close

Вход

Забыли?

вход по аккаунту

?

j.jastp.2018.08.007

код для вставкиСкачать
Journal of Atmospheric and Solar-Terrestrial Physics xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Journal of Atmospheric and Solar-Terrestrial Physics
journal homepage: www.elsevier.com/locate/jastp
The performance of the IRI-Plas model as compared with Alouette II and
GIM-TEC data over the midlatitude station Alma-Ata
G.I. Gordiyenkoa, O.A. Maltsevab, F. Arikanc,∗, A.F. Yakovetsa
a
Institute of Ionosphere, National Center for Space Research and Technology, Almaty, Kazakhstan
Institute for Physics, Southern Federal University, Rostov-on-Don, Russia
c
Hacettepe University, Dept. of Electrical and Electronics Engineering, Ankara, Turkey
b
A R T I C LE I N FO
A B S T R A C T
Keywords:
Midlatitude ionosphere
IRI-Plas model
Alouette Ne(h)-Profiles
Total electron content (TEC)
GIM
Climatic ionospheric models present an important medium for both investigation of physical structure and
correction of detrimental effects of ionospheric variability on space based communication, navigation and positioning systems. International Reference Ionosphere Extended to Plasmasphere (IRI-Plas) is one of the empirical models that can provide ionospheric layer parameters and electron density profiles up to GPS satellite
orbital height of 20,200 km. IRI-Plas can input not only F2 layer critical parameters, foF2 and hmF2, but also the
Total Electron Content (TEC) values. In this study, Ne(h)-profiles and TEC values obtained from the IRI-Plas
model are compared with topside Alouette II satellite profiles and Global Ionospheric Map (GIM) TEC values,
respectively. The satellite Ne(h) profiles are averaged over a midlatitude region for various seasons of 1966 and
1969. It is observed that IRI-Plas is in good agreement with Alouette profiles between 300 km and 500 km. After
500 km up to 2000 km, the profiles seem to differ where IRI-Plas Ne(h) usually overestimates the Alouette
profile. When IRI-Plas TEC is compared with GIM-TEC, it is observed that model TEC generally overestimates the
GIM-TEC values during daytime hours. The lowest ΔTEC values (10–20%) are observed in March and September,
and the highest values are observed in the months of January (around 30%) and July (up to 67%). During
nighttime, the IRI-Plas model results mainly underestimate the observational ones, up to 30% in equinoctial
months, and with ± 10% discrepancies in January and July. Since the IRI-Plas model can scale foF2 and hmF2
values using an external TEC input by equating instantaneous and median slab thickness (TEC/NmF2 ratio), the
correlation between GIM-TEC and ionosonde NmF2 values measured at the midlatitude station Alma-Ata located
at [43.25°N, 76.92°E] is obtained for various seasons during 1999, 2000, 2008, 2009 and 2012. The study
showed a positive correlation for all seasons and levels of solar activity. The correlation coefficients between the
data sets were very high (greater than or equal to 0.79) for hourly data and (greater than or equal to 0.91) for
monthly-medians at a significance level of 5%. Thus, for the practical applications, the correlation results can be
used to update slab thickness model of IRI-Plas which will lead to better scaling of foF2 and hmF2 values and
plasmaspheric Ne(h)-profile with external instantaneous TEC input.
1. Introduction
Modeling of the Earth's ionosphere is an important and challenging
task that is based on both theoretical and experimental studies of the
ionosphere, plasmasphere, Earth's magnetic field, solar events, and interactions between them. A special importance of the ionospheric
models is their use by Global Navigation Satellite Systems (GNSS) to
correct the effects of the ionosphere on their operation in terms of the
signal group delay and phase advance that can affect the accuracy of
positioning, especially, in case of using single frequency Global
Positioning System (GPS) receivers.
The ionospheric time delay is proportional to the total electron
content (TEC) along the path of the radio wave propagation. TEC is
defined as the line integral of electron density along a ray path. The unit
of TEC is TECU and 1 TECU = 1016 el/m2. The TEC computed between
a GPS receiver and satellite is called as Slant TEC (STEC). STEC is
converted to Vertical TEC (VTEC) using a simple mapping function
which may introduce additional errors to the computation of TEC
(Arikan et al., 2003). Since GPS-TEC is estimated in the local zenith
direction of a GPS receiver, the values contain the contribution from all
∗
Corresponding author.
E-mail addresses: ggordiyenko@mail.ru (G.I. Gordiyenko), mal@ip.rsu.rc (O.A. Maltseva), arikan@hacettepe.edu.tr (F. Arikan),
artyak40@mail.ru (A.F. Yakovets).
https://doi.org/10.1016/j.jastp.2018.08.007
Received 21 March 2018; Received in revised form 8 August 2018; Accepted 10 August 2018
1364-6826/ © 2018 Elsevier Ltd. All rights reserved.
Please cite this article as: Gordiyenko, G.I., Journal of Atmospheric and Solar-Terrestrial Physics (2018),
https://doi.org/10.1016/j.jastp.2018.08.007
Journal of Atmospheric and Solar-Terrestrial Physics xxx (xxxx) xxx–xxx
G.I. Gordiyenko et al.
et al., 2017; Sezen et al., 2018). In the above listed studies, it is observed that up to peak F2 layer height, the electron density profiles of
IRI and NeQuick show great similarity with those of IRI-Plas since all
the above mentioned models employ IRI for this section of the ionosphere. During geomagnetically quiet days, the topside profiles up to
the height of 2000 km are in general accordance with those of IRI and
NeQuick models. The update or scaling of model coefficients is
achieved by the use of slab thickness which is defined as the ratio of
TEC to peak electron density value, NmF2. In (Okoh et al., 2018), TEC
obtained from the NeQuick model and the IRI-Plas model are compared
with respect to GPS-TEC obtained from a single station and it has been
observed that IRI-Plas, without an external input, can overestimate
GPS-TEC especially during local daytime. As indicated in various studies in the literature including but not limited to (Maltseva et al., 2013,
2015; Panda et al., 2015; Zakharenkova et al., 2015; Adebiyi et al.,
2016, 2017; Bolaji et al., 2017; Ezquer et al., 2017; Sezen et al., 2018)
that the assimilation of TEC into the IRI-Plas model improves the representation of current ionospheric state significantly.
In the present study, we extend the recent investigations to topside
ionosphere and compare the Alouette II Ne(h)-profiles with the IRI-Plas
vertical electron density distributions for the first time for a midlatitude
region. The model TEC values are also compared with those of the
Alouette TEC and GIM-TEC for various years and ionospheric seasons
for a midlatitude station, namely Alma-Ata (43.25°N, 76.92°E). In order
to improve the possible update of model parameters for assimilation of
instantaneous TEC, the relationship between Alma-Ata ionosonde
NmF2 and GIM-TEC used to derive a relationship for midlatitude ionospheric slab thickness trend variability. Section 2 includes model and
data information. Results are provided in Section 3 and the paper ends
with Discussion and Conclusion sections.
layers of ionosphere as well as plasmasphere. The TEC values show
regular (depending upon local time, season, latitude, longitude and
solar cycle) and irregular (during geomagnetic disturbances) changes as
discussed in various studies in the literature including but not limited to
(Gulyaeva, 2011, 2012; Blanch et al., 2013; Gulyaeva et al., 2013;
Astafyeva et al., 2015; Nava et al., 2016).
To obtain a global TEC distribution, it is necessary to utilize the
model data. For that, some empirical ionosphere models based on integration of Ne(h)-profiles, or integrated electron density maps can be
used. Examples of empirical models that can describe electron density
profiles are the Klobuchar model (1987), the International Reference
Ionosphere (IRI) model (Bilitza and Reinisch, 2008; Bilitza, 2009;
Bilitza et al., 2014), the International Reference Ionosphere Extended to
Plasmasphere (IRI-Plas) model (Gulyaeva, 2003a, 2011, 2012;
Gulyaeva et al., 2011; Gulyaeva and Bilitza, 2012) and the Semi-empirical Low-latitude Ionospheric Model (SLIM) (Anderson et al., 1987).
