JOURNAL OF GEOPHYSICAL RESEARCH: SPACE PHYSICS, VOL. 118, 45944608, doi:10.1002/jgra.

50413, 2013

Global TEC maps based on GNSS data: 1. Empirical background

TEC model
P. Mukhtarov,1 D. Pancheva,1 B. Andonov,1 and L. Pashova 1
Received 17 January 2013; revised 18 April 2013; accepted 19 June 2013; published 24 July 2013.

[1] A global background total electron content (TEC) model is built by using the Center for

Orbit Determination of Europe (CODE) TEC data for full 13 years, 19992011. It describes
the climatological behavior of the ionosphere under both its primary external driver, i.e., the
direct photo-ionization by incident solar radiation, and regular wave particularly tidal
forcing from the lower atmosphere. The model construction is based on the very different
time scales of the solar cycle, seasonal, and diurnal TEC variabilities (at least an order of
magnitude); this leads to modulations of shorter-period variabilities with periods of the
longer ones. Then the TEC spatial-temporal variability is presented as a multiplication of
three separable functions. The solar activity is described by both parameters: F10.7 and its
linear rate of change KF while the seasonal variability is presented by sine functions
including four subharmonics of the year. The diurnal variability of the TEC model is
described by 2D (longitude-time) sine functions with zonal wave numbers up to 4 and 4
subharmonics of the solar day. The model offers TEC maps which depend on geographic
coordinates (5
5 in latitude and longitude) and UT at given solar activity and day of the
year. The presented background model ts to the CODE TEC input data with a zero
systematic error and an RMS error of 3.387 TECU. It is able to reproduce the well-known
ionospheric structures as Weddell Sea Anomaly and some longitudinal wave-like structures.
Citation: Mukhtarov, P., D. Pancheva, B. Andonov, and L. Pashova (2013), Global TEC maps based on GNSS data: 1.
Empirical background TEC model, J. Geophys. Res. Space Physics, 118, 45944608, doi:10.1002/jgra.50413.

1.

Introduction

[2] The conventional sources of ionospheric structure and

variability are related to changes in solar extreme ultraviolet
(EUV) and ultraviolet (UV) radiation and geomagnetic
activity, together with the subsequent response of the thermosphere-ionosphere system and interaction between the components [Roble, 1995]. The ionosphere also varies in response to
neutral winds that move ions and electrons along inclined magnetic eld lines [Schunk and Nagy, 2009], electrodynamical
coupling with the overlying plasmasphere and magnetosphere
[Huba et al., 2005], and dynamical coupling with the underlying atmosphere particularly effective during low solar activity
conditions [Mendillo et al., 2002; Rishbeth, 2006]. The high
sensitivity of the ionosphere to the external forcing and the continuous action of the lower atmospheric forcing causes signicant ionospheric variability on different time and spatial scales.
To understand and forecast such variability is one of the main
tasks of space weather research.
This article is a companion to Mukhtarov et al. [2013], doi:10.1002/
jgra.50412.
1
National Institute of Geophysics, Geodesy and Geography, Bulgarian
Academy of Sciences, Soa, Bulgaria.
Corresponding author: P. Mukhtarov, National Institute of Geophysics,
Geodesy and Geography, Bulgarian Academy of Sciences, Soa, Bulgaria.
(dpancheva@geophys.bas.bg)
2013. American Geophysical Union. All Rights Reserved.
2169-9380/13/10.1002/jgra.50413

[3] Ionospheric models have been playing an important role

for many years in specifying the ionospheric environment
through which the radio waves propagate as realistically as
possible. These ionospheric models can be classied into three
main categories [Feltens et al., 2011]: (i) rst principles
models (physical models) based on ionospheric physics and
chemistry [Schunk et al., 1986; Fuller-Rowell et al., 1987];
(ii) parametric models which simplify the physical models
reducing the number of parameters [Anderson et al., 1989;
Daniell et al., 1993a, 1993b], and (iii) empirical models based
on observations [Hedin et al., 1996; Drob et al., 2008].
The present two companion papers are devoted to empirical
modeling and that is why this type of models is going to
be discussed further. Ionospheric empirical models are
established on statistical analysis of long-term data sets and
utilize appropriate functions to describe the intrinsic variations
of the ionospheric parameters. They usually represent the main
features of the ionosphere quite well, but are limited to the way
the model was constructed, the used data, and the conditions
occurring while the data set was taken [Ercha et al., 2012].
These models are often formulated in terms of monthly
median parameters and represent the state of the ionosphere
for a particular epoch of solar activity. Hence, they can
describe and predict the long-time average conditions of the
ionosphere and are usually used for planning radio system
operations. A typical example of a monthly median empirical
model is the International Reference Ionosphere (IRI) [Bilitza,
2001; Bilitza and Reinisch, 2008].

4594

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

[4] Among ionospheric parameters, the total electron

content (TEC) is one of the particularly important physical
quantities of the ionosphere. TEC has been extensively investigated and modeled for both scientic research of
ionosphere and for applications. The main reason for the
TEC importance is that the trans-ionospheric radio signals,
used by the Global Navigation Satellite Systems (GNSS),
may experience quiet large range errors, and these errors
are proportional to the integral of the electron density along
the ray path, i.e., proportional to slant TEC. In this way, the
ionospheric effect has become the largest error source in
GNSS positioning, timing, and navigation. Hence, the accuracy of the GNSS such as the Global Positioning System
(GPS), the Russian GLONASS, the Chinese BeiDou, and
the European Galileo is heavily affected by the ionosphere.
Under normal solar activity conditions, the ionospheric inuence on GPS signals is usually in the range from a few meters
to tens of meters, but it could reach more than 100 m during
severe ionosphere storms [Gao and Liu, 2002]. Using the
dispersive properties of the ionosphere, it is possible to
eliminate most of this error, or so called rst-order range
error, by differential measurements in dual-frequency systems like GPS, 1575.42 MHz at L1 and 1227.60 MHz at
L2. However, ionosphere dual-frequency algorithms used
for positioning applications remove only rst-order range error [Brunner and Gu, 1991] but do not take into account its
higher-order terms. Also, the ray paths and TEC are assumed
to be the same for both frequencies which is away from the
reality particularly considering the horizontal gradients of
the ionosphere electron density [Kashcheyev et al., 2012].
Additionally, there are still numerous single-frequency applications which need additional information for mitigating the
ionospheric propagation error. Such GNSS users can be provided by adequate ionospheric corrections obtained by an autonomous ionospheric TEC model (without any ionospheric
measurements). Recently, for precise GNSS applications,
higher-order ionospheric terms are considered and taken into
account especially at higher solar activity [Hoque and
Jakowski, 2007; Elmas et al., 2011].
[5] TEC empirical models can be constructed by the following two different ways: (i) from existing empirical
models of the electron density distribution such as IRI
[Bilitza and Reinisch, 2008] or NeQuick [Hochegger et al.,
2000; Radicella and Leitinger, 2001; Nava et al., 2008],
which describes the prole of electron density and is driven
by geographic/geomagnetic location, F10.7, and geomagnetic activity level, and (ii) by using different TEC measurements at single, regionally and globally distributed sites. Most
early empirical TEC models were constructed on the basis of
three different techniques for measuring the ionospheric TEC:
Faraday rotation, Differential Doppler, and Group Delay
methods. Some details about the models built on the TEC data
from the above mentioned techniques can be found in Mao
et al. [2008] and Ercha et al. [2012] and the references therein.
[6] Recently, the GPS measurements obtained from the
global and regional networks of International GNSS
Service (IGS) ground receivers have become a major source
of TEC data over large geographic areas [Wilson et al., 1995;
Komjathy, 1997; Mannucci et al., 1998; Hernndez-Pajares
et al., 1999; Ors et al., 2005]. The GPS has become a powerful measurement tool to investigate the global and regional
ionospheric structures owing to its continuous, easy

operation, and worldwide distributed receivers. This system

offers low-cost information characterized by its accuracy,
high temporal and spatial resolution, and availability. The
GPS technique has great advantage in producing real-time
global ionosphere maps (GIMs) and regional ionosphere
maps (RIMs) of TEC distribution by using different methods
[Iijima et al., 1999; Ping et al., 2002; Otsuka et al., 2002;
Meggs et al., 2004; Ors et al., 2005; Stolle et al., 2005;
Fuller-Rowell et al., 2006; Zapfe et al., 2006; Sayin et al.,
2008]. Therefore, many new TEC empirical models based
on the instantaneous snapshots of the global/regional ionospheric TEC represented by GIMs/RIMs have been well
developed. For example, Mao et al. [2008] built a climatological model of TEC over China using nine years (19962004)
of GPS data and utilizing empirical orthogonal function
(EOF) analyses. Lean et al. [2011a] reported a linear model
that includes solar EUV irradiance variations (0103 nm), four
oscillations (at semiannual, annual, terannual, and biennial periods), geomagnetic activity, and a positive long-term trend
[Lean et al., 2011b] which capture more than 98% of the variance in the GPS-derived daily averaged global TEC during
the period from 1995 to 2010. Using historical estimates of solar EUV irradiance and geomagnetic activity, this model also
facilitates the reconstruction of TEC climatology in the past,
which has been illustrated since 1950. Ercha et al. [2012]
constructed a global empirical TEC model based also on
EOF analysis and by using the GIMs provided by Jet
Propulsion Laboratory (JPL) during the years 19992009.
Bouya et al. [2010] used Spherical Cap Harmonic Analysis
to construct a regional ionospheric TEC model over
Australia. Habarulema et al. [2010, 2011] built regional
GPS-based TEC models over Southern Africa by using a neural network analysis method. Wan et al. [2012] developed a
global ionospheric TEC model using a statistical Eigen mode
analysis method. In this way, the GPS-based ionospheric
TEC data have been extensively utilized recently in ionospheric studies to analyze and validate the ionospheric models
and for space weather monitoring applications.
[7] The basic aim of the present paper (Part 1) is to present a
global empirical background TEC model which is built by
using long-term TEC data from the Center for Orbit
Determination of Europe (CODE) [Schaer, 1999]. The main
purpose is to describe the mean behavior of the ionosphere under both its primary external driver, i.e., the direct photo-ionization by incident solar radiation, and regular wave particularly
tidal forcing from the lower atmosphere. Moreover, we aim
to make this model applicable under quiet geomagnetic conditions for long-term prediction of the average TEC variability.
Not only a detailed statistical evaluation of the global empirical
background TEC model is accomplished in Part 2, but an error
model is established as well; the latter differentiates the present
empirical model from others.

