation, except for 2004. Moreover, this Figure
does not include the values for 2005 in connection
with abnormal values of foF2, shown in Fig. 12
requiring additional study (the top panel). One
reason may be an insufficient quantity of data: the
lower panel shows the deviation |ΔfoF2| along with
the number of days reduced by 10 times.
Figure 12: An illustration of the abnormal behavior of
foF2 at the station Mawson in 2005.
In December results are comparable with the
values typical for the northern hemisphere.
In (Maltseva et al., 2012; 2013; Maltseva, 2014)
a lot of attention was paid to the comparison of
different empirical models of foF2 and TEC
(Gulyaeva, 2011; Hoque, Jakowski, 2011; Jakowski
et al., 2011). In addition to these results, a huge
volume of data has been used in (Maltseva et al.,
2014; 2015) for comparison, covering more than 10-
15 stations in different regions of the globe and
several years. Despite this, an unambiguous answer
was not obtained.
Great expectations are linked to a database of
satellites COSMIC. The COSMIC EDP data are
obtained from the COSMIC Data Analysis and
Archive Center (CDAAC) at University Corporation
for Atmospheric Research (UCAR). COSMIC RO
Usage of the Internet Resources for Research of the Ionosphere and the Determination of Radio Wave
Propagation Conditions
15
measurements and products (such as electron density
profiles) can be available from the Taiwan Analysis
Center for COSMIC (TACC, http://
tacc.cwb.gov.tw/en/) and the COSMIC Data
Analysis and Archive Center (CDAAC,
http://www.cosmic.ucar.edu/cdaac/).
6 CONCLUSIONS
Only a few examples illustrate the possibility of
obtaining new results using the resources of the
Internet. For ionospheric studies, the most important
are databases of the vertical sounding, measurements
of TEC by high-orbit navigation satellites,
measurements of plasma frequency by low-orbit
satellites, data of solar and geomagnetic conditions.
The obtained results show the large possibilities of
use of resources of the Internet for study of the
ionosphere and the representation of conditions of
propagation of radio-waves. A series of new
knowledge in this area was obtained.
But there is another side to such use: (1) it is not
always possible to reconcile different data to each
other because of their inaccuracies and ambiguities,
(2) there is a danger in the access closing of certain
countries, as it may be in an emergency situation.
ACKNOWLEDGEMENTS
The author thanks the scientists who provided the
data of SPIDR, satellites DMSP, global maps of
TEC, operation and modification of the IRI model,
Dr A. Karpachev (IZMIRAN, karp@izmiran.ru) for
CHAMP data, Southern Federal University for
finantial support (project N 213.01-11/2014-22).
REFERENCES
Bilitza, D., 2001. International Reference Ionosphere,
Radio Sci., 36(2), 261-275.
Bilitza, D., Reinisch, B.W., 2008. International Reference
Ionosphere 2007: Improvements and New Parameters.
Adv. Space Res., 42, 599-609.
Bilitza, D., Altadill, D., Zhang, Y., Mertens, C., Truhlik,
V., Richards, P., McKinnell, L.-A., Reinisch, B., 2014.
The International Reference Ionosphere 2012 a
model of international collaboration, J. Space Weather
Space Clim., 4, A07, 12p. DOI:
10.1051/swsc/2014004.
De Franceschi, G., Candidi, M., 2013. GRAPE, GNSS
Research and Application for Polar Environment,
Expert Group of SCAR, Annals of Geophysics, Special
Issue 56. 2. P0215; Edited by Giorgiana,
doi:10.4401/ag-6366.
Fuller-Rowell, T., Araujo-Pradere, E., Minter, C.,
Codrescu, M., Spencer, P., Robertson, D., Jacobson,
A.R., 2006. A new data assimilation product from the
Space Environment Center characterizing the
ionospheric total electron content using real-time GPS
data US-TEC, Radio Sci. 41, RS6003, doi:
10.1029/2005RS003393.
Gulyaeva, T.L., 2011. Storm time behavior of topside
scale height inferred from the ionosphere-
plasmosphere model driven by the F2 layer peak and
GPS-TEC observations. Adv. Space Res., 47, 913-920.
Hoque, M.M., Jakowski, N., 2011. A new global empirical
NmF2 model for operational use in radio systems.