In the studies given in the literature, it is reported that the
Klobuchar ICA (the Ionosphere Correction Algorithm) TEC model
which is based on the Bent model (Llewellyn and Bent, 1973) provides
approximately 50% in the ionospheric delay correction (Klobuchar,
1987). Recently, some improvements have been suggested for the
Klobuchar model that is described in details by Wang et al. (2016) and
Bi et al. (2017). In the case of IRI-2012 model, all three options (NeQuick, IRI01-corr, and IRI 2001) for topside electron density profile
show that the discrepancy in IRI-2012 as compared to GPS-TEC in
midlatitude regions (Mongolia and Russia) for 2009–2013 is found to
be within 5–10 TECU (Kumar et al., 2015) for quiet geomagnetic conditions. It should be noted that a change of 1 TECU (for example for GPS
L1) introduces a pseudo range error of about 0.16 m or it is about approximately 1 m for 6 TECU (Galav et al., 2012). Since the upper altitude limit for the TEC computation in the IRI model is 2000 km, it can
be a possible reason for the observed discrepancies. A significant difference between the model Ne predictions and in-situ measurements in
the altitude range 400–500 km during the unusually low and extended
solar minimum of 2008/9 is found by Lühr and Xiong (2010), as well as
a difference in shapes of the topside part in the Ne(h)-profiles
(Zakharenkova et al., 2015) which can be listed as additional causes for
the model versus GPS-TEC discrepancies. Recently, the Alouette II
electron density profiles were compared with the IRI2012 model
(Gordiyenko and Yakovets, 2017) where the authors showed that none
of the three IRI2012 topside options represented the Alouette Ne(h)profile correctly, by showing a significant disagreement in the Ne(h)
shape. In many cases, the IRI2012 models significantly underestimate
the experimental data, while in some others, the models overestimate
them. It should be noted here that these results are rather surprising
because the NeQuick and IRI-corr options were developed with different topside sounder data including the Alouette II data.
In this study, we have chosen IRI-Plas model to serve as a basis for
global representation of ionospheric parameters due to the fact that IRIPlas, whose region of interest can extend to Middle Earth Orbit (MEO)
GPS satellite altitude of 20,200 km, is the one of the few climatic
models that can input both F2 layer critical layer parameters and TEC
(Gulyaeva et al., 2011; Sezen et al., 2017, 2018). IRI-Plas can be obtained from http://ftp.izmiran.ru/pub/izmiran/SPIM/ as a FORTRAN
code and at www.ionolab.org where IRI-Plas is offered online as a space
weather service (Sezen et al., 2017, 2018). The IRI model (and the IRIPlas as one of the extensions to the plasmasphere) has been reinstated
as an international standard by ISO (ISO 16457:2014) as given in
https://www.iso.org/standard/61556.html in 2014.
Recently, there have been various studies that compared the performance of ionospheric models with IRI-Plas such as those given in
(Gulyaeva et al., 2011; Maltseva et al., 2015; Panda et al., 2015;
Zakharenkova et al., 2015; Adebiyi et al., 2016; Bolaji et al., 2017;
Ezquer et al., 2017; Okoh et al., 2018). Also, the model parameter
outputs such as TEC and foF2 are compared with those from measurement as given in (Maltseva et al., 2013; Kumar et al., 2015; Adebiyi
2. IRI-Plas model and the data sets
The online IRI-Plas model (www.ionolab.org/iriplasonline/) is
capable of calculating the electron density profile (in terms of latitude,
longitude and height) and the ionospheric vertical Total Electron
Content in TECU at a selected location for any desired date and time in
four altitude ranges: from 80 km to hmF2 (ECbot), from hmF2 to
1336 km (ECtop), from 1336 km to HPL (plasma-pause height limited in
IRI-Plas by GPS satellite orbital altitude of 20,200 km, ECpl), and finally
from 80 km to HPL (TEC) for all levels of solar activity. IRI-Plas is based
on the ionosphere part of IRI model. The Topside Ne(h) profile,
Plasmasphere Ne(h) and Te(h) models of IRI-Plas are described in
(Gulyaeva, 2003b; Gulyaeva et al., 2002; Gulyaeva and Titheridge,
2006). The altitude 1336 km refers to orbit of Topex satellite. This is
used to provide the 'ionosphere' and 'plasmasphere' contribution to total
TEC: TECbot (h < 1336 km) and TECtop (h > 1336 km) while the
plasmasphere model is merged with topside ionosphere model at height
h05top (at Ne = 0.5 NmF2). Thus, TECbot includes Ne(h) bottomside
(h < hmF2), topside ionosphere (h from hmF2 to h05top) and part of
plasmasphere (h from h05top to 1336 km). Similar to IRI, the altitude
1336 km could be changed inside IRI-Plas code to any desirable altitude
h < 2000 km.
As mentioned in the previous section, the IRI-Plas model has the
advantage (compared to other models) that it can input F2 layer peak
parameters, F2 layer critical frequency foF2 that is a measure of the
peak electron concentration NmF2 (NmF2 = 1.24f2 × 104 el/sm3
where f in MHz) and maximum ionization height hmF2, as well as TEC
values (Gulyaeva, 2011, 2012; Gulyaeva et al., 2011; Sezen et al., 2017,
2018). TEC input is used to change foF2 and hmF2 peak parameters in
IRI-Plas, and hence it is changing the total Ne(h) profile fitted to updated peak density and height.
The TEC input can be obtained from any GPS receiver using
IONOLAB-TEC algorithm available at www.ionolab.org online or as an
executable file (Sezen et al., 2013). Another option is to utilize the
automatic pop-up window that offers the Global Ionospheric Map (GIM)
2
Journal of Atmospheric and Solar-Terrestrial Physics xxx (xxxx) xxx–xxx
G.I. Gordiyenko et al.
content (SHSAT) between the level of F2-peak, i.e. HMAX, and the
satellite position (HTSAT) in units of 1012 el/cm2; the geodetic longitude (GDLON) and latitude (GDLAT); the geomagnetic planetary indices such as Kp. One possible bias in the data can come from the fact
that the peak is assumed to be at the lowest measured point in the
profile. Because the satellite measures downward from above, and the
F2 peak and the profiles in the database are given by discrete points, it
is possible that the lowest point is above the peak, but it can never be
below the peak. It is assumed that any bias introduced in this way is
small, and it is typically 10 km above the true value, even if there are no
problems with the profile (Verhulst and Stankov, 2013; and references
therein).
In this study, a number of the Alouette II satellite data is used to
compare the observed Ne(h) profiles of the topside ionosphere (above
the F2-layer peak altitude, hmF2) with those predicted by the IRI-Plas
model. The Alouette data consisting of Ne(h)-profiles (i.e. Ne values at
number of real altitudes in the topside ionosphere) that were published
in the form of Tables in the data books (Data on Topside Ionosphere,
1970, 1974, 1977) are converted to digital form by the authors. To do
the comparison, we selected several periods of quiet geomagnetic activity mainly for noon and near midnight hours of winter, equinox and
summer months available for 1966, 1968 and 1969 (Data on Topside
Ionosphere, 1970, 1974, 1977), when the Alouette II data were highly
reliable (qual = 3) and included the F2-layer peak altitude hmF2, and
the density NmF2 (which is proportional to foF2) and the Electron
Content (EC) between the level of F2-peak, i.e. HMAX, and the satellite
position (the total number of the selected profiles amounted to 48).
For identification of geomagnetic quiet days, we used the geomagnetic Kp index with a 3-h resolution, which was given in the Tables for
Ne(h)-profiles. Geomagnetic conditions were considered to be quiet
when Kp˂4 (http://spaceweather.com).
The ionosonde data that is necessary to compute the relationship
between TEC and peak electron density, NmF2, are obtained from the
ionosonde at the Institute of the Ionosphere (the Almaty station
[43.25°N, 76.92°E]) using the PARUS digital ionosonde (e.g. Karpenko
et al., 2009 and references therein; for a description, please also see
http://www.izmiran.rssi.ru). The Almaty ionosonde is connected to a
computer that is used to collect, store, and process ionograms (h'(f)) in a
digital form. The ionosonde provides readings of the virtual reflection
height h' as a function of signal frequency f with a resolution of 2.5 km
and 0.05 MHz, respectively. The information needed for the study was
read from the ionograms using the semi-automatic method with the
participation of an experienced operator. Stankov et al. (2012) showed
that this method of the ionogram processing has a higher accuracy for
reading ionospheric parameters and a larger statistical output of ionograms fit for processing as compared to the automatic method. The data
used for the present study are the F2-layer critical frequencies, deduced
from the ionograms, and the F-layer peak height (hmF2), calculated by
using the standard program POLAN (Polynomial Analysis program;
Titheridge, 1985). The ionosonde data (median foF2 deduced from ionograms and calculated hmF2 values) from Alma-Ata station [43.25°N,
76.92°E] are used for adapting of the model to the current ionospheric
conditions.
TEC interpolations for the user defined location, date and time offered
at Online IRI-Plas service of IONOLAB group at www.ionolab.org
(Sezen et al., 2018).