2.

TEC Data Set

[8] The ionospheric TEC is derived by mapping the slant

path delay of the signal from dual-frequency L1 and L2
bands observed by the global networks of IGS ground receivers [Ge et al., 2004; Dow et al., 2009]. IGS provides
the highest precision of GPS satellite orbits, and precise
positions (5 mm) for 350 worldwide reference stations
[Ercha et al., 2012]. Usually, the single ionospheric layer

4595

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

assumption is considered to convert the slant path TEC to

vertical TEC with a mapping function. Currently, ve analysis centers routinely provide GIMs of vertical TEC using the
growing global network of dual-frequency GNSS receivers
[Ercha et al., 2012]. These are CODE [Schaer, 1999], JPL
[Ho et al., 1996], European Space Agency [Feltens and
Schaer, 1998; Feltens, 2007], Polytechnical University of
Catalonia [Hernndez-Pajares et al., 1997], and the Energy
Mines and Resources Canada [Gao et al., 1994].
[9] It has been already mentioned that the present background TEC model is constructed on the base of TEC maps
generated by the CODE at Astronomical and Physical
Institutes of the University of Bern (http://cmslive3.unibe.ch/
unibe/philnat/aiub/content/e15/e59/e126/e440/e447/index_eng.
html). For the current study, we chose data for full 13 years, 1
January 1999 31 December 2011, provided from the
CODE FTP directory: ftp://ftp.unibe.ch/aiub/CODE/. At
CODE, the vertical TEC is modeled with a spherical harmonic
expansion up to degree of order 15 referring to a solar-geomagnetic reference frame [Schaer, 1999]. The two-hourly sets
are derived from GPS data of the global IGS network of about
200 stations. The GIM/CODE is regarded as one of the most
precise TEC maps generated from GNSS observations. The
used global IGS TEC data have a time resolution of 2 h and
a grid spacing of 5
2.5 in longitude and latitude,
respectively, with errors of several TEC Units (TECU, 1
TECU = 1016 el/m2) [Hernndez-Pajares et al., 2009]. The
errors of the GIM/CODE are determined by comparing with
an independent source of TEC. The reference TEC values
are provided by dual-frequency altimeters on board of
TOPEX, JASON, and ENVISAT satellites. Because the
altimeters are working over oceans, this comparison is considered as a pessimistic determination of the global TEC map
actual errors.
[10] The original global TEC data were arrayed in terms of
the coordinate system of geographical latitude (from 87.5
to 87.5 at each 2.5) and longitude (from 180 to 180 at
each 5). It is known, however, that the neutral wind and
electric eld effects on the ionosphere are dependent on the
geomagnetic eld conguration as the electrons are
constrained to the magnetic eld lines. That is why the distribution of the ionospheric parameters, including TEC as well,
is usually presented in geomagnetic latitude instead of geographic one. Early investigations [reported, e.g., in Rawer,
1984] demonstrated the benet of using the modied dip
(modip) latitude, introduced by Rawer [1963], to describe
the variability of the densest part of the ionosphere, particularly at mid and low latitudes. The modip latitude which is
adapted to the real magnetic eld, e.g.,
. to the magnetic incliI p
, where is the
nation (dip), is dened by: tan
cos

modip latitude, I is the true magnetic dip (usually at a height

of 350 km), and is the geographic latitude. Modip equator
is the locus of points where the magnetic dip (or inclination)
is 0. In the equatorial zone, the lines of constant modip are
practically identical to those of the magnetic inclination,
but as latitude increases, they deviate and come nearer to
those of constant geographical latitude. The poles are identical to the geographic ones [Azpilicueta et al., 2006]. Then for
the purpose of this study, the global TEC data were rearrayed in terms of the coordinate system of modip latitude,
from 80 to 80 at each 5, and geographic longitude, from

180 to 180 at each 15. The TEC data falling into the area
5 (modip latitude)
15 (longitude) were averaged.
[11] Usually the background ionospheric models are formulated in terms of monthly median parameters because
such parameters are not affected by large but short-time lasting disturbances generated by strong geomagnetic storms. In
the present study, we use sliding medians dened by a 31 day
moving window, and the median value is assigned to the central day of the window, i.e., to the 16th day of the window.
The sliding medians are calculated independently for each
point of the grid (as it is done with single station data). In this
way, the daily TEC time series are obtained at each modip
latitude, longitude, and UT. It is worth noting that the perturbations from geomagnetic origin or related to solar rotation
period are ltered from these time series.

3.

Basic Approach of the Model Construction

[12] The basic idea of each global empirical background

TEC model, used for long-term prediction, is to dene a set
of empirical functions which describes the most probable
TEC values at given solar activity, day of the year (DOY),
geographic location, and UT. In the present study, we accepted: (i) longitude and UT as independent variable quantities; the conversion into local time (LT) is a simple
procedure, and (ii) at each modip latitude, a separate model
is constructed; the values of the model TEC which do not belong to the 5 modip grid are obtained by an interpolation
procedure that will be described later. The latter is done because if a latitudinal approximation is used, rst the number
of model constants will increase, and second an additional error will be introduced in the model.
[13] According to the above mentioned approach, the TEC
values at each modip latitude circle can be presented as a
function of:
TEC solar activity; DOY ; longitude; UT

(1)

[14] Ideally, the solar activity should be described by an index that tracks the solar cycle changes in the EUV wavelength range, since this part of the solar spectrum affects
the ionosphere (Bilitza, 2001). However, such indices cannot
be observed at the ground and are only available for relatively short time periods covered by satellite UV instruments.
Actually recently Lean et al. [2011c] have reported new
models (two and three components) of solar EUV irradiance
variability (from 0 to 120 nm), by using multiple regression
of the Mg II and F10.7 solar activity indices with irradiance
observations made during the descending phase of solar cycle 23. The authors used the Solar EUV Experiment on the
Thermospheric Ionospheric Mesospheric Energy and
Dynamics (TIMED) spacecraft observations of solar EUV irradiance. The models have been used to reconstruct EUV
spectra daily since 1950, annually since 1610, to forecast
daily EUV irradiance, and to estimate future levels in cycle
24. These models, however, have not yet been openly accessible for the scientic community. Thus, most ionospheric
modelers use the sunspot number (number of dark spots on
the solar disc) and the solar radio ux at 10.7 cm wavelength
(F10.7) as solar indices, since both can be observed from the
ground, long data records exist, and they can be predicted.
These indices together with their 6 month predictions are

4596

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

4

200

1999-2011
F10.7
KF

180

140

120

-2

KF

F10.7

160

100
-4

80

-6

60
1999

2001

2003

2005

2007

2009

2011

Figure 1. Monthly mean solar radio ux F10.7 (red line)

and its linear rate of change KF (blue line) obtained by a sliding 11 month time derivatives of F10.7.
regularly published by NOAA Space Weather Prediction
Center (http://www.swpc.noaa.gov/index.html). Some preliminary experiments were preformed with both indices,
and the results revealed that particularly for the considered
last solar cycle F10.7 describes better the TEC variability.
Thus, in the present study, F10.7 is used as a proxy for the solar activity. It is known, however, that the ionosphere behaves differently at the rising and declining phases of the
solar cycle at one and the same F10.7, i.e., at one and the
same F10.7, the TEC values in the rising and in the falling
part of the solar cycle are different [Huang, 1963; Gopal
Rao and Sambasiva Rao, 1969; Apostolov et al., 1994]. To
include this ionospheric feature in the model, an additional
parameter KF is used which describes the linear rate of
change of F10.7. The idea for describing the solar activity
by such two parameters, i.e., by the level of solar activity
and its tendency, was introduced for the rst time by
Pancheva and Mukhtarov [1996] in modeling the monthly
median critical frequency of the ionospheric F region, foF2,
above Soa. Later, this idea was used for long-term prediction
of other ionospheric parameters, and the model approach was
successfully applied to other European ionosonde stations
[Pancheva and Mukhtarov, 1998]. It was found that the inclusion of the KF in the monthly median foF2 model decreased
its mean standard deviation by ~0.5 MHz. Figure 1 shows the
temporal variability of the used two solar parameters F10.7 in
solar ux units (1022 W m2 Hz1) and KF for the considered
13 years (19992011). It can be distinguished that this solar
cycle is characterized by double-peak solar maximum; while
the rst peak occurs in 2000, the second one is seen in 2002.
[15] The seasonal variability of the vertical TEC is characterized mainly by annual and semiannual components [Natali
and Meza, 2011] whose amplitudes and phases depend on
solar activity, geographic/geomagnetic location and UT.
The seasonal components with periods shorter than 6 months
have also some contribution, but they are weaker than annual
and semiannual components. In general, the semiannual
behavior is characterized by a larger peak in MarchApril
than that in SeptemberOctober; the difference is particularly
strong (up to 40%) during high solar activity [Natali and
Meza, 2011].
[16] In all existing so far empirical models, the diurnal
TEC variability is described only by the migrating tidal
components, i.e., components with periods 24, 12, 8 h, etc.,
which propagate with the apparent motion of the Sun to a

