global modeling of equatorial spread f with sami3/waccm-x
TRANSCRIPT
Global Modeling of Equatorial Spread F
with SAMI3/WACCM-X
J.D.Huba1 and H.-L. Liu2
1 Syntek Technologies, Fairfax, VA
2 High Altitude Observatory, National Center for Atmospheric Research, Boulder,
CO
Abstract
We report the first results of a global ionosphere/thermosphere simulation study
that self-consistently generates large-scale equatorial spread F (ESF) plasma bubbles
in the post-sunset ionosphere. The coupled model comprises the ionospheric code
SAMI3 and the atmosphere/thermosphere code WACCM-X. Two cases are modeled
for different seasons and geophysical conditions: the March case (low solar activity:
F10.7 = 70) and the July case (high solar activity: F10.7 = 170). We find that
equatorial plasma bubbles formed and penetrated into the topside F layer for the
March case but not the July case. For the March case a series of bubbles formed
in the Atlantic sector with irregularity spacings in the range 400 - 1200 km, rose
to over 800 km, and persisted until after midnight. These results are consistent
with recent GOLD observations. Calculation of the generalized Rayleigh-Taylor
instability (GRTI) growth rate shows that the e-folding time was shorter for the
March case than the July case.
1
1 INTRODUCTION
Post-sunset ionospheric irregularities in the equatorial F region were first observed
by Booker and Wells (1938) using ionosondes. This phenomenon eventually became
known as equatorial spread F (ESF). It is now known that during ESF the equatorial
ionosphere becomes unstable because of a Rayleigh-Taylor-like instability: large scale
(10s km) electron density ‘bubbles’ can develop and rise to high altitudes (1000 km
or greater at times) [Haerendel, 1974; Ossakow, 1981; Hysell, 2000]. Modeling and
forecasting ESF is very important because of its impact on space weather: it causes
radio wave scintillation that degrades communication and navigation systems.
Given the complexity and non-linear development of equatorial plasma bubbles,
computational models are needed to tackle this problem. Scannapieco and Os-
sakow (1976) presented the first numerical results of ESF and demonstrated that
perturbations in the bottomside F layer can develop into bubbles that pene-
trate the topside. Subsequently, numerous and more sophisticated models have
been developed. An excellent review of these models is presented in Yokoyama (2017).
Virtually all of these models are regional: they cover a relatively narrow range of the
ionosphere in latitude and longitude. Typical longitudinal widths are 2 - 4 over
a latitude range ±15 − 30 with spatial resolutions in the range 1 km [Yokoyama
et al., 2015] to 5 km [Huba et al., 2008]. One exception is the work of Huba and
Joyce (2010) who embedded a high-resolution grid (longitude cell size 0.0625) in the
global model SAMI3. They seeded the pre- and post-sunset ionosphere with density
perturbations which developed into full-scale plasma bubbles. However, this model
used the empirical thermosphere models NRLMSISE00 [Picone et al., 2002] and
HWM07 [Drob et al., 2008] to specify the background neutral densities, temperature,
and winds.
In this Letter we report results from simulations using a coupled SAMI3/WACCM-X
2
model at high-resolution (the spatial grid is ≲ 70 km in the low-latitude ionosphere).
The codes are one-way coupled in that the thermospheric variables (i.e., neutral
densities, temperature, winds) from WACCM-X are inputs to SAMI3 but the SAMI3
ionosphere variables (i.e., ion density, temperature, velocity) do not feed back into
WACCM-X. Two seasons (equinox and solstice) are modeled with different levels of
solar activity: the March case (low solar activity: F10.7 = 70) and the July case
((high solar activity: F10.7 = 170). We find that the coupled model self-consistently
generates equatorial spread F plasma bubbles in the post-sunset ionosphere for the
March case. The growth rate of the generalized Rayleigh-Taylor instability (GRTI)
is calculated; we find that the e-folding time is much shorter for the March case than
the July case. Most significantly, for the March case, a series of bubbles formed in
the Atlantic sector with wavelengths in the range 400 - 1200 km, rose to over 800
km, and persisted until after midnight. These results are remarkably consistent with
recent GOLD observations.