The Global Ionospheric TEC Maps (GIM) have proven themselves as
a useful source for global distribution of interpolated TEC values hourly
or 2 hourly intervals on a grid of 2.5° × 5° resolution in latitude and
longitude, respectively. The GIM are calculated in several independent
analysis centers whose IONOspheric EXchange (IONEX) files that contain the GIM-TEC values can be obtained from ftp://cddis.gsfc.nasa.
gov/pub/gps/products/ionex/. Generally, these centers use different
algorithms for computation of the ionospheric maps (e.g. different
methods for data interpolation on the longitude-latitude grid, and also
different receiver and satellite bias solutions and station/data distributions) that result in some difference in the global ionospheric TEC
distribution represented by the centers (Arikan et al., 2003; HernándezPajares et al., 2009; Orus et al., 2010; Gulyaeva et al., 2013; Sezen
et al., 2013; Hernández-Pajares, 2014). It follows that there are still
uncertainties of several TECU in the determination of vertical TEC values. This error will affect the IRI-Plas adjustment of NmF2 and hmF2
when IRI-Plas is run with external GIM-TEC input. In this study, we
have used CODE (Centre for Orbit Determination in Europe, Astronomical Institute, University of Berne, Switzerland) TEC for comparison
of IRI-Plas TEC due to its relatively satisfactory performance in midlatitude regions (Sezen et al., 2018). The CODE GIM-TEC are obtained
for a location closest to the Alma-Ata point [42.5°N, 75°E] (closer than
200 km) on the IONEX grid. The CODE GIM-TEC values are compared
with the IRI-Plas TEC for January, March, July and September of 2009,
low solar activity year.
The original form of the data from the Alouette II was the topside
ionograms printed on film. The satellite data were received at the
telemetry station Kashima [35°57.2′ N, 140°40.0′ E] since 1966, and
they were processed and analyzed at the Radio Research Laboratories
(RRL). The ionogram data on the graphic display unit mounted at the
operator console were scaled to remove the echo traces through a
graphic input unit. The scaled data were transferred to the slave computer where a computation of the N(h) analysis were carried out using
the ONLIP (On-Line Ionogram Processor) constructed in 1975 as an
ionogram analysis system serving both for Alouette-ISIS satellites and
ISS. Theoretical h'(f) traces based on the calculated N(h) profiles were
returned to the graphic display unit where the theoretical h'(f) traces as
well as the original traces were displayed simultaneously for comparison. If necessary, some corrections for the scaling traces are performed
so that better consistency may be attained. Degree of validity of the
computed electron density profiles was examined in two ways: one is a
calculation of the electron density profiles computed from O-mode, Xmode and Z-mode traces and examining consistency between them; the
other is a calculation of theoretical h'(f) traces for O-mode, X-mode and
Z-mode waves by using each electron density profile and comparison of
them with observed h'(f) traces to examine their consistency. The Ne
(h)-profiles obtained from Alouette II ionograms were published in the
form of Tables in the data books (Data on Topside Ionosphere, 1970,
1974, 1977) that was one of the ISIS (International Satellites for Ionospheric Studies) project in RRL. In the Tables, degree of validity of
the electron density profiles is shown by the quality factor “qual” (in
one-digit number (3, 2, 1) and one-letter code (o, x, z)), where 3 indicates that the electron density profiles are highly reliable, 2 denotes
that the electron density profiles are moderately reliable, 1 is a value
indicating that the electron density profiles should be used with caution; the one-letter code indicates what mode trace (O, X, Z) in the
ionogram was used to obtain the electron density profile. The reduction
program was developed by Hojo and Nishizaki (1971). In the Tables,
the electron density (in units of 105 el/cm3) is given in altitude intervals
of 50 km; the date and times (in UT) of each ionogram are given at the
caption of each Table. Other data corresponding to each ionogram are
also listed as the peak electron density of F2 layer (NMAX) in units of
105 el/cm3; the height of F2-layer (HMAX) in units of km; the electron
3. Results
In this section, the performance of IRI-Plas topside electron density
profile and model TEC values is obtained by comparison to Alouette II
and GIM-TEC data as discussed in detail in the following.
3.1. Comparison between the Alouette II observations and the IRI-Plas
model outputs
In order to compare the model profiles with Aloutte data, we selected several periods of quiet geomagnetic activity when the Alouette
II data were highly reliable (“qual” = 3) and included both the Ne(h)3
Journal of Atmospheric and Solar-Terrestrial Physics xxx (xxxx) xxx–xxx
G.I. Gordiyenko et al.
Table 1
The date, Local Time (LMST), Geodetic Latitudes and Longitudes (GDLAT, GDLON), Kp/qual indices, observed the F2-layer peak parameters (foF2obs/hmF2obs), the
satellite altitudes (HTSAT), observed electron content (SHSAT), IRI-Plas TEC between the level of F2-peak and 1336 km (IRI-Plas TECtop), and differences between
observed electron content (SHSAT) values and those calculated by the IRI-Plas model (DTECtop).
Date
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
LMST
GDLAT
GDLON
Kp/qual
yy/mm/dd
hh/mm/ss
deg
deg
1
2
3
4
5
12/10/1966
12/10/1966
12/10/1966
12/10/1966
12/10/1966
12/10/1966
12/10/1966
12/10/1966
12/10/1966
14/10/1966
14/10/1966
14/10/1966
14/10/1966
14/10/1966
26/02/1969
26/02/1969
26/02/1969
26/02/1969
26/02/1969
26/02/1969
28/02/1969
28/02/1969
28/02/1969
28/02/1969
28/02/1969
20/12/1969
20/12/1969
20.12.1969
20.12.1969
20.12.1969
21.12.1969
21.12.1969
23.12.1969
23.12.1969
23.12.1969
23.12.1969
23.12.1969
02.03.1969
02.03.1969
02.03.1969
05.03.1969
05.03.1969
16.03.1969
26.03.1969
08.09.1969
02.06.1969
03.06.1969
27.08.1969
20/09/48
20/11/43
20/16/02
20/18/23
20/20/50
20/23/33
20/26/29
20/29/32
20/32/44
19/59/03
20/01/12
20/03/33
20/06/05
20/08/42
23/45/16
23/47/37
23/49/51
23/51/58
23/54/00
23/55/59
23/31/04
23/33/29
23/35/48
23/37/13
23/40/06
12/48/20
12/50/59
12/53/26
12/55/45
12/59/12
12/48/00
12/52/11
12/29/49
12/32/15
12/35/42
12/38/09
12/39/45
23/20/57
23/23/14
23/27/32
23/00/26
23/05/01
11/14/59
10/04/20
1/07/11
12/44/16
12/44/42
2/32/06
33.4
35.3
39.1
41
42.9
44.8
46.8
48.7
50.5
36.8
38.7
40.7
42.6
44.5
49.4
48.1
46.7
45.2
43.8
42.3
49.9
48.5
47.1
46.2
44.2
49.5
48
46.4
44.8
42.3
44.9
41.8
49.5
47.9
45.5
43.4
41.9
47.6
46.1
43.1
47.3
44.3
45.5
43.5
43.8
47.1
41.6
41.1
150.2
150.5
151.4
151.8
152.3
152.8
153.4
154.1
154.7
138.6
139
139.5
140
140.5
127.8
128.2
128.7
129.1
129.4
129.8
147.6
148.1
148.5
148.8
149.3
132.8
133.3
133.8
134.2
134.9
129.2
130
149.2
149.7
150.3
150.7
151
137.9
138.3
139.1
122.2
123.1
127.7
135.5
133.1
151.9
148.4
133.9
2+/3x
2+/3x
2+/3x
2+/3x
2+/3x
2+/3x
2+/3x
2+/3x
2+/3x
1-/3x
1-/3x
1-/3x
1-/3x
1-/3x
2/3x
2/3x
2/3x
2/3x
2/3x
2/3x
3-/3x
3-/3x
3-/3x
3-/3x
3-/3x
0+/3x
0+/3x
0+/3x
0+/3x
0+/3x
2/3x
2/3x
2-/3x
2-/3x
2-/3x
2-/3x
2-/3x
2/3x
2/3x
2/3x
2-/3x
2-/3x
1/3x
3+/3x
2-/3x
1-/3x
2-/3x
2/3x
foF2obs/hmF2obs MHz/km
HTSAT
SHSAT
16
IRI-Plas TECtop
2
16
2
DTECtop
km
(´10 el/m )
(´10 el/m )
(´1016el/m2)
6
7
8
9
10
3.8/304
3.8/320
4.5/278
4.7/271
5.0/266
5.0/283
5.2/278
5.1/273
5.0/283
3.8/322
3.8/298
3.8/336
3.9/322
3.9/329
5.7/348
5.8/310
5.7/299
5.8/317
5.8/332
5.8/320
6.1/335
5.8/368
5.8/352
6.1/300
5.7/337
9.3/214
9.3/208
9.1/209
8.8/215
9.0/203
8.5/228
8.9/221
9.2/219
9.4/185
9.1/184
9.3/226
9.5/258
5.2/356
5.5/313
5.6/319
6.2/370
6.4/364
13.3/249
11.9/337
5.2/318
6.6/253
8.8/219
4.5/474
951
923
868
842
818
793
769
747
726
848
823
798
774
751
2084
2052
2020
1988
1956
1922
2015
1982
1950
1929
1884
1541
1575
1608
1642
1696
1681
1748
1661
1694
1749
1795
1828
1879
1845
1777
1745
1676
2899
2710
1753
1151
1288
2184
3.692
3.458
4.706
4.894
5.510
5.107
5.195
5.002
4.470
3.103
3.412
2.706
2.862
2.901
8.868
10.56
11.13
10.50
10.29
10.35
10.70
9.385
10.30
11.58
10.94
17.50
17.31
18.38
16.56
18.04
14.08
15.17
15.50
18.66
17.68
16.07
14.03
6.591
9.165
9.632
9.169
9.444
35.79
29.59
9.512
12.71
23.88
7.354
–
–
–
–
–
–
–
–
–
–
–
–
–
–
11.04
10.94
10.45
10.98
11.01
10.96
12.39
11.64
11.44
11.87
10.87
22.53
22.34
21.49
20.37
20.85
19.42
21.19
22.33
22.20
20.90
22.87
24.63
9.36
9.92
10.26
13.04
13.67
47.61
43.71
8.91
13.30
20.98
7.76
–
–
–
–
–
–
–
–
–
–
–
–
–
–
−2.2
−0.4
0.7
−0.5
−0.7
−0.6
−1.7
−2.3
−1.1
−0.3
0.1
−5.0
−5.0
−3.1
−3.8
−2.8
−5.3
−6.0
−6.8
−3.5
−3.2
−6.8
−10.6
−2.8
−0.8
−0.6
−3.9
−4.2
−11.8
−14.1
0.6
−0.6
2.9
−0.4
satellite position were calculated according to the conditions given in
Table 2 to examine the fitting performance of the model; the model
calculations were adapted to hmF2obs and foF2obs .