ground-based observer. In this way, it is assumed that the

diurnal variability of the ionosphere is forced mainly by the
diurnal cycle of the photo-ionization. However, only within
the past 56 years has the realization emerged that troposphere weather contributes signicantly to the space
weather of the thermosphere, especially during solar minimum conditions. Much of the attendant variability is attributable to upward-propagating solar tides excited by latent
heating due to deep tropical convection, and solar radiation
absorption primarily by water vapor and ozone in the troposphere and stratosphere, respectively. Recent studies based
on the modern satellite-board data (electron densities from
COSMIC and temperatures from TIMED) revealed that
while the ionospheric migrating semidiurnal tidal (SW2, S
for semidiurnal, W means westward, and 2 claries the wave
number) response is predominantly shaped by the migrating
SW2 tide forced from below, the migrating diurnal tidal
(DW1, D for diurnal) response is mainly due to daily variability of the photo-ionization [Pancheva and Mukhtarov,
2012]. The nonmigrating tides which amplify in the lower
thermosphere are particularly important for the ionosphere.
Through coupling mechanisms (such as the wind dynamo),
they can play an important role in generating the longitudinal
variability in the ionosphere. A special attention has been
paid recently on the nonmigrating tides DE3 and DE2 (diurnal eastward propagating tides with zonal wave numbers, respectively, 3 and 2) as a link between the troposphere and
ionosphere [Immel et al., 2006; Kil et al., 2008; Hartman
and Heelis, 2007; Ren et al., 2008; Wan et al., 2008;
Pancheva and Mukhtarov, 2010]. Mukhtarov and
Pancheva [2011] investigated in detail the ionospheric response to the above mentioned nonmigrating tides by using
COSMIC global electron density data. Later, their experimental results have been supported by the numerical simulations done with the recently developed Earths whole
atmospheric model from the troposphere to the ionosphere,
called GAIA [Pancheva et al., 2012]. The above mentioned
effects of the tidal forcing which induce longitude variability
are not yet emulated in empirical ionospheric models. In the
present study, we include these inuences.
[17] There is another reason also for including the
nonmigrating tidal variability in the TEC model. It is related
to the offset between the geographic and modip latitudes. The
dynamics of the thermosphere (dened mainly by prevailing
winds and atmospheric tides) is dened in a geographic frame,
but its effect on the ionosphere depends on the geomagnetic
eld conguration. The photo-ionization depends also on geographic frame (solar zenith angle); however, as the electrons
are constrained to the magnetic eld lines, its effect is geomagnetic eld dependent. The offset between geographic and
geomagnetic frames generates additional tidal ionospheric
responses which can originally not be present in the forcing
from below. These additional ionospheric tides are much
weaker than the real ones (on the average ~310%), but they
have some contribution to the diurnal variability of the ionosphere particularly in shaping some quasi-stationary structures.
[18] As the time scales of the solar cycle, seasonal, and diurnal inuences on the TEC variability are very different
(they differ at least an order of magnitude), then the
shorter-period TEC variabilities are usually modulated by
the longer ones. In this case, the TEC spatial-temporal variability can be represented as a multiplication of three

4597

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

Mean TEC (1999-2011)

Modip Latitude (degree)

80
80

60

70

40

60

20

50

40

-20

30
20

-40

10

-60
-80
12

24

36

48

60

72

84

96 108 120 132 144 156

Month Number (start January 1999)

Amp. DW1 TEC (1999-2011)
Modip Latitude (degree)

80
60
40
20

50

60

40

40

30

9
7

20
5

0
20

-20

10

-40

-20

-40
1

-60

-60

-80

-80
12

24

36

48

60

72

84

96 108 120 132 144 156

12

Amp. SW2 TEC (1999-2011)

24

36

60

72

84

96 108 120 132 144 156

60

11

60

11

40

40

20

20

-20

-20
3

-40
1

-60

-40

-60
-80

-80
12

24

36

48

60

72

84

12

96 108 120 132 144 156

24

36

Amp. TW3 TEC (1999-2011)

48

60

72

84

96 108 120 132 144 156

Amp. DE3 TEC (1999-2011)

80

80

Modip Latitude (degree)

48

Amp. D0 TEC (1999-2011)

80

80

Modip Latitude (degree)

Amp. SPW1 TEC (1999-2011)

80

60

60

40

40

20

20

-20

-40

-20

-40
-60

-60
-80

-80
12

24

36

48

60

72

84

96 108 120 132 144 156

Month Number (start January 1999)

12

24

36

48

60

72

84

96 108 120 132 144 156

Month Number (start January 1999)

Figure 2. Latitude-time cross sections of the following TEC components: (i) zonal and time mean TEC
(upper most plot); (ii) left column of plots shows the amplitudes of migrating diurnal (DW1, upper plot),
semidiurnal (SW2, middle plot), and terdiurnal (TW3, bottom plot) tides, and (iii) right column of plots
shows the amplitudes of SPW1 (upper plot), nonmigrating zonally symmetric diurnal tide (D0, middle plot)
and nonmigrating eastward propagating tide with zonal wave number 3 (DE3, bottom plot). The considered
years from 1999 to 2011 are separated by thin white lines; the TEC components are measured in TECU.

4598

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

separable functions, i.e., (1) can be expressed in the following way [Pancheva et al., 2005]:
TEC F; K F ; DOY ; lon; UT
1 F; K F 2 DOY 3 lon; UT

[19] The above right-hand side unknown functions k

(k = 1,2,3) can be represented by their series expansions;
1 can be expanded in Taylor series, while 2 and 3,
which are periodic functions with periods, respectively, a
year and a solar day, can be expanded in Fourier series.
Therefore, the background TEC model can be described by
the following functions:
TEC F; K F ; DOY ; lon; UT


0 1 F 2 K F 3 FK F 4 F 2 5 K F 2



4
2

0 i cos i
day  i
365
i1


4 4
2
2

0 is cos i UT  s lon  is
24 
360
i1s4

4
2
s cos s
lon  is
360
s1

[20] The expression in the rst right-hand bracket, i.e., the

Taylor series expansion up to degree of 2, represents the solar
activity term which modulates the seasonal and diurnal behavior of the ionosphere. The seasonal term (expression in
the second right-hand bracket) includes four subharmonics
of the year, i.e., annual, semiannual, 4, and 3 month components; it modulates the diurnal behavior of the ionosphere.
The diurnal variability of the TEC model (expression in the
third right-hand bracket) is composed by three parts: mean
TEC (0), a part describing migrating and nonmigrating tides,
and a part representing the effect of the stationary planetary
waves (SPWs). The contribution of the migrating and
nonmigrating tides is presented by 2D (longitude-time) sine
functions with zonal wave numbers up to 4 and
subharmonics of the solar day with periods 24, 12, 8, and
6 h. The last part, describing the contribution of the SPWs,
includes waves with zonal wave numbers up to 4. The presence of these wave structures in the ionosphere can be related
to a few reasons: (i) offset between geographic and modip
frames; (ii) can be generated by coupling processes between
migrating and nonmigrating tides with one and the same periods [Oberheide et al., 2011], and (iii) some effect of the
SPW1 temperature eld in the lower thermosphere on the
ionosphere, particularly at middle-high and high-latitude
ionosphere; Mukhtarov et al. [2010] found strong evidence
indicating that the auroral heating is a main origin of the
lower thermospheric SPW1 structure.
[21] The background TEC model described by (3) contains
4374 constants, and they are determined by least squares
tting techniques. The numbers of the included components
in the above described Taylor and Fourier expansion series
are dened experimentally. The above mentioned solar, seasonal, and diurnal components have been determined on the
basis of trial and error method by using the following criterion: the addition of more components has been discarded
when their inclusion leads to an error improvement only after
the third decimal point.