Radio Sci., 46, RS6015, 1-13.
Houminer, Z., Soicher H., 1996. Improved short term
predictions of foF2 using GPS time delay
measurements. Radio Sci., 31(5), 1099-1108.
Jakowski, N., Hoque, M.M., Mayer, C., 2011. A new
global TEC model for estimating transionospheric
radio wave propagation errors. Journal of Geodesy,
85(12), 965-974.
Khattatov, B., Murphy, M., Gnedin, M., Sheffel, J.,
Adams, J., Cruickshank, B., Yudin, V., Fuller-Rowell,
T., Retterer, J. 2005. Ionospheric nowcasting via
assimilation of GPS measurements of ionospheric
electron content in a global physics-based time-
dependent model. Q.J.R.Meteorol. Soc., 131, 3543
3559.
Maltseva, O., 2014. Choice of the definition method for
the total electron content to describe the conditions in
the ionosphere. Proceedings of Third International
Conference on Telecommunications and Remote
Sensing, Luxemburg, Grand Duchy of Luxemburg, 26-
27 June 2014, 51-61.
Maltseva, O.A., Mozhaeva, N.S., Nikitenko, T.V., 2014.
Validation of the Neustrelitz Global Model according
to the low latitude ionosphere. Adv. Space Res.,54,
463472, http://dx.doi.org/10.1016/j.asr.2013.11.005.
Maltseva, O.A., Mozhaeva, N.S., Nikitenko, T.V., 2015.
Comparative analysis of two new empirical models
IRI-Plas and NGM (the Neustrelitz Global Model).
Adv. Space Res.,55, 20862098.
Maltseva, O., Mozhaeva N., Vinnik E., 2013. Validation
of two new empirical ionospheric models IRI-Plas and
NGM describing conditions of radio wave propagation
in space. Proceedings of Second International
Conference on Telecommunications and Remote
Sensing, Noordwijkerhout, The Netherlands, 11-12
July, 109-118.
Maltseva, O., Mozhaeva, N., Zhbankov, G., 2012. A new
model of the International Reference Ionosphere IRI
for Telecommunication and Navigation Systems.
Proceedings of the First International Conference on
Telecommunication and Remote Sensing, 129-138.
Maltseva, O.A., Poltavsk, O.S., Schlyupkin, A.S., 2007.
The IRI model residual difference and the new method
16
of N(h)-profile determination. Acta Geophysica, 55,
441-458.
McNamara, L.F., 1985. The use of total electron density
measurements to validate empirical models of the
ionosphere. Adv. Space Res., 5(7), 81-90.
McNamara, L.F., 2006. Quality figures and error bars for
autoscaled Digisonde vertical incidence ionograms.
Radio Sci., 41, RS4011, doi:10.1029/2005RS003444.
Ozhogin, P., Tu, J., Song, P., Reinisch, B.W., 2012. Field-
aligned distribution of the plasmaspheric electron
density: An empirical model derived from the IMAGE
RPI measurements, J. Geophys. Res., 117, A06225.
Reinisch, B.W., Galkin, I.A., Khmyrov, G.M., Kozlov,
A.V., Bibl, K., Lisysyan, I.A., Cheney, G.P., Huang,
X., Kitrosser, D.F., Paznukhov, V.V., Luo, Y., Jones,
W., Stelmash, S., Hamel, R., Grochmal, J., 2009. The
New Digisonde for Research and Monitoring
Applications. Radio Sci.,RS004115.
URSI Atlantic Radio Science Conference, 18-22 May
2015, ExpoMeloneras, Grand Canaria, http://www.at-
rasc.com
Yao, Y., Tang, J., Chen, P., Zhang, S., Chen, J., 2014. An
Improved Iterative Algorithm for 3-D Ionospheric
Tomography Reconstruction. IEEE Transactions on
Geoscience and Remote sensing, 52, 8.
Zabotin, N.A., Wright, J.W., Bullett, T.W., Zabotina,
L.Ye., 2005. Dynasonde 21: principles of data
processing, transmission and web service. 11
International Ionospheric Effects Symposium,
Alexandria, USA, 3-5 May 2005, 7В51-7В58.
Usage of the Internet Resources for Research of the Ionosphere and the Determination of Radio Wave
Propagation Conditions
17