Figs. 1–3 demonstrate examples of the averaged Alouette Ne(h)profiles in comparison with the model outputs. Here, one can see that
for near noon hours of winter (Fig. 1a and b), summer (Fig. 2b) and
equinoctial (Fig. 3d) seasons the IRI-Plas model gives a general overestimation of the actual Ne values at the altitudes from 400 and 500 km
up to 2000 km reaching a factor of 3 or 4 at the heights of 600–800 km.
The same tendency can be observed for 20 LT (Fig. 3a and b) but due to
the low altitude of the satellite position, it is not possible to determine
the altitude range for the exceeding Ne. For near midnight hours of
winter, summer and equinoctial seasons, the IRI-Plas model overestimates the Alouette Ne values at the altitudes from 600 km and
700 km to 1500 km and 1600 km, while for the higher altitudes
profiles and the F2-layer peak parameters, hmF2 (HMAX) and foF2
(proportional to NMAX), height of the satellite position (HTSAT),
electron content between the level of F2-peak and satellite position
(SHSAT), (which makes 48 cases in total) as provided in Table 1. The
data set is divided into several parts (groups) for specific time intervals
relevant to the daytime and nighttime periods such as 10 to 11 LT, 12 to
13 LT, around 20 LT, and 23 to 03 LT of winter, equinox and summer
seasons. Then, by averaging the measured Ne densities for corresponding altitudes, the mean Ne(h)-profiles are calculated for each
group.
The F2 layer peak parameters are also correspondingly averaged.
Mean local time (LT), location (φ, λ) in latitude and longitude, respectively, averaged hmF2obs and foF2obs for the chosen data groups are
given in Table 2 (columns 1–6). Using the IRI-Plas models the corresponding Ne(h)-profiles from hmF2obs to the average height of the
4
Journal of Atmospheric and Solar-Terrestrial Physics xxx (xxxx) xxx–xxx
G.I. Gordiyenko et al.
Table 2
Date, location (φ, λ), Local Time (LT), the averaged observed F2-layer peak parameters (hmF2 obs, foF2obs ), the averaged model F2-layer peak parameters (hmF2mod ,
foF2mod ), the difference between the observed and model parameters (DhmF2, DfoF2), Total Electron Content (in TECU) calculated by the IRI-Plas 2017 model with
using of the observed and model F2-layer parameters (TEC1 and TEC2 correspondingly), difference between TEC1 and TEC2 (DTEC). The number of Ne(h) profiles
used for averaging are shown in parentheses in the first column.
Date dd.mm.yy (N)
φ, deg.
λ, deg.
Time (LT) hh:mm
hmF2 obs km
foF2 obs MHz
hmF2 mod km
foF2 mod MHz
DhmF2 km
DfoF2 MHz
TEC1
TEC2
DTEC
1
2
3
4
5
6
7
8
9
10
11
12
13
43°N
48°N
43°N
48°N
40°N
40°N
45°N
44°N
41°N
45°N
140°E
140°E
140°E
140°E
152°E
140°E
130°E
130°E
134°E
150°E
13:00
13:00
23:30
23:30
20:00
20:00
24:00
10:40
02:32
13:00
225
203
330
329
284
321
340
293
474
236
8.98
9.24
5.75
5.84
4.68
3.84
5.69
12.6
4.5
7.7
266
263
353
359
310
306
356
286
348
305
10.7
10.8
4.9
4.6
5.1
5.3
5.1
11.0
6.0
7.6
−41
−60
−23
−30
−26
15
−16
7
126
−69
−1.72
−1.56
0.9
1.24
−0.42
−1.46
0.6
1.6
−1.5
0.1
29.60
29.58
15.34
15.84
10.56
8.13
15.24
66.14
12.64
25.78
43.1
44.1
12.2
11.2
13.5
13.5
13.4
51.1
17.2
29.5
−13.50
−14.52
3.14
4.64
−2.94
−5.37
1.84
15.04
−4.56
−3.74
21.12.69
21.12.69
27.02.69
27.02.69
12.10.66
14.10.66
03.03.69
20.03.69
27.08.69
03.06.69
(6)
(6)
(3)
(8)
(9)
(5)
(6)
(2)
(1)
(2)
Fig. 1. The average Ne(h)-profiles for the topside ionosphere derived from the Alouette II measurements (solid line) and IRI-Plas (cross) model for selected time
periods of December and February 1969. The IRI-Plas profiles were normalized to the measured hmF2 and NmF2.
Fig. 2. The average Ne(h)-profiles for the topside ionosphere derived from the Alouette II measurements (solid line) and IRI-Plas (cross) models for selected time
periods of June and August 1969. The IRI-Plas profiles were normalized to the measured hmF2 and NmF2.
5
Journal of Atmospheric and Solar-Terrestrial Physics xxx (xxxx) xxx–xxx
G.I. Gordiyenko et al.
Fig. 3. The average Ne(h)-profiles for the topside ionosphere derived from the Alouette II measurements (solid line) and IRI-Plas (cross) models for selected time
periods of October 1966 and March 1969. The IRI-Plas profiles were normalized to the measured hmF2 and NmF2.
factor of 1.3–1.5) in daytime and 2 to 5 TECU (a factor of 1.1–1.7) in
nighttime. It means that the hmF2 and foF2 errors have a strong impact
on the TEC estimations. At that, using a number of different pairs of the
observed and model hmF2 and foF2 data, we found that the foF2 errors
have a stronger impact on the TEC estimations than the hmF2 errors.
Hence, these results indicated a direction to examine the dependency of
observed TEC and NmF2 values as a possible correction for model
outputs.
(h > 1600 km) the actual Ne values are significantly underestimated
as shown in Figs. 1c, d, 2a, and 3c.
To assess the performance of the IRI-Plas model in terms of total
electron content output, the topside TEC values (TECtop) were calculated using the IRI-Plas model for all selected cases given in Table 1
according to dates, local time, locations, the F2-layer peak height hmF2
and frequency foF2. Note, using the IRI-Plas model, the TECtop values
were calculated for the height range of hmF2 at 1332 km (an available
option in the IRI-Plas model) only for cases where the HTSAT values
(Table 1, column 7) were close or more than 1335 km. Also, Table 1
shows the Alouette TECtop (SHSAT, column 8), IRI-Plas-TECtop (column
9) values and computed absolute deviations ΔTECtop (Table 1, column
10) between the Alouette TECtop data and the modeled ones for all cases
selected. There are examples (daytime of December and March 1969)
when the difference between modeled TECtop and measurements is
sufficiently large exceeding the experimental data by about 5–14 TECU
(Table 1, lines 26–27, 31–33, 36–37, 43–44), and we can conclude that
there is a strong dependence of the topside electron content from the
topside Ne(h)-profile shape (see Figs. 1a, b and 3d). As a whole, it is
seen that the IRI-Plas model overestimates the Alouette TECtop values
for all cases considered not depending on the fact that the upper
boundary (1336 km) for the TEC calculation in the model is significantly lower than the altitude of the satellite.
Based on the above results, we made an attempt to examine the
capability of the model in predicting the ionospheric Total Electron
Content (TEC) results. First of all, averaged hmF2mod and foF2mod values were calculated (Table 2, columns 7 and 8) for the corresponding
data groups which showed that the model can overestimate or underestimate the Alouette II experimental F2 layer peak parameters. Then,
the IRI-Plas-TEC values were calculated for all data groups in the altitude range 80 km to 20,200 km. The model outputs are provided in
Table 2 where the calculations were adapted to the experimental (TEC1,
column 11) or the model (TEC2, column 12) F2-layer peak parameters.