[22] Figure 2 presents examples of solar cycle and seasonal

modulations of some diurnal components included in the
TEC model. For this purpose, the TEC data are decomposed
to mean TEC, migrating and nonmigrating tides, and SPWs
(i.e., expressions in the third right-hand bracket of (3)) by
using a 31 day window. Then the 31 day window is moved
forward through the time series with steps of 1 day in order
to obtain the daily values of the wave characteristics for
period of time 1 January 199931 December 2011. The
monthly mean wave characteristics, shown in Figure 2, are
obtained by vector averaging for each calendar month
[Pancheva and Mukhtarov, 2012]. We note that the color
scale of Figure 2 and all other gures is in TECU. The uppermost plot shows the latitude-time cross section of the zonal
and time mean of the TEC (rst term, 0, in the third bracket);
the considered years from 1999 to 2011 are separated by thin
white lines. This diurnal TEC component follows strictly the
solar activity, even the two maxima, a main maximum in
2002 (~80 TECU) and a secondary one in 2000 (~75
TECU), can be clearly distinguished. The semiannual variability is a dominant component of the seasonal behavior;
on the average, the vernal equinox maxima are stronger than
the autumnal ones. The left column of plots present the latitude-time cross sections of the rst three migrating tides, 24
h (DW1, upper plot), 12 h (SW2, middle plot), and 8 h
(TW3, bottom plot). The DW1 component is the strongest
tidal component (maximum amplitude reaches ~50 TECU)
and is shaped mainly by the diurnal variability of the
photo-ionization. The SW2 component is signicantly
weaker than the DW1 one (maximum amplitude reaches
~11 TECU) and as it has been already mentioned is formed
mainly by the SW2 tide forced from the lower atmosphere
[Pancheva and Mukhtarov, 2012]. A clear evidence for the
dominant role of the lower atmospheric SW2 tide on the
TEC SW2 variability is the existence of the local winter maxima at modip latitude around 60. Wu et al. [2011] investigated the mesosphere-lower thermosphere SW2 tide in the
neutral winds measured by the instrument TIDI on board of
the TIMED satellite and found that at high latitudes the
SW2 amplies in winter solstice. The solar cycle and seasonal (mainly semiannual) modulations can be seen well
for these diurnal components as well. While at high solar activity, the separation at both sides of the equator can be well
seen for the mean (0) and DW1 components, for the SW2
component, such separation is evident better at low solar activity; at high solar activity, the SW2 amplies predominantly over the equator. The amplications of the mean
TEC, tidal DW1, and SW2 amplitudes around (2030)
modip are related to the equatorial ionization anomaly
(EIA) observed mainly during the daytime. Figure 2 shows
that the solar activity affects not only the amplitude of the
TEC EIA but also the location of the crests; at high solar activity, they are located close to 30, while at low solar activity, the crests move close to the equator, around 20. The
solar cycle and seasonal modulations are seen also on the
third migrating component, TW3. The latitude structure of
this diurnal component shows a main amplication over the
equator and secondary ones around 50 and 65; the latter
are particularly well seen during high solar activity.
[23] The right column of plots present the latitude-time
cross sections of the amplitudes of the SPW1 (upper plot),
nonmigrating zonally symmetric diurnal (D0, middle plot)

4599

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

12 Dec 2012, 08 UT (TEC Model)

90
70

Latitude (degree)

60
60

30

50
40

30

-30

20
10

-60
-90
-180 -150 -120 -90 -60 -30

30

60

90 120 150 180

12 Dec 2012, 08 UT (TEC Data)

90
70

Latitude (degree)

60

60
50

30

40

0
30

-30

20
10

-60
-90
-180 -150 -120 -90 -60 -30

30

60

90 120 150 180

solar cycle and shows clear semiannual variability. In this

case, however, the autumnal maxima are stronger than the
vernal ones. This is due to fact that the neutral DE3 tide
forced from below is the main driver of the ionospheric
DE3 variability. Recent studies on the temperature and
neutral wind DE3 tide based on the satellite measurements
revealed that this tide amplies at lower thermosphere
(~105110 km) during AugustOctober [Oberheide and
Forbes, 2008; Pancheva et al., 2010]. Pancheva and
Mukhtarov [2010] presented unambiguous evidence that the
ionospheric DE3 variability is forced from below. Later, in
Mukhtarov and Pancheva [2011], on the basis of global electron density data from COSMIC satellites, the authors found
the main spatial structure and temporal variability of the electron density DE3 tide which have been supported by the whole
atmosphere-ionosphere coupling GAIA model [Pancheva
et al., 2012]. We pay special attention to the TEC DE3
nonmigrating diurnal component as it has the main contribution to the so-called wave number 4 (WN4) longitude structure
observed in many ionospheric parameters particularly during
low solar activity [Immel et al., 2006]. The other longitude
wave-like structure, observed mainly in December solstice
conditions, is the WN3 structure; the main contributor for it
is the nonmigrating DE2 tide that forced the ionospheric
DE2 variability [Mukhtarov and Pancheva, 2011]. Later, we
will demonstrate the ability of this background model to reproduce the WN4 and WN3 ionospheric structures.

Longitude (degree)

Figure 3. Global map in geographical frame calculated

from the empirical background TEC model (upper plot) compared with the CODE TEC map (bottom plot) at 08 UT for 12
December 2012; the Weddell Sea Anomaly is apparent between 80 to 50 S and 120 to 0 E.
and nonmigrating eastward propagating diurnal tide with
zonal wave number 3 (DE3, bottom plot). All these diurnal
components demonstrate regular solar cycle and seasonal
variability. Their amplitudes are weaker than the DW1 tide
but are comparable with those of the SW2 and TW3 tides.
The D0 is the strongest nonmigrating component with maximum amplitude of ~11 TECU (the same as that of SW2). It
amplies mainly in the Southern Hemisphere (SH) at high
latitudes; similar distributions have also the other zonally
symmetric tidal components, but their amplitudes are weaker
than that of D0 (not shown here). All zonally symmetric tidal
components show amplications like stripes between 40
and 70 modip latitude that can be distinguished even at
low solar activity. Similar amplications are evident for
SPW components, but they are present at both hemispheres.
Later, it will be shown that just these zonally symmetric
nonmigrating and SPW components have predominant contribution to the so called Weddell Sea Anomaly (WSA) [He
et al., 2009; Chen et al., 2011]. This anomaly appears as an
evening enhancement in electron density, i.e., larger nighttime electron density than during the day, in the region near
the Weddell Sea, Antarctica peninsula. The WSA occurs
mostly in southern summer and can extend from South
America and Antarctica to the central Pacic.
[24] The TEC DE3 component (bottom plot), similarly to
the other diurnal components, is strongly modulated by the

4.

Model Results

[25] The basic aim of each global TEC model used for
long-term prediction is to construct the global distribution
of the TEC, i.e., to obtain global TEC maps, at given solar
activity, DOY, and UT. The TEC maps are constructed by
interpolation of the TEC values from the used grid with a
5 step in modip latitude and 15 in longitude. The interpolation between obtained from the model TEC values is done by
using Inverse Distance to a Power Method. This gridding
method is a weighted average interpolator that can be either
an exact or a smoothing interpolator. The data points are
weighted during interpolation such that the inuence of one
data point relative to another declines with distance from
the grid node. Weighting is assigned to data points through
the use of a weighting power that controls how the weighting
factors drop off as distance from a grid node increases. The
greater the weighting power the less effect points far from
the grid node has during interpolation. For a smaller power,
the weights are more evenly distributed among the neighboring data points [Shepard, 1968]. After completing the interpolation, the modip frame is converted to geographical one.
The TEC values assigned to both poles are found by interpolation between the known from the model points which have
the highest northern and southern latitudes. The model maps
are arrayed in terms of the coordinate system of geographical
latitude from 90 to 90 at each 5 and longitude from
180 to 180 at each 5.
[26] First we will show how the background TEC model
describes the WSA. The zone of anomalous diurnal variations in foF2, which is characterized by an excess of nighttime foF2 values over daytime ones, occupies the
longitudes of 0180W and the latitudes of 4080S as
the effect is maximal (up to ~5 MHz for the critical frequency

4600

Modip Latitude (degree)

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

80

WN4

WN3

October 2008 (TEC Model, 23LT)

January 2008 (TEC Model, 14LT)

80

13

60

28

40

40
20
0
-20

20

30

60

-60

90 120 150 180

30

60

90 120 150 180

January 2008 (TEC Data, 14LT)

12.3

60

13
8

-80
-180 -150 -120 -90 -60 -30

October 2008 (TEC Data, 23LT)

80

18

-40

-60

23

-20

-40
-80
-180 -150 -120 -90 -60 -30

Modip Latitude (degree)

33

60

11

10.3

40

80

35

60

30

40

20

8.3

20

6.3

-20
4.3

-40
-60

2.3

-80
-180 -150 -120 -90 -60 -30

30

60

25
20

-20

15

-40

10

-60

-80
-180 -150 -120 -90 -60 -30

90 120 150 180

30

60

90 120 150 180

Longitude (degree)

Longitude (degree)

Figure 4. Comparison between the global TEC model maps in modip latitude (upper row of plots) and
CODE TEC maps (bottom row of plots) which shows the ionospheric WN4 (left column of plots, at 23
UT for 15 October 2008) and WN3 (right column of plots, at 14 UT for 15 January 2008) structures.

of the F region, foF2) at longitudes of 40105W and latitudes of 6070S [Karpachev et al., 2011]. Figure 3 presents the global map in geographical coordinate system
calculated from the empirical background TEC model (upper
plot) and compared with the CODE TEC map (bottom plot)
at 08UT for 12 December 2012. The stripe TEC amplication in the Western Hemisphere at latitudes of ~5080S, i.
e., the WSA, can be clearly distinguished at both model
and CODE TEC maps; the maximal effect at both maps is
near 70S and longitudes of ~0120W. The presence of
the WSA is a reason for appearing of an additional to the
EIA TEC amplication around 3040S and at the most
Western Hemisphere; this feature is also well reproduced
by the model. The model TEC map describes well also the
hemispheric asymmetry of the EIA revealing that the summer crest is stronger than the winter one.
[27] To demonstrate how the model reproduces some longitude wave-like structures, we re-arrayed the model and
CODE TEC data sets in LT. It has been already mentioned
that usually the ionospheric wave-like longitude structures
are observed during low solar activity as the WN4 is seen
in AugustOctober, while the WN3 in DecemberJanuary.
Figure 4 shows the comparison between the global TEC
model maps in modip latitude (upper row of plots) and
CODE TEC maps (bottom row of plots) which represent
the ionospheric WN4 (left column of plots) and WN3 (right
column of plots) structures. The example for the WN4 structure is for October 2008 at 23LT, i.e., nighttime, and that is
why there is no splitting of the irregularities at both sides of
the equator. Four peaks around longitudes: 150, 60,
30, and 120 can be clearly distinguished at both model
and CODE TEC maps. There is not only qualitative but also
quantitative similarity between the model and observations.