2 MODELS
WACCM-X is an atmospheric component of the National Center for Atmospheric
Research (NCAR) Community Earth System Model, which couples atmosphere,
ocean, land surface, sea and land ice, and carbon cycle components through
exchanging fluxes and state information [Hurrell et al., 2013]. It is based on the
Community Atmosphere Model and WACCM. The first version of WACCMX is
described by H.L. Liu et al. (2010) and the most recent version is described in H.L.
Liu et al. (2018). The top boundary of WACCM-X v.2.0 is set at 4.0 × 10−10 hPa
(= 500 to 700 km altitude, depending on solar activity). The vertical resolution in
the mesosphere and thermosphere is a quarter of a scaleheight, and the horizontal
resolution is 0.474 × 0.625 in latitude and longitude, respectively. WACCM-X
has the option to have the tropospheric and stratospheric dynamics constrained to
meteorological reanalysis fields for specifically targeted time periods.
3
SAMI3 (Sami3 is Also a Model of the Ionosphere) is a seamless, global, three-
dimensional, physics-based model of the ionosphere. It is based on SAMI2 [Huba et
al., 2000]. SAMI3 models the plasma and chemical evolution of seven ion species
(H+, He+, N+, O+, N+2 , NO
+ and O+2 ). The temperature equation is solved for three
ion species (H+, He+ and O+) and for the electrons. Ion inertia is included in the
ion momentum equation for motion along the geomagnetic field. SAMI3 nominally
uses the EUVAC [Richards et al., 1994] model for solar radiation and the Richmond
apex model for the magnetic field [Richmond, 1995]. The neutral composition,
temperature, and winds can be specified in SAMI3 by analytical models, empirical
models, or physics-based models. The electrostatic potential used in SAMI3 in
the low- to mid-latitude ionosphere is determined by the solution of a potential
equation driven by the neutral wind dynamo [Huba et al., 2008]. The potential in the
high-latitude region is specified by the Weimer05 model [Weimer, 2005]. One new
feature of the SAMI3 model used in this study is the implementation of a 4th order
flux-corrected transport scheme for E × B transport perpendicular to the magnetic
field. The partial donor cell method [Hain, 1987; Huba, 2003] is used which reduces
numerical diffusion and allows steeper density gradients to develop.
Details of the simulation parameters are as follows. The WACCM-X grid is 0.474
× 0.625 in latitude and longitude, and covers the entire earth. The SAMI3 grid
is also 0.625 in longitude but is variable in latitude. The grid in latitude is ∼ 1
for latitudes 40, decreases to ∼ 0.15 near the magnetic equator, and increases to
∼ 1.5 in the high-latitude region. The SAMI3 grid extends to ±80 latitude in
magnetic coordinates.
Two simulation studies were performed for different seasons and solar activity. The
first simulation study is for March 21, 22 and 23 for low solar activity (F10.7 = 70
and F10.7A). The second study is for July 2, 3 and 4 for high solar activity (F10.7
= 170 and F10.7A = 170). We will refer to the first case as the ’March case’ and the
4
second case as the ’July case.’ Only results from the third day are presented. The
rationale for these choices is somewhat arbitrary but was made to highlight potential
differences under very different geophysical conditions with a limited set of runs.
3 RESULTS
In Fig. 1 we show the total electron content (TEC) calculated from SAMI3 at 23:59
UT for the July case (top panel) and the March case (bottom panel). Daytime
is over the Pacific sector with sunset in the American sector. The most striking
feature of the March case is the development of large depletions in TEC over eastern
South America, the Atlantic ocean, and western Africa. These depletions in TEC
are associated with large-scale plasma bubbles that developed after sunset and are
initiated by atmospheric gravity waves in the WACCM-X simulation data. The
spacing (i.e., wavelength) of the irregularities in longitude varies