The calculations indicate that there is a certain amount of difference
between TEC1 and TEC2 values reaching at times up to 13 to 15 TECU (a
3.2. What is the dependence between observed TEC and NmF2 values?
As mentioned in the previous sections, IRI-Plas can update its critical parameters foF2 and hmF2 by using an external TEC input. This is
achieved by equating the instantaneous slab thickness (ratio of TEC to
NmF2) to the median slab thickness values (which are already available
in the coefficients that are necessary to run the model). Therefore, when
the user inputs external TEC, the model computes instantaneous NmF2,
which in turn is used in update of foF2 and hmF2.
Due to the above mentioned results which report certain discrepancies between ‘measured/experimental’ and ‘model’ values, a
further investigation is carried out in this study to determine the regional slab thickness trend model using hourly and monthly-median
(indicated with subscript “med”) values of ionosonde NmF2 and CODE
GIM-TEC values as mentioned in Section 2 for the Alma-Ata location.
The NmF2 and CODE GIM-TEC data were obtained for the periods of
1999–2000, 2008–2009 and 2012 which are characterized by high
(F10.7an. av. = 154 to 179), low (F10.7 an. av. = 69 to 71) and moderate
(F10.7 an. av. = 120) solar activity where the subscript “an. av.” denotes
annual average values of F10.7 for the years of interest. For the above
mentioned years, the ionosonde data were relatively complete except
for a two month gap in 2000.
As an example, Figs. 4–6 illustrate scatter plots of the GIM-TEC
versus NmF2 values over the same years for a number of months selected to be typical for winter, summer and equinoctial seasons. The left
panels indicate the hourly data; the right panels show the median ones.
6
Journal of Atmospheric and Solar-Terrestrial Physics xxx (xxxx) xxx–xxx
G.I. Gordiyenko et al.
Fig. 4. Correlation of hourly (left panels) and monthly-median (right panels) GIM-TEC versus NmF2 at Alma-Ata for selected months of 1999.
Units in these figures are TECU (Total Electron Content Units, 1
TECU = 1016 el/m2) for GIM-TEC and PECU (the F2-layer Peak Electron Concentration Units, 1 PECU = 104 el/cm3) for NmF2. The full
dots represent the GIM-TEC and NmF2 values, dashed-lines represent
the linear fit for the dependence of GIM-TEC on NmF2 using the least
squares estimator. The correlation coefficients (r), the coefficients of
determination (R2), standard deviations (σ), and the derived regression
equations are shown in the figures. The results show that although there
is some scattering of the data the correlation coefficients are high (from
0.84 to 0.97 for hourly data and from 0.94 to 0.99 for monthly-medians), for all cases, the data sets are strongly correlated regardless of the
season and the solar activity levels. The results confirm the fact that
NmF2 and GIM-TEC variations may be driven by the same main factor,
namely the solar activity.
7
Journal of Atmospheric and Solar-Terrestrial Physics xxx (xxxx) xxx–xxx
G.I. Gordiyenko et al.
Fig. 5. Correlation of hourly (left panels) and monthly-median (right panels) GIM-TEC versus NmF2 at Alma-Ata for the year of 2008.
data (not less than 0.79). Standard deviations (σ in TECU) showed an
evident dependence on the level of solar activity. In high solar activity
(1999–2000), the deviations ranged from 2.76 to 6.0 for the hourly data
and from 1.2 to 4.3 for monthly medians. For low solar activity years,
they varied from 0.57 to 1.5 for hourly correlations and from 0.3 to
1.05 for median correlations. The larger scattering of the data for the
years 1999 and 2000 may be caused by stronger variability of the
The correlation coefficients (r), standard deviations (σ), the regression coefficients (a) and constants (b) of the linear fits are summarized in Table 3, where Table 3a is for hourly values, and Table 3b is
for monthly medians.
The results in Table 3 indicate that despite the different seasons and
levels of solar activity, the correlation between median values tend to
be higher (not less than 0.91) as compared with those for the hourly
8
Journal of Atmospheric and Solar-Terrestrial Physics xxx (xxxx) xxx–xxx
G.I. Gordiyenko et al.
Fig. 6. Correlation of hourly (left panels) and monthly-median (right panels) GIM-TEC versus NmF2 at Alma-Ata for the year of 2012.
4. Discussion
ionosphere in the periods of high solar activity. The regression parameters “a” and “b” show a very evident annual dependence with extreme values in the summer season as given in Tables 3a and 3b.
The results are discussed in detail in the following section.
In Section 3.1, it was shown that the IRI-Plas model mainly overestimates the Alouette TECtop values, which is counter intuitive to the
fact that the upper boundary (1336 km) for the IRI-Plas-TEC calculation
in the model is significantly lower than the altitude of the satellite.
Thus, it follows that we can expect an overestimation of TEC values by
9
Journal of Atmospheric and Solar-Terrestrial Physics xxx (xxxx) xxx–xxx
G.I. Gordiyenko et al.
Table 3a
The linear correlation coefficients (r), standard deviations (σ), and coefficients a and b obtained from the least squares fitting of the relation y = ax + b, where
x = NmF2 and y = TEC for hourly values.
Year/Month
r
1999
2000
2008
2009
2012
σ
1999
2000
2008
2009
2012
a
1999
2000
2008
2009
2012
b
1999
2000
2008
2009
2012
Jan.
Feb.
Mar.
Apr.
May
Jun.
Jul.
Aug.
Sep.
Oct.
Nov.
Dec.
0.94
0.96
0.88
0.85
0.97
0.94
0.97
0.89
0.90
0.97
0.95
0.96
0.93
0.89
0.97
0.95
0.96
0.93
0.90
0.97
0.88
0.93
0.88
0.89
0.92
0.84
0.85
0.85
0.86
0.90
0.90
0.79
0.87
0.87
0.91
0.89
0.92
0.86
0.89
0.93
0.93
0.95
0.91
0.91
0.96
0.93
–
0.91
0.90
0.95
0.95
–
0.87
0.85
0.94
0.96
0.93
0.84
0.84
0.96
2.76
2.92
0.86
0.96
1.70
3.28
4.10
0.97
0.88
1.50
4.15
5.70
1.20
1.12
2.20
3.50
6.00
1.20
0.57
2.40
5.00
4.60
1.30
1.31
3.10
4.70
4.80
1.50
1.32
3.00
3.80
5.90
1.20
1.23
2.70
3.90
4.30
1.20
1.10
2.40
3.60
3.80
1.00
1.05
2.40
5.30
–
1.10
1.14
2.70
5.40
–
0.93
1.10
2.40
2.90
4.60
0.80
1.01
1.00
0.180
0.199
0.139
0.147
0.177
0.213
0.242
0.157
0.160
0.191
0.204
0.301
0.201
0.178
0.211
0.266
0.327
0.232
0.202
0.251
0.303
0.321
0.268
0.246
0.287
0.327
0.301
0.287
0.264
0.340
0.316
0.307
0.282
0.244
0.324
0.295
0.277
0.237
0.242
0.289
0.233
0.254
0.186
0.204
0.244
0.206
–
0.153
0.148
0.207
0.219
–
0.131
0.130
0.187
0.199
0.200
0.116
0124
0.160
5.713
4.270
4.307
3.537
5.900
4.824
3.810
4.205
3.615
5.948
5.640
3.680
3.761
4.165
6.002
2.487
0.871
3.526
4.107
4.751
3.484
1.749
3.113
2.932
4.232
4.292
5.580
2.330
2.379
1.601
4.084
5.440
2.101
2.710
2.504
4.086
3.884
3.313
2.735
3.677
5.672
3.403
4.044
4.352
5.497
6.882
–
4.355
5.590
5.522
5.351
–
4.494
5.302
5.825
4.667
5.576
4.048
5.155
5.481
Table 3b
The linear correlation coefficients (r), standard deviations (σ), and coefficients a and b obtained from the least squares fitting of the relation y = ax + b, where
x = NmF2med and y = TECmed for monthly medians.
Year/Month
r
1999
2000
2008
2009
2012
σ
1999
2000
2008
2009
2012
a
1999
2000
2008
2009
2012
b
1999
2000
2008
2009
2012
Jan.
Feb.
Mar.
Apr.
May
Jun.
Jul.
Aug.
Sep.
Oct.
Nov.
Dec.