The example for the WN3 structure is for January 2008 at

14LT, i.e., daytime, when the EIA is present. The splitting
of the irregularities at both sides of the equator is seen at both
model and CODE TEC maps. However, the WN3 structure is
well developed and signicantly stronger in SH. This
hemispheric asymmetry is due to the asymmetry of the
ionospheric DE2 variability, reported by Mukhtarov and
Pancheva [2011], which is the main contributor of the
WN3 structure. The three peaks particularly in the SH are
located at longitudes of 150, 0, and 120. They are not
exactly equidistant most probably because the contribution
of other nonmigrating tides, as DW4 and SE1, and SPW3
[Pancheva and Mukhtarov, 2012]. There are some signatures
for the rst and third peaks at NH evident at both model and
CODE TEC maps.
[28] The ability of the background TEC model to reproduce the temporal-spatial feature of the input CODE TEC
data will be displayed by a comparison between the model
and CODE TEC maps for different solar activity, seasons
and UT. While Figure 5a shows the global maps in geographical coordinate system calculated from the empirical
background TEC model (left column of plots) which are
compared with the CODE TEC maps (right column of plots)
at 12UT for 15 January 2001 (upper row of plots) and 15
March 2001 (bottom row of plots) during high solar activity
2001, Figure 5b shows the same but during low solar activity
2008. The modip latitude is also marked at the plots by white
line as the low-latitude plasma bulk follows the modip frame.
At both solar cycle conditions, it is seen that the model maps
reproduce very well the main features of the CODE TEC
maps; some quantitative difference is evident only at winter
model map where the EIA is slightly weaker than that at
the CODE TEC map (Figure 5a, upper row of plots). The

4601

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

TEC Model

TEC Data

15 Jan 2001, 12UT

15 Jan 2001, 12UT

90

90
100
90
80
70
60
50
40
30
20
10
0

30
0
-30
-60
-90
-180 -150 -120 -90 -60 -30

60

Latitude (deg)

Latitude (deg)

60

30
0
-30
-60
-90
-180 -150 -120 -90 -60 -30

30 60 90 120 150 180

Longitude (deg)
15 Mar 2001, 12UT

30 60 90 120 150 180

15 Mar 2001, 12UT

90
110
100
90
80
70
60
50
40
30
20
10
0

30
0
-30
-60
-90
-180 -150 -120 -90 -60 -30

60

Latitude (deg)

60

Latitude (deg)

Longitude (deg)

90

110
100
90
80
70
60
50
40
30
20
10
0

30
0
-30
-60
-90
-180 -150 -120 -90 -60 -30

30 60 90 120 150 180

Longitude (deg)

30 60 90 120 150 180

Longitude (deg)

TEC Model

TEC Data

15 Jan 2008, 12UT

15 Jan 2008, 12UT

90

90

60

60

28
24

30

20
16

12
8

-30

Latitude (deg)

Latitude (deg)

100
90
80
70
60
50
40
30
20
10
0

28
24

30

20
16

12
8

-30

-60

-60

-90
-180 -150 -120 -90 -60 -30

-90
-180 -150 -120 -90 -60 -30

30 60 90 120 150 180

Longitude (deg)

30 60 90 120 150 180

Longitude (deg)

15 Mar 2008, 12UT

15 Mar 2008, 12UT

90

90
32
28
24
20
16
12
8
4
0

30
0
-30
-60
-90
-180 -150 -120 -90 -60 -30

60

Latitude (deg)

60

Latitude (deg)

32
28
24
20
16
12
8
4
0

30
0
-30
-60
-90
-180 -150 -120 -90 -60 -30

30 60 90 120 150 180

Longitude (deg)

30 60 90 120 150 180

Longitude (deg)

Figure 5. (a) Global maps in geographical coordinate system calculated from the empirical background
TEC model (left column of plots) which are compared with the CODE TEC maps (right column of plots) at
12 UT for 15 January 2001 (upper row of plots) and 15 March 2001 (bottom row of plots) during high solar
activity 2001. The modip latitude is also marked by white line. (b) The same as in Figure 5a, but during low
solar activity 2008.
4602

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

TEC Model

TEC Data

15 July 2001, 12UT

15 July 2001, 12UT

90

90
60
60
55
50
45
40
35
30
25
20
15
10
5
0

30
0
-30
-60
-90
-180 -150 -120 -90 -60 -30

Latitude (deg)

Latitude (deg)

60

0
-30
-60
-90
-180 -150 -120 -90 -60 -30

30 60 90 120 150 180

Longitude (deg)
15 Oct 2001, 12UT

30 60 90 120 150 180

Longitude (deg)
15 Oct 2001, 12UT

90

90
60
120

30

100
80

60
40

-30

Latitude (deg)

60

Latitude (deg)

60
55
50
45
40
35
30
25
20
15
10
5
0

30

120

30

100
80

60
40

-30

20

20

-60

-60

-90
-180 -150 -120 -90 -60 -30

-90
-180 -150 -120 -90 -60 -30

30 60 90 120 150 180

Longitude (deg)

30 60 90 120 150 180

TEC Model

TEC Data

15 July 2008, 12UT

15 July 2008, 12UT

90
18
16
14
12
10
8
6
4
2
0

30
0
-30
-60
-90
-180 -150 -120 -90 -60 -30

60

Latitude (deg)

60

Latitude (deg)

Longitude (deg)

90

18
16
14
12
10
8
6
4
2
0

30
0
-30
-60
-90
-180 -150 -120 -90 -60 -30

30 60 90 120 150 180

Longitude (deg)

30 60 90 120 150 180

Longitude (deg)
15 Oct 2008, 12UT

15 Oct 2008, 12UT

90

90

60

60

28
24

30

20
16

12
8

-30

Latitude (deg)

Latitude (deg)

28
24

30

20
16

12
8

-30

-60

-90
-180 -150 -120 -90 -60 -30

-60

-90
-180 -150 -120 -90 -60 -30

30 60 90 120 150 180

Longitude (deg)

30 60 90 120 150 180

Longitude (deg)

Figure 6. (a) Global maps in geographical coordinate system calculated from the empirical background
TEC model (left column of plots) which are compared with the CODE TEC maps (right column of plots) at
12 UT for 15 July 2001 (upper row of plots) and 15 October 2001 (bottom row of plots) during high solar
activity 2001. The modip latitude is also marked by white line. (b) The same as in Figure 6a, but during low
solar activity 2008.
4603

TEC Model

TEC Data

15 Jul 2004, 00UT

15 Jul 2004, 00UT

90

90

60

60

Latitude (deg)

Latitude (deg)

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

30
0
-30
-60

30
0
-30
-60

-90
-180 -150 -120 -90 -60 -30

-90
-180 -150 -120 -90 -60 -30

30 60 90 120 150 180

Longitude (deg)

60

60

0
-30
-60

30
0
-30
-60

-90
-180 -150 -120 -90 -60 -30

-90
-180 -150 -120 -90 -60 -30

30 60 90 120 150 180

30 60 90 120 150 180

Longitude (deg)

TEC Model

TEC Data

15 July 2004, 12UT

15 July 2004, 12UT

90

90

60

60

Latitude (deg)

Latitude (deg)

Longitude (deg)

30
0
-30
-60

30
0
-30
-60

-90
-180 -150 -120 -90 -60 -30

-90
-180 -150 -120 -90 -60 -30

30 60 90 120 150 180

Longitude (deg)

60

60

Latitude (deg)

90

0
-30

30 60 90 120 150 180

15 July 2004, 18UT

90

30

Longitude (deg)

15 July 2004, 18UT

Latitude (deg)

30 60 90 120 150 180

15 Jul 2004, 06UT

90

Latitude (deg)

Latitude (deg)

15 Jul 2004, 06UT

90

30

Longitude (deg)

-60

30
0
-30
-60

-90
-180 -150 -120 -90 -60 -30

-90
-180 -150 -120 -90 -60 -30

30 60 90 120 150 180

Longitude (deg)

30 60 90 120 150 180

Longitude (deg)

Figure 7. (a) Global maps in geographical coordinate system calculated from the empirical background TEC
model (left column of plots) which are compared with the CODE TEC maps (right column of plots) at 00 UT (upper
row of plots) and 06 UT (bottom row of plots) for 15 July 2004 during middle solar activity (2004). The modip latitude is also marked by white line. (b) The same as in Figure 7a, but at 12 UT (upper row of plots) and at 18 UT
(bottom row of plots).
4604

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

hemispheric asymmetry of the EIA, generated mainly by the

thermospheric transequatorial neutral winds blowing from
the summer to winter hemisphere, is evident well at both
model and CODE maps but only at low solar activity
(Figure 5b, upper row of plots). The hemispheric symmetry
of the EIA during vernal equinox is also well evident at both
model and CODE maps during high and low solar activity
(Figures 5a and 5b, bottom rows of plots).
[29] Figure 6a presents the global maps from the background TEC model (left column of plots) compared with
the CODE TEC maps (right column of plots) at 12UT for
15 July 2001 (upper row of plots) and 15 October 2001
(bottom row of plots) during high solar activity 2001, while
Figure 6b shows the same but for low solar activity 2008.
Again, the degree of similarity between model and CODE
TEC maps is very high. Some underestimation of the model
TEC is seen in July at most northern latitudes at both high
and low solar activity (Figures 6a and 6b, upper row of plots).
The model, however, reproduces very well the four TEC
amplications seen between ~50N and ~40S at the most
western longitudes and two TEC amplications at most
eastern longitudes in July 2008 (Figure 6b, upper row of
plots). It is worth noting that both model and CODE maps
show the following features: (i) do not display hemispheric
asymmetry of the EIA during July at both solar activity
conditions, (ii) the January TEC is larger than that in July
at high and low solar activity (so-called winter anomaly),
and (iii) while the March TEC is higher than the October
one during low solar activity, the opposite feature is evident
during high solar activity. The last result does not support
that of Natali and Meza [2011] which however is found for
2000 (not 2001 as here). We remind also that all the above
mentioned features are seen at 12UT.
[30] The comparisons for the middle solar activity 2004 are
presented in Figures 7a and 7b; in this case, only maps for
July at different UT are shown in order to trace out better
the diurnal variability of the low latitude plasma bulk.
Figure 7a shows global maps from the background TEC
model (left column of plots) which are compared with the
CODE TEC maps (right column of plots) at 00UT (upper
row of plots) and 06UT (bottom row of plots) for 15 July
2004 while Figure 7b presents the same but at 12UT (upper
row of plots) and at 18UT (bottom row of plots). Again the
comparison shows high degree of similarity; even some
details in the spatial TEC distribution are well reproduced
(see, for example, Figure 7a). Some hemispheric asymmetry
of the whole low-latitude plasma bulk can be distinguished
only at 00UT and 18UT; the TEC values at positive modip
latitudes are larger than those at the respective negative ones.
[31] The above shown comparisons indicate that the
empirical background TEC model can reproduce very well
the main spatial-temporal variability of the CODE TEC
maps. Each empirical model needs to assess the quality of
the adjustment procedure by model residuals, i.e., by calculating the differences between input data and model values
[Jakowski et al., 2011]. A detailed description of the model
residuals will be done in the companion paper. Here however
only the main statistics based on the entire data set will be
presented. It is worth noting that actually the residuals reveal
the nature of the modelling error. Due to this, it has been
accepted that the mean (systematic) error (ME), root mean
squares error (RMSE), and the standard deviation error

(STDE) are the main characteristics of each model. They

are dened as:
1 N
TEC mod  TEC obs
N i1
s
1 N
TEC mod  TEC obs 2
RMSE
N i1
p
STDE RMSE2  M E 2

ME

[32] The application of (4) to all data for the considered

period of time (1 January 199931 December 2011) gives
the following errors: ME = 0.003 TECU, i.e., the model
has practically zero systematic error. In this case,
RMSE = STDE = 3.387 TECU. In order to have an idea if such
errors are high or low, we compare this model with the similar
to some extend new global TEC model built recently by
Jakowski et al. [2011] and called NTCM-GL model. The statistical assessments of the NTCM-GL model are: ME = 0.3
TECU and RMSE = 7.5 TECU. Hence, the errors of the present background TEC model are signicantly smaller than
those of the NTCM-GL model. We have to note nevertheless
that both models are climatological, i.e., they describe the average behavior under quiet geomagnetic conditions, the TEC
model constructed by Jakowski et al. [2011] needs only 12 coefcients and a few empirically xed parameters for describing a broad spectrum of TEC variation at all levels of solar
activity. We, however, do not consider the large number of coefcients in the present background TEC model, 4374, as its
weak point. They are calculated only once and are xed later
at the model applications.

5.

Summary

[33] In this study, we present a global background TEC

model built on the basis of full 13 years (19992011) of
CODE TEC data. The model describes the climatological,
i.e., under quiet geomagnetic conditions, behavior of the
ionosphere, and can be used for long-term prediction. For
this purpose, at given DOY, geographic location, and UT,
the model needs as input parameters only the predicted level
of solar activity (F10.7 is used here as a proxy for solar
activity). The model maps are arrayed in terms of the coordinate system of geographical latitude from 90 to 90 at each
5 and longitude from 180 to 180 at each 5.
[34] The approach for building this model is based on the
very different time scales of the solar cycle, seasonal, and diurnal TEC variabilities (they differ at least an order of
magnitude). This leads to modulations of shorter-period variabilities with periods of the longer ones. Then the TEC spatial-temporal variability has been presented as a multiplication
of three separable functions (as it is shown in (2)). In
constructing the present background model, besides the seasonal course that is described in a standard way, i.e., by the
subharmonics of the year, the two new ideas are used in empirical modeling of the TEC solar cycle and diurnal variabilities:
[35] 1. solar cycle variability is presented not only by
F10.7 but also by its linear rate of change, KF, i.e., its tendency; in this way, the model describes better the different
ionospheric behavior at rising and declining parts of the solar
cycle when F10.7 has the same values;

4605

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

[36] 2. diurnal TEC variability is described not only by migrating tidal components, as it has been done so far, but
nonmigrating tides and SPW structures are also included; in
this way, the contribution of both tidal forcing from below
and the effects arising from the offset between geographic
and modip latitudes are included in the model.
[37] It was shown that the model describes very well such
structures as the WSA (Figure 3) and the well-known WN4
and WN3 longitude structures (Figure 4). This was possible
mainly because of the nonmigrating tides and SPW inclusion
in describing the diurnal variability of the TEC. The
presented comparison between the model and CODE TEC
maps at different solar activities and seasons (Figures 5, 6,
and 7) also demonstrated high degree of similarity. The zero
systematic error and its low RMSE (3.387 TECU) provides
the model signicant advantage over the other similar
models. The detailed statistical evaluation of the present
model will be done in the companion paper. There, an error
model will be presented as well.
[38] The present background model can be used for both
science and applications. In science, the model can be utilized as a background condition on the basis of which the perturbations can be estimated. It is particularly useful for
investigating the geomagnetic perturbations, or ionospheric
disturbances related to the sudden stratospheric warmings,
by incoherent scatter radars where the measurements are
available only for several days; in this case, the background
condition described by the monthly median TEC values cannot
be determined. This model can be useful for numerous singlefrequency GPS applications which need additional information
for mitigating the ionospheric propagation error. Such GNSS
users can be provided by adequate ionospheric corrections
obtained by this autonomous ionospheric TEC model.
[39] Acknowledgments. We are grateful to the CODE TEC team for
the access to the TEC data provided from the CODE FTP directory: ftp://
ftp.unibe.ch/aiub/CODE/. This work was supported by the European
Ofce of Aerospace Research and Development (EOARD), Air Force
Ofce of Scientic Research, Air Force Material Command, USAF, under
grant FA8655-12-1-2057 to D. Pancheva. We thank the anonymous
reviewers for their insightful comments on the original manuscript.
[40] Robert Lysak thanks the reviewers for their assistance in evaluating
this paper.

References
Anderson, D. N., J. M. Forbes, and M. Codrescu (1989), A fully analytical,
low- and middle-latitude ionospheric model, J. Geophys. Res., 94(A2),
15201524, doi:10.1029/JA094iA02p01520.
Apostolov, E., L. Alberca, and D. Pancheva (1994), Long-term prediction of
the foF2 on the rising and falling parts of the solar cycle, Adv. Space Res.,
14(12), 4750.
Azpilicueta, F., C. Brunini, and S. M. Radicella (2006), Global ionospheric
maps from GPS observations using modip latitude, Adv. Space Res., 38,
23242331.
Bilitza, D. (2001), International Reference Ionosphere 2000, Radio Sci., 36,
261275, doi:10.1029/2000RS002432.
Bilitza, D., and B. W. Reinisch (2008), International Reference Ionosphere
2007: Improvements and new parameters, Adv. Space Res., 42, 599609,
doi:10.1016/j.asr.2007.07.048.
Bouya, Z., M. Terkildsen, and D. Neudegg (2010), Regional GPS-based ionospheric TEC model over Australia using Spherical Cap Harmonic
Analysis, paper presented at 38th COSPAR Scientic Assembly, Comm.
on Space Res., Bremen, Germany.
Brunner, F. K., and M. Gu (1991), An improved model for the dual frequency ionospheric correction of GPS observations, Manuscr. Geodaet.,
16, 205214.
Chen, C. H., J. D. Huba, A. Saito, C. H. Lin, and J. Y. Liu (2011),
Theoretical study of the ionospheric Weddell Sea Anomaly using
SAMI2, J. Geophys. Res., 116, A04305, doi:10.1029/2010JA015573.