0.99
0.98
0.95
0.98
0.99
0.99
0.99
0.98
0.98
0.99
0.99
0.98
0.97
0.98
0.99
0.99
0.99
0.98
0.98
0.99
0.98
0.99
0.97
0.98
0.98
0.97
0.96
0.94
0.91
0.94
0.98
0.99
0.95
0.92
0.98
0.99
0.99
0.97
0.97
0.98
0.99
0.99
0.98
0.97
0.99
0.99
–
0.98
0.98
0.99
0.98
–
0.94
0.96
0.98
0.99
0.98
0.98
0.94
0.99
1.24
1.82
0.55
0.32
0.90
1.29
2.54
0.40
0.32
0.50
1.03
4.30
0.70
0.49
0.89
1.30
3.70
0.70
0.50
1.10
2.10
1.56
0.70
0.54
1.50
2.40
2.53
0.90
1.05
2.10
1.80
1.20
0.70
0.87
1.10
1.20
1.30
0.50
0.50
1.30
1.20
1.80
0.37
0.50
0.50
2.00
–
0.50
0.51
1.60
3.90
–
0.60
0.52
1.30
1.50
2.80
0.30
0.56
0.44
0.201
0.184
0.144
0.160
0.180
0.236
0.233
0.163
0.166
0.187
0.223
0.303
0.229
0.198
0.220
0.317
0.337
0.268
0.214
0.249
0.357
0.371
0.361
0.290
0.319
0.432
0.343
0.332
0.266
0.402
0.407
0.456
0.306
0.286
0.338
0.312
0.304
0.267
0.273
0.319
0.260
0.287
0.201
0.216
0.249
0.213
–
0.168
0.149
0.227
0.223
–
0.131
0.145
0.183
0.199
0.209
0.121
0.119
0.162
4.802
4.650
4.288
3.242
5.829
4.351
4.300
3.965
3.459
6.039
3.897
3.190
3.184
3.775
5.496
−0.812
−0.876
2.590
3.883
5.076
−1.203
−4.053
1.081
1.534
1.446
−3.221
1.38
1.322
2.145
2.115
−3.206
−5.710
1.718
1.823
1.459
2.520
1.701
2.688
1.941
2.126
3.476
1.226
3.747
4.163
5.151
5.260
–
3.931
5.550
4.596
4.965
–
4.501
5.035
5.765
4.606
4.626
3.981
5.128
5.369
GIM-TECmed values during daytime hours. The lowest ΔTEC values
(10%–20%) are observed in March and September, and the highest
values are observed in the months of January (around 30%) and July
(up to 67%). During nighttime, the IRI-Plas model results mainly underestimate the observational ones, up to 30% in equinoctial months,
and with ± 10% discrepancies in January and July.
As stated above, there are many papers (see Section 1) that have
compared IRI and IRI-Plas TEC with TEC measurements. One of them is
the paper by Zakharenkova et al. (2015) where the IRI-Plas TEC are
compared with the GPS vertical TEC data (vTEC) recorded at a European midlatitudinal GPS station POTS (Potsdam, Germany) [52.4°N;
13.1°E]. The authors reported that IRI-Plas generally overestimates GPS
vTEC for low and moderate solar activity years (2008 and 2003). These
the model. On the assumption of this fact, we calculated the electron
content in the 80 to 20,200 km altitude range by using the IRI-Plas
model to estimate the difference between the GIM-TEC and IRI-PlasTEC results. To demonstrate the results, January, March, July and
September of 2009 were chosen. The IRI-Plas-TEC values were calculated for each hour on the 15th day of selected months, where the ionosonde data (median foF2 deduced from ionograms and calculated
hmF2 values from Alma-Ata station) were used as an input to update
the model. Fig. 7 illustrates results of the comparison in terms of the
difference between the IRI-Plas-TEC and GIM-TECmed values expressed
in percentages ((IRI-Plas-TEC - GIM-TECmed)/GIM-TECmed, %) where
GIM-TECmed is the monthly-median of GIM-TEC for each month. It
appears that in most of the cases the IRI-Plas model overestimates the
10
Journal of Atmospheric and Solar-Terrestrial Physics xxx (xxxx) xxx–xxx
G.I. Gordiyenko et al.
a significance level of 5%. These results allow us to assume an availability to calculate the TEC values by using linear correlations with
NmF2. However, although the correlation coefficients are very high,
almost all above 0.9, the parameters “a” and “b” both have significant
dependencies on the months and years. For example, it is evident that
the parameter “a” is systematically larger in the summer months
compared to those in winter. Two reasons can be assumed why the “a”
and “b” parameters show the seasonal dependence: 1- the topside Ne
profiles are different in different seasons, 2 – seasonal variations of the
NmF2 and GIM-TEC values are different.
To test this assumption, the diurnal NmF2 and GIM-TEC variations
were compared for winter (January) and summer (July) months of the
year 2000 (high solar activity) and 2009 (low solar activity). Fig. 8a
shows the results where one can see the known seasonal anomaly (Zou
et al., 2000; Danilov, 2017) in the diurnal NmF2 variations when NmF2
is greater in winter than in summer by day, but the anomaly disappears
at night. In contrast, Fig. 8b shows that this seasonal feature does not
appear in the case for the GIM-TEC values, that is the daytime GIM-TEC
has greater values in the summer season compared to those in winter
season. The next figure (Fig. 8c) shows examples of a comparison between the averaged topside (the Alouette data) Ne(h)-profiles taken for
daytime of winter and summer conditions of the years 1969 (Section 2,
Tables 1–2) and 1968 (Data on Topside Ionosphere, 1974).
It should be noted that the seasonal anomaly appears only up to
approximately 300–400 km, but at higher altitudes, the Ne(h)-profiles
show the lowest Ne values in the winter season compared with the
summer season, that is the seasonal anomaly disappears in the topside
ionosphere at the altitudes above 300 and 400 km. Thus, our present
study shows that the seasonal variations of the F-region electron density
(NmF2) and TEC are different. Also, the shape of the topside Ne(h)profile varies significantly in different seasons which may be a factor for
the seasonal variations of “a” and “b” parameters. These results can be
utilized to update the model median slab thickness parameters for
better representation of regional ionospheric climatology.
Fig. 7. Diurnal variations of the difference (in %) between the IRI-Plas model
results and GIM-TECmed for January, March, July and September 2009. The IRIPlas-TEC values were calculated up to 20,200 km using the median foF2 and
hmF2 values obtained from the observations at the Alma-Ata ionospheric station.
results confirm our results for daytime hours, except for the nighttime
when the IRI-Plas TEC mainly underestimates the GIM TEC results at
the Alma-Ata location (Kazakhstan).
As the above mentioned ‘model’ versus ‘measurement’ discrepancies
for Ne(h) and TEC are observed, the computation of experimental slab
thickness has come to attention. As given in previous sections, ionospheric slab thickness (TEC/NmF2) is one of the major parameters that
are involved in model development. In the literature, there have been
many efforts to investigate the variability structure of slab thickness.
For example, in (Davies and Liu, 1991), it is observed that the slab
thickness varies approximately linearly with the 12-month smoothed
values of the 10.7-cm solar radio flux. In middle latitudes, during
winter, midnight slab thickness is essentially independent of the flux,
whereas in summer and equinox seasons, the midnight thickness increases with the increase of solar flux. The noon thickness is always
linearly proportional to the increase of solar flux in all seasons. In
(Jayachandran et al., 2004), it is reported that hourly values of slab
thickness reach to maximum during pre-sunrise and post-sunset hours.
Same pre-sunrise peak and higher night time variability are also observed in (Chuo et al., 2010; Guo et al., 2011). In midlatitude region,
during post-noon hours and equinox seasons, higher slab thickness
values are observed (Jin et al., 2007; Venkatesh et al., 2011). In
(Stankov and Warnant, 2009), it is mentioned that slab thickness is one
of the main parameters in ionospheric model and examples of climatological midlatitude slab thickness behavior are presented. The midlatitude diurnal, seasonal and solar activity dependence of slab thickness is provided in detail in (Kouris et al., 2008, 2009; Mosert et al.,
2013). Using the similar regression analysis technique that is applied in
the previous section, it is reported that both TEC and slab thickness are
seasonal dependent and slab thickness is practically independent of
solar activity since both NmF2 and TEC are highly correlated. Long
term investigations reported in (Jakowski et al., 2017) indicate that
slab thickness reflects the dynamics in both ionosphere and plasmasphere especially in geomagnetically disturbed days.
This study indicated a high positive correlation between NmF2 and
TEC parameters for all seasons and levels of solar activity which is in
accordance with the general trends reported in the literature as given
above. The correlation coefficients between the data sets are not less
than 0.79 for hourly data and not less than 0.91 for monthly-medians at
5. Conclusion
In this paper, we present results of a comparison of the TEC values
and Ne(h)-profiles with the IRI-Plas model predictions using GIM-TEC
results for Alma-Ata location (43.250 N, 76.920 E) and Alouette II data.