Daniell, R. E., W. G. Whartenby, and L. D. Brown (1993a), Parameterized

Real-time Ionospheric Specication Model, PRISM version 1.2,
Algorithm description, Comput. Phys., Inc., Newton, Mass.
Daniell, R. E., W. G. Whartenby, and L. D. Brown (1993b), Parameterized
Real-time Ionospheric Specication Model, PRISM version 1.2,
Validation report, Comput. Phys., Inc., Newton, Mass.
Dow, J. M., R. E. Neilan, and C. Rizos (2009), The International GNSS
Service in a changing landscape of Global Navigation Satellite Systems,
J. Geod., 83, 191198, doi:10.1007/s00190-008-0300-3.
Drob, D. P., et al. (2008), An empirical model of the Earths horizontal
wind elds: HWM07, J. Geophys. Res., 113, A12304, doi:10.1029/
2008JA013668.
Elmas, Z. G., M. Aquino, H. A. Marques, and J. F. G. Monico (2011), Higher
order ionospheric effects in GNSS positioning in the European region,
Ann. Geophys., 29, 13831399, doi:10.5194/angeo-29-1383-2011.
Ercha, A., D. Zhang, A. J. Ridley, Z. Xiao, and Y. Hao (2012), A global model:
Empirical orthogonal function analysis of total electron content 19992009
data, J. Geophys. Res., 117, A03328, doi:10.1029/2011JA017238.
Feltens, J. (2007), Development of a new three-dimensional mathematical
ionosphere model at European Space Agency/European Space Operations
Centre, Space Weather, 5, S12002, doi:10.1029/2006SW000294.
Feltens, J., and S. Schaer (1998), IGS products for the ionosphere, IGS
Position Paper, in IGS 1998 Analysis Center Workshop: Proceedings,
edited by J. M. Dow, J. Kouba, and T. Springer, pp. 225232, Eur.
Space Oper. Cent., Darmstadt, Germany.
Feltens, J., M. Angling, N. Jackson-Booth, N. Jakowski, M. Hoque,
M. Hernndez-Pajares, A. Aragn-ngel, R. Ors, and R. Zandbergen
(2011), Comparative testing of four ionospheric models driven with GPS
measurements, Radio Sci., 46, RS0D12, doi:10.1029/2010RS004584.
Fuller-Rowell, T. J., D. Rees, S. Quegan, R. J. Moffett, and G. J. Bailey
(1987), Interaction between neutral thermosphere composition and the
polar ionosphere using a coupled ionosphere-thermosphere model,
J. Geophys. Res., 92(A7), 77447748, doi:10.1029/JA092iA07p07744.
Fuller-Rowell, T., E. Araujo-Pradere, C. Minter, M. Codrescu, P. Spencer,
D. Robertson, and A. R. Jacobson (2006), US-TEC: A new data assimilation product from the Space Environment Center characterizing the ionospheric total electron content using real-time GPS data, Radio Sci., 41,
RS6003, doi:10.1029/2005RS003393.
Gao, Y., and Z. Z. Liu (2002), Precise ionosphere modeling using regional
GPS network data, J. Global Pos. Systems, 1, 1824.
Gao, Y., P. Heroux, and J. Kouba (1994), Estimation of GPS receiver and
satellite L1/L2 signal delay biases using data from CACS, paper presented
at the International Symposium on Kinematic Systems in Geodesy,
Geomatics, and Navigation, Univ. of Calgary, Banff, Alberta, Canada.
Ge, S., et al. (2004), Comparison of TEC measurements from dual frequency
space geodetic techniques, Eos. Trans. AGU, 85(47), Fall Meet. Suppl.,
Abstract G53A-0118.
Gopal Rao, M. S. V., and R. Sambasiva Rao (1969), The hysteresis variation
in the F2-layer parameters, J. Atmos. Terr. Phys., 31, 11191125.
Habarulema, J. B., L.-A. McKinnell, and B. D. L. Opperman (2010), TEC
measurements and modelling over Southern Africa during magnetic
storms; a comparative analysis, J. Atmos. Sol. Terr. Phys., 72, 509520,
doi:10.1016/j.jastp.2010.01.012.
Habarulema, J. B., L.-A. McKinnell, and B. D. L. Opperman (2011),
Regional GPS TEC modeling; Attempted spatial and temporal extrapolation of TEC using neural networks, J. Geophys. Res., 116, A04314,
doi:10.1029/2010JA016269.
Hartman, W. A., and R. A. Heelis (2007), Longitudinal variations in the
equatorial vertical drift in the topside ionosphere, J. Geophys. Res., 112,
A03305, doi:10.1029/2006JA011773.
He, M., L. Liu, W. Wan, B. Ning, B. Zhao, J. Wen, X. Yue, and H. Le
(2009), A study of the Weddell Sea Anomaly observed by
FORMOSAT-3/COSMIC, J. Geophys. Res., 114, A12309, doi:10.1029/
2009JA014175.
Hedin, A. E., et al. (1996), Empirical wind model for the upper, middle and
lower atmosphere, J. Atmos. Terr. Phys., 58, 14211447.
Hernndez-Pajares, M., J. M. Juan, and J. Sanz (1997), High-resolution TEC
monitoring method using permanent ground GPS receivers, Geophys. Res.
Lett., 24, 16431646, doi:10.1029/97GL01591.
Hernndez-Pajares, M., J. M. Juan, and J. Sanz (1999), New approaches in
global ionospheric determination using ground GPS data, J. Atmos. Sol.
Terr. Phys., 61, 12371247, doi:10.1016/S1364-6826(99)00054-1.
Hernndez-Pajares, M., J. M. Juan, J. Sanz, R. Orus, A. Garcia-Rigo,
J. Feltens, A. Komjathy, S. C. Schaer, and A. Krankowski (2009), The
IGS VTEC maps: A reliable source of ionospheric information since
1998, J. Geod., 83, 263275, doi:10.1007/s00190-008-0266-1.
Ho, C. M., A. J. Mannucci, U. J. Lindqwister, X. Pi, and B. T. Tsurutani
(1996), Global ionosphere perturbations monitored by the worldwide
GPS network, Geophys. Res. Lett., 23, 32193222, doi:10.1029/
1996GL02763.

4606

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

Hochegger, G., B. Nava, S. M. Radicella, and R. Leitinger (2000), A family
of ionospheric models for different uses, Phys. Chem. Earth C Sol. Terr.
Planet. Sci., 25(4), 307310.
Hoque, M. M., and N. Jakowski (2007), Higher order ionospheric effects
in precise GNSS positioning, J. Geod., 81(4), 259268, doi:10.1007/
s00190-006-0106-0.
Huang, J.-N. (1963), The hysteresis variation of the semi-thickness of the
F2-layer and its relevant phenomena at Kokubunji, Japan, J. Atmos.
Terr. Phys., 25, 647658.
Huba, J. D., G. Joyce, S. Sazykin, R. Wolf, and R. Spiro (2005), Simulation
study of penetration electric eld effects on the low- to mid-latitude ionosphere, Geophys. Res. Lett., 32, L23101, doi:10.1029/2005GL024162.
Iijima, B. A., I. L. Harris, C. M. Ho, U. J. Lindqwister, A. J. Mannucci, X. Pi,
M. J. Reyes, L. C. Sparks, and B. D. Wilson (1999), Automated daily
process for global ionospheric total electron content maps and satellite
ocean altimeter ionospheric calibration based on Global Positioning
System data, J. Atmos. Sol. Terr. Phys., 61, 12051218, doi:10.1016/
S1364-6826(99)00067-X.
Immel, T. J., E. Sagawa, S. L. England, S. B. Henderson, M. E. Hagan,
S. B. Mende, H. U. Frey, C. M. Swenson, and L. J. Paxton (2006),
Control of equatorial ionospheric morphology by atmospheric tides,
Geophys. Res. Lett., 33, L15108, doi:10.1029/2006GL026161.
Jakowski, N., M. M. Hoque, and C. Mayer (2011), A new global TEC model
estimating trans-ionospheric radio wave propagation errors, J. Geod., 85,
965974, doi:10.1007/s00190-011-0455-1.
Karpachev, A. T., N. A. Gasilov, and O. A. Karpachev (2011), Morphology
and causes of the Weddel Sea Anomaly, Geomag. Aeron., 51(6), 812824,
doi:10.1134/S0016793211050070.
Kashcheyev, A., B. Nava, and S. M. Radicella (2012), Estimation of
higher-order ionospheric errors in GNSS positioning using a realistic
3-D electron density model, Radio Sci., 47, RS4008, doi:10.1029/
2011RS004976.
Kil, H., E. R. Talaat, S.-J. Oh, L. J. Paxton, S. L. England, and S.-Y. Su
(2008), The wave structures of the plasma density and vertical E
B drift
in low-latitude F region, J. Geophys. Res., 113, A09312, doi:10.1029/
2008JA013106.
Komjathy, A. (1997), Global ionospheric total electron content mapping
using the Global Positioning System, Ph.D. thesis, Univ. of New
Brunswick, Frederiction, New Brunswick, Canada.
Lean, J. L., R. R. Meier, J. M. Picone, and J. T. Emmert (2011a), Ionospheric
total electron content: Global and hemispheric climatology, J. Geophys.
Res., 116, A10318, doi:10.1029/2011JA016567.
Lean, J. L., J. T. Emmert, J. M. Picone, and R. R. Meier (2011b), Global and
regional trends in ionospheric total electron content, J. Geophys. Res., 116,
A00H04, doi:10.1029/2010JA016378.
Lean, J. L., T. N. Woods, F. G. Eparvier, R. R. Meier, D. J. Strickland,
J. T. Correira, and J. S. Evans (2011c), Solar extreme ultraviolet irradiance: Present, past, and future, J. Geophys. Res., 116, A01102,
doi:10.1029/2010JA015901.
Mannucci, A. J., B. D. Wilson, D. N. Yuan, C. H. Ho, U. J. Lindqwister, and
T. F. Runge (1998), A global mapping technique for GPS-derived ionospheric total electron content measurements, Radio Sci., 33, 565582,
doi:10.1029/97RS02707.
Mao, T., W. Wan, X. Yue, L. Sun, B. Zhao, and J. Guo (2008), An empirical
orthogonal function model of total electron content over China, Radio Sci.,
43, RS2009, doi:10.1029/2007RS003629.
Meggs, R. W., C. N. Mitchell, and P. S. J. Spencer (2004), A comparison of
techniques for mapping total electron content over Europe using GPS
signals, Radio Sci., 39, RS1S10, doi:10.1029/2002RS002846.
Mendillo, M., H. Rishbeth, R. G. Roble, and J. Wroten (2002), Modelling
F2-layer seasonal trends and day-to-day variability driven by coupling
with the lower atmosphere, J. Atmos. Sol. Terr. Phys., 64, 19111931,
doi:10.1016/S1364-6826(02)00193-1.
Mukhtarov, P. and D. Pancheva (2011), Global ionospheric response to
nonmigrating DE3 and DE2 tides forced from below, J. Geophys. Res.,
116, A05323, doi:10.1029/2010JA016099.
Mukhtarov, P., D. Pancheva, and B. Andonov (2010), Climatology of the
stationary planetary waves seen in the SABER/TIMED temperatures
(20022007), J. Geophys. Res., 115, A06315, doi:10.1029/2009JA015156.
Natali, M. P., and A. Meza (2011), Annual and semiannual variations of vertical total electron content during high solar activity based on GPS observations, Ann. Geophys., 29, 865873.
Nava, B., P. Coisson, and S. M. Radicella (2008), A new version of the
NeQuick ionosphere electron density model, J. Atmos. Sol. Terr. Phys.,
70, 18561862.
Oberheide, J., and J. M. Forbes (2008), Tidal propagation of deep tropical
cloud signatures into the thermosphere from TIMED observations.
Geophys. Res. Lett. 35, L04816, doi:10.1029/ 2007GL032397.
Oberheide, J., J. M. Forbes, X. Zhang, and S. L. Bruinsma (2011), Wavedriven variability in the ionosphere-thermosphere-mesosphere system