The results indicate the IRI-Plas model gives a general overestimation of
the daytime Alouette Ne values at the altitudes from 400 and 500 km up
to 2000 km for winter, summer and equinoctial seasons reaching a
factor of 3 or 4 between 600 and 800 km. For near midnight hours, the
IRI-Plas model overestimates the Alouette Ne values at the altitudes
starting from 600 or 700 km up to 1500 and 1600 km, while for the
higher altitudes (h > 1600 km), the experimental Ne values are found
to be significantly underestimated. Our results show that there is a
strong dependence of the topside electron content from the topside Ne
(h)-profile shape. It was also found that the foF2 errors have stronger
impact on the TEC estimations compared to those on the hmF2. An
attempt was made to study the correlation dependence between GIMTEC and NmF2 values. The study showed a positive correlation between
the two parameters, and despite the different seasons and levels of solar
activity the correlation coefficients between the data sets were very
high (not less than 0.79 for hourly data and not less than 0.91 for
monthly-medians) at a significance level of 5%. So, to use the model in
practical application it needs to be adapted to the observed foF2 and
hmF2 values and have a reliable plasmasphere part in the altitude Ne
description.
11
Journal of Atmospheric and Solar-Terrestrial Physics xxx (xxxx) xxx–xxx
G.I. Gordiyenko et al.
Fig. 8. Diurnal NmF2 (a) and GIM-TEC (b) variations for January and July 2000. The topside (the Alouette data) Ne(h)-profiles for daytime of winter and summer
conditions of 1969 and 1968 (c).
Acknowledgments
Appendix A. Supplementary data
The online IRI-Plas model is presented at the site www.ionolab.org/
iriplasonline/. The IONEX files that contain GIM-TEC are obtained from
IGS Iono Working Group Data Analysis Centre of Jet Propulsion
Laboratory at ftp://cddis.gsfc.nasa.gov/pub/gps/products/ionex/. The
Alouette data can be reached at ‘Data on Topside Ionosphere’ files given
in References section. The ionosonde data is made available at http://
www.izmiran.rssi.ru and ftp://ftp.ngdc.noaa.gov/ionosonde/data/.
The Institute of Ionosphere Working Group (GG and AYa) was supported by the Kazakhstan Foundation for Basic Research (Grant Nos.
0115PR00399 and 0115PR01275). OAM was supported by Grant under
the state task N3.9696.2017/8.9 from Ministry of Education and
Science of Russia. FA was supported by Türkiye Bilimsel ve Teknolojik
Araştırma Kurumu (TÜBİTAK) (Scientific and Technological Research
Council of Turkey (TUBITAK)) , Turkey Grant number EEEAG 115E915.
Finally, the authors wish to express their thanks to Dr. T.L. Gulyaeva for
insightful information on IRI-Plas model.
Supplementary data related to this article can be found at https://
doi.org/10.1016/j.jastp.2018.08.007.
References
Adebiyi, S.J., Adimula, I.A., Oladipo, O.A., Joshua, B.W., 2016. Assessment of IRI and
IRI‐Plas models over the African equatorial and low‐latitude region. J. Geophys. Res.
Space Phys. 121 (7), 7287–7300.
Adebiyi, S.J., Adebesin, B.O., Ikubanni, S.O., Joshua, B.W., 2017. Performance evaluation
of GIM-TEC assimilation of the IRI-Plas model at two equatorial stations in the
American sector. Space Weather 15, 726–736. https://doi.org/10.1002/
2017SW001596.
Anderson, D.N., Mendillo, M., Herniter, B.C., 1987. A semi-empirical low-latitude ionospheric model. Radio Sci. 22 (2), 292–306. https://doi.org/10.1029/
RS022i002p00292.
Arikan, F., Erol, C.B., Arikan, O., 2003. Regularized estimation of vertical total electron
content from Global Positioning System data. J. Geophys. Res. 108 (A12), 1469.
https://doi.org/10.1029/2002JA009605. 12p.
Astafyeva, E., Zakharenkova, I., Förster, M., 2015. Ionospheric response to the 2015 St.
Patrick's Day storm: a global multi‐instrumental overview. J. Geophys. Res. Space
12
Journal of Atmospheric and Solar-Terrestrial Physics xxx (xxxx) xxx–xxx
G.I. Gordiyenko et al.
Hojo, H., Nishizaki, R., 1971. Reduction of topside ionograms for field-aligned propagation paths. Nature 233, 121–123. https://doi.org/10.1038/physci233121b0.
Jakowski, N., Hoque, M.M., Mielich, J., Hall, C., 2017. Equivalent slab thickness of the
ionosphere over Europe as an indicator of long-term temperature changes in the
thermosphere. J. Atmos. Sol. Terr. Phys. 163, 91–102.
Jayachandran, B., Krishnankutty, T.N., Gulyaeva, T.L., 2004. Climatology of ionospheric
slab thickness. Ann. Geophys. 22 (1), 25–33.
Jin, S., Cho, J.H., Park, J.U., 2007. Ionospheric slab thickness and its seasonal variations
observed by GPS. J. Atmos. Sol. Terr. Phys. 69 (15), 1864–1870.
Karpenko, A.L., Leshchenko, L.N., Manaenkova, N.I., 2009. Peculiarities and perspectives
of network digital ionospheric station ‘‘PARUS”. In: Progress in Electromagnetics
Research (PIERS) Symposium Proceedings, Moscow, Russia, August 18–21, pp.
219–222.
Klobuchar, J.A., 1987. Ionospheric time-delay algorithm for single-frequency GPS users.
IEEE Trans. Aero. Electron. Syst. AES-23 (3), 325–331. http://www.dtic.mil/dtic/tr/
fulltext/u2/a229968.pdf.
Kouris, S.S., Polimeris, K.V., Cander, L.R., Ciraolo, L., 2008. Solar and latitude dependence of TEC and SLAB thickness. J. Atmos. Sol. Terr. Phys. 70 (10), 1351–1365.
Kouris, S.S., Polimeris, K.V., Ciraolo, L., Fotiadis, D.N., 2009. Seasonal dependence of TEC
and SLAB thickness. Adv. Space Res. 44 (6), 715–724.
Kumar, S., Tan, E.L., Murti, D.S., 2015. Impacts of solar activity on performance of the
IRI-2012 model predictions from low to mid-latitudes. Earth Planets Space 67 (42),
1–17. https://doi.org/10.1186/s40623-015-0205-3.
Llewellyn, S.K., Bent, R.B., 1973. Documentation and Description of the Bent Ionospheric
Model, Air Force Geophysics Laboratory. Report AFCRL-TR-73-0657. Hanscom AFB,
Massachusetts.
Lühr, H., Xiong, C., 2010. IRI-2007 model overestimates electron density during the 23/
24 solar minimum. Geophys. Res. Lett. 37, L23101. https://doi.org/10.1029/
2010GL045430.
Maltseva, O.A., Zhbankov, G.A., Mozhaeva, N.S., 2013. Advantages of the new model of
IRI (IRI-Plas) to simulate the ionospheric electron density: case of the European area.
Adv. Radio. Sci. 11 (GHJ. 2), 307–311.
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 (8), 2086–2098.
Mosert, M., Magdaleno, S., Buresova, D., Altadill, D., Gende, M., Gularte, E., Scida, L.,
2013. Behavior of the equivalent slab thickness over three European stations. Adv.
Space Res. 51 (4), 677–682.
Nava, B., Rodríguez‐Zuluaga, J., Alazo‐Cuartas, K., Kashcheyev, A., Migoya‐Orué, Y.,
Radicella, S.M., Amory‐Mazaudier, C., Fleury, R., 2016. Middle‐and low‐latitude ionosphere response to 2015 St. Patrick's Day geomagnetic storm. J. Geophys. Res.:
Space Phys. 121 (4), 3421–3438.
Okoh, D., Onwuneme, S., Seemala, G., Jin, S., Rabiu, B., Nava, B., Uwamahoro, J., 2018.
Assessment of the NeQuick-2 and IRI-Plas 2017 models using global and long-term
GNSS measurements. J. Atmos. Sol. Terr. Phys. 170, 1–10. https://doi.org/10.1016/j.
jastp.2018.02.006.
Orus, R., Hernández-Pajares, M., Juan, J.M., Sanz, J., Aragon, A., Garcia-Rigo, A., 2010.
Real time application of TOMION model. In: The International Beacon Satellite
Symposium BSS2010, 4p.
Panda, S.K., Gedam, S.S., Rajaram, G., 2015. Study of Ionospheric TEC from GPS observations and comparisons with IRI and SPIM model predictions in the low latitude
anomaly Indian subcontinental region. Adv. Space Res. 55 (8), 1948–1964.
Sezen, U., Arikan, F., Arikan, O., Ugurlu, O., Sadeghimorad, A., 2013. Online, automatic,
near‐real time estimation of GPS‐TEC: IONOLAB‐TEC. Space Weather 11 (5),
297–305.
Sezen, U., Gulyaeva, T.L., Arikan, F., 2017. Online international reference ionosphere
extended to plasmasphere (IRI-Plas) model. In: 32nd URSI GASS, Montreal, 19–26
August 2017.
Sezen, U., Gulyaeva, T.L., Arikan, F., 2018. Performance of solar proxy options of IRI‐Plas
model for equinox seasons. J. Geophys. Res.: Space Phys. 123 (2), 1441–1456.
https://doi.org/10.1002/2017JA024994.
Stankov, S.M., Warnant, R., 2009. Ionospheric slab thickness–analysis, modelling and
monitoring. Adv. Space Res. 44 (11), 1295–1303.
Stankov, S., Jodogne, J.-C., Kutiev, I., Stegen, K., Warnant, R., 2012. Evaluation of automatic ionogram scaling for use in real-time ionospheric density profile specification: Dourbes DGS-256/ARTIST-4 performance. Ann. Geophys. 55 (2), 283–291.
https://doi.org/10.4401/ag-4976.
Titheridge, J.E., 1985. Ionogram Analysis with the Generalised Program POLAN. Report
UAG-93. University of Auckland, New Zealand.
Venkatesh, K., Rao, P.V.S., Prasad, D.S.V.V.D., Niranjan, K., Saranya, P.L., 2011. Study of
TEC, slab-thickness and neutral temperature of the thermosphere in the Indian low
latitude sector. Ann. Geophys. (09927689) 29 (9).
Verhulst, T., Stankov, S.M., 2013. The topside sounder database – data screening and
systematic biases. Adv. Space Res. 51, 2010–2017.
Wang, N., Yuan, Y., Li, Z., Huo, X., 2016. Improvement of Klobuchar model for GNSS
single-frequency ionospheric delay corrections. Adv. Space Res. 57, 1555–1569.
https://doi.org/10.1016/j.asr.2016.01.010.
Zakharenkova, I.E., Cherniak, IuV., Krankowski, A., Shagimuratov, I.I., 2015. Vertical
TEC representation by IRI 2012 and IRI Plas models for European midlatitudes. Adv.
Space Res. 55, 2070–2076. https://doi.org/10.1016/j.asr.2014.07.027.
Zou, L., Rishbeth, H., Muller-Wodarg, I.C.F., Aylward, A.D., Millward, G.H., FullerRowell, T.J., Idenden, D.W., Moffett, R.J., 2000. Annual and semiannual variations in
the ionospheric F2-layer. I. Modelling. Ann. Geophys. 18, 927–944.
Phys. 120 (10), 9023–9037.
Bi, T., An, J., Liu, Sh, 2017. A modified Klobuchar model for single-frequency GNSS users
over the polar region. Adv. Space Res. 59 (3), 833–842. https://doi.org/10.1016/j.
asr.2016.10.029.
Bilitza, D., Reinisch, B.W., 2008. International reference ionosphere 2007: improvements
and new parameters. Adv. Space Res. 42, 599–609. https://doi.org/10.1016/j.asr.
2007.07.048.
Bilitza, D., 2009. Evaluation of the IRI-2007 model options for the topside electron
density. Adv. Space Res. 44 (6), 701–706. https://doi.org/10.1016/j.asr.2009.04.
036.
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, 1–12. https://doi.org/10.
1051/swsc/2014004.
Blanch, E., Marsal, S., Segarra, A., Torta, J.M., Altadill, D., Curto, J.J., 2013. Space
weather effects on Earth's environment associated to the 24–25 October 2011 geomagnetic storm. Space Weather 11 (4), 153–168.
Bolaji, O.S., Oyeyemi, E.O., Adewale, A.O., Wu, Q., Okoh, D., Doherty, P.H., et al., 2017.
Assessment of IRI-2012, NeQuick-2 and IRI-Plas 2015 models with observed equatorial ionization anomaly in Africa during 2009 sudden stratospheric warming event.
J. Atmos. Sol. Terr. Phys. 164, 203–214.
Chuo, Y.J., Lee, C.C., Chen, W.S., 2010. Comparison of ionospheric equivalent slab
thickness with bottomside digisonde profile over Wuhan. J. Atmos. Sol. Terr. Phys. 72
(5–6), 528–533.
Danilov, A., 2017. New results in studying foF2 trends. J. Atmos. Sol. Terr. Phys. 163,
103–113.
Davies, K., Liu, X.M., 1991. Ionospheric slab thickness in middle and low latitudes. Radio
Sci. 26 (04), 997–1005.
Data on Topside Ionosphere, 1970. Electron densities and scale heights from Alouette II
observations over Japan during October, November and December 1966. In:
Prepared by Radio Research Laboratories Ministry of Posts and Telecommunications.
Tokyo, Japan, vol. 1 135 p.
Data on Topside Ionosphere, 1974. Electron densities and scale heights from Alouette II
observations over Japan during 1968. In: Prepared by Radio Research Laboratories
Ministry of Posts and Telecommunications. Tokyo, Japan, vol. 5 242 p.
Data on Topside Ionosphere, 1977. Electron densities and scale heights from Alouette II
observations over Japan during 1969. In: Prepared by Radio Research Laboratories
Ministry of Posts and Telecommunications. Tokyo, Japan, vol. 6 220 p.
Ezquer, R.G., Scidá, L.A., Orué, Y.M., Nava, B., Cabrera, M.A., Brunini, C., 2017.
NEQUICK 2 and IRI PLAS VTEC predictions for low latitude and South American
sector. Adv. Space Res. https://doi.org/10.1016/j.asr.2017.10.003.
Galav, P., Sharma, S., Pandey, R., 2012. Observations of TEC depletions over Indian low
latitudes. Adv. Space Res. 49, 686–693. https://doi.org/10.1016/j.asr.2011.11.023.
Gordiyenko, G., Yakovets, A., 2017. Comparison of mid-latitude ionospheric F region
peak critical frequencies, heights and topside Ne profiles from IRI2012 model prediction with ground based ionosonde and Alouette II observations. Adv. Space Res.
60, 461–474. https://doi.org/10.1016/j.asr.2017.01.006.
Gulyaeva, T.L., 2003a. International standard model of the Earth's ionosphere and plasmasphere. Astron. AstroPhys. Trans. 22 (4), 639–643. https://doi.org/10.1080/
10556790310001722410.
Gulyaeva, T.L., 2003b. Variations of the half-width of the topside ionosphere according to
the observations by space ionosondes ISIS 1, ISIS 2, and Ik 19. Int. J. Geomagn.
Aeron. 4 (3), 201–207.
Gulyaeva, T.L., Huang, X., Reinisch, B.W., 2002. Plasmaspheric extension of topside
electron density profiles. Adv. Space Res. 29 (6), 825–831. https://doi.org/10.1016/
S0273-1177(02)00038-8.
Gulyaeva, T.L., Titheridge, J.E., 2006. Advanced specification of electron density and
temperature in the IRI ionosphere-plasmasphere model. Adv. Space Res. 38 (11),
2587–2595. https://doi.org/10.1016/j.asr.2005.08.045.
Gulyaeva, T.L., Arikan, F., Stanislawska, I., 2011. Inter-hemispheric imaging of the ionosphere with the upgraded IRI-Plas model during the space weather storms. Earth
Planets Space 63 (8), 929–939.
Gulyaeva, T.L., 2011. Storm time behavior of topside scale height inferred from the ionosphere–plasmasphere model driven by the F2 layer peak and GPS-TEC observations. Adv. Space Res. 47 (6), 913–920.
Gulyaeva, T.L., 2012. Empirical model of ionospheric storm effects on the F2 layer peak
height associated with changes of peak electron density. J. Geophys. Res.: Space
Phys. 117 (A2).
Gulyaeva, T.L., Bilitza, D., 2012. Towards ISO standard Earth ionosphere and plasmasphere model. In: Larsen, R.J. (Ed.), New Developments in the Standard Model.
NOVA, Hauppauge, New York, pp. 1–39.
Gulyaeva, T.L., Arikan, F., Hernandez-Pajares, M., Stanislawska, I., 2013. GIM-TEC
adaptive ionospheric weather assessment and forecast system. J. Atmos. Sol. Terr.
Phys. 102, 329–340.
Guo, P., Xu, X., Zhang, G.X., 2011. Analysis of the ionospheric equivalent slab thickness
based on ground-based GPS/TEC and GPS/COSMIC RO measurements. J. Atmos. Sol.
Terr. Phys. 73 (7–8), 839–846.
Hernández-Pajares, M., Juan, J.M., Sanz, J., Orus, R., Garcia-Rigo, A., Feltens, J.,
Komjathy, A., Schaer, S.C., Krankowski, A., 2009. The IGS VTEC maps: a reliable
source of ionospheric information since 1998. J. Geodes. 83 (3), 263–275.
Hernández-Pajares, M., on behalf of IGS RT WG UPC-IonSAT, Barcelona (Spain), June 23,
2014. Inputs received from Ionospheric research groups on activities related with RT
global electron content determination. In: IGS RT WG Splinter Meeting, Pasadena,
CA, USA.
13
Документ
Категория
Без категории
Просмотров
2
Размер файла
1 957 Кб
Теги
007, jastp, 2018
1/--страниц
Пожаловаться на содержимое документа