from TIMED observations: What contributes to the wave4?,

J. Geophys. Res., 116, A01306, doi:10.1029/2009JA015911.
Ors, R., M. Hernndez-Pajares, J. M. Juan, and J. Sanz (2005),
Improvement of global ionospheric VTEC maps by using Kriging interpolation technique, J. Atmos. Sol. Terr. Phys., 67, 15981609, doi:10.1016/j.
jastp.2005.07.017.
Otsuka, Y., et al. (2002), A new technique for mapping of total electron content using GPS network in Japan, Earth Planet Space, 54, 6370.
Pancheva, D., and P. Mukhtarov (1996), A single-station spectral model of the
monthly median F-region critical frequency, Ann. Geosc., 39(4), 807818.
Pancheva, D., and P. Mukhtarov (1998), A single station spectral model of
the monthly median foF2 and M(3000)F2, Studia geophys. et geod.,
42(2), 183196, doi:10.1023/A:1023361105552.
Pancheva D., and P. Mukhtarov (2010), Strong evidence for the tidal control
on the longitudinal structure of the ionospheric F-region, Geophys. Res.
Lett., 37, L14105, doi:10.1029/ 2010GL044039.
Pancheva, D., and P. Mukhtarov (2012), Global response of the ionosphere to
atmospheric tides forced from below: Recent progress based on satellite
measurements, Space Sci. Rev., 168(14), 175209, doi:10.1007/s11214011-9837-1.
Pancheva, D., P. Mukhtarov, N. J. Mitchell, and H. G. Muller (2005),
Empirical Model of the Dynamics in the Mesosphere and Lower
Thermosphere Region over the UK, Including Solar and Geomagnetic
Activity, J. Atmos. Sol. Terr. Phys., 67, 197209.
Pancheva, D., P. Mukhtarov, and B. Andonov (2010), Climatology of the
eastward propagating tides seen in the SABER/TIMED temperatures
(20022007), Adv. Space Res., 46, 257274, doi:10.1016/j.asr.2010.03.026.
Pancheva, D., Y. Miyoshi, P. Mukhtarov, H. Jin, H. Shinagava, and
H. Fujiwara (2012), Global response of the ionosphere to atmospheric
tides forced from below: Comparison between COSMIC measurements
and simulations by Atmosphereionosphere Coupled Model GAIA,
J. Geophys. Res., 117, A07319, doi:10.1029/2011JA017452.
Ping, J., Y. Kono, K. Matsumoto, Y. Otsuka, A. Saito, C. Shum, K. Heki,
and N. Kawano (2002), Regional ionosphere map over Japanese Islands,
Earth Planet. Space, 54, e13e16.
Radicella, S. M., and R. Leitinger (2001), The evolution of the DGR
approach to model electron density proles, Adv. Space Res., 27, 3540.
Rawer, K. (1963), in Meteorological and Astronomical Inuences on
Radio Wave Propagation, edited by B. Landmark, pp. 221250, Pergamon
Press, Oxford.
Rawer, K. (1984), (Ed.), Encyclopedia of Physics, Geophysics III, Part VII,
pp. 389391, Springer-Verlag, Berlin.
Ren, Z., W. Wan, L. Liu, B. Zhao, Y. Wei, X. Yue, and R. A. Heelis (2008),
Longitudinal variations of electron temperature and total ion density in the
sunset equatorial topside ionosphere, Geophys. Res. Lett., 35, L05108,
doi:10.1029/2007GL032998.
Rishbeth, H. (2006), F region links with the lower atmosphere?, J. Atmos.
Sol. Terr. Phys., 68, 469478, doi:10.1016/j.jastp.2005.03.017.
Roble, R. (1995), Energetics of the mesosphere and thermosphere, in The
Upper Mesosphere and Lower Thermosphere: A Review of Experiment
and Theory, Geophys. Monogr. Ser., vol. 87, edited by R. M. Johnson and
T. L. Killeen, pp. 122, AGU, Washington, D. C., doi:10.1029/GM087.
Sayin, I., F. Arikan, and O. Arikan (2008), Regional TEC mapping with
Random Field Priors and Kriging, Radio Sci., 43, RS5012, doi:10.1029/
2007RS003786.
Schaer, S. (1999), Mapping and Predicting the Earths Ionosphere Using the
Global Positioning System, vol. 59, Geod. Geophys. Arb. Schweiz.Inst.
fr Geod. und Photogramm, Zurich, Switzerland.
Schunk, R. W., and A. F. Nagy (2009), IonospheresPhysics, Plasma
Physics, and Chemistry, 628 pp., Cambridge Univ. Press, New York,
doi:10.1017/CBO9780511635342.
Schunk, R. W., J. J. Sojka, and M. D. Bowline (1986), Theoretical study of
the electron temperature in the high latitude ionosphere for solar maximum
and winter conditions, J. Geophys. Res., 91(A11), 12,04112,054,
doi:10.1029/JA091iA11p12041.
Shepard, D. (1968), A two-dimensional interpolation function for
irregularly-spaced data, ACM 68 Proc., pp. 517524, ACM New York,
New York, doi:10.1145/800186.810616.
Stolle, C., S. Schlter, S. Heise, C. Jacobi, N. Jakowski, S. Friedel,
D. Krschner, and H. Lhr (2005), GPS ionospheric imaging of the north
polar ionosphere on 30 October 2003, Adv. Space Res., 36, 22012206,
doi:10.1016/j.asr.2005.08.047.
Wan, W., L. Liu, X. Pi, M.-L. Zhang, B. Ning, J. Xiong, and F. Ding
(2008), Wavenumber-4 patterns of the total electron content over the
low latitude ionosphere, Geophys. Res. Lett., 35, L12104, doi:10.1029/
2008GL033755.
Wan, W., F. Ding, M. Zhang, L. Liu, and B. Ning (2012), Modeling the
global ionospheric total electron content with empirical orthogonal
function analysis, Sci. China, Ser. E, 55, 11611168, doi:10.1007/
s11431-012-4823-8.

4607

MUKHTAROV ET AL.: 1. EMPIRICAL BACKGROUND TEC MODEL

Wilson, B. D., A. J. Mannucci, and C. D. Edwards (1995), Subdaily northern
hemisphere maps using an extensive network of GPS receivers, Radio Sci.,
30, 639.
Wu, Q., D. A. Ortland, S. C. Solomon, W. R. Skinner, and R. J. Niciejewski
(2011), Global distribution, seasonal, and inter-annual variations of

mesospheric semidiurnal tide observed by TIMED TIDI, J. Atmos. Sol.

Terr. Phys., 73(1718), 24822502, doi:10.1016/ j.jastp.2011.08.007.
Zapfe, B. D., M. Materassi, C. N. Mitchell, and P. Spalla (2006), Imaging of
the equatorial ionospheric anomaly over South AmericaA simulation
study of total electron content, J. Atmos. Sol. Terr. Phys., 68, 18191833.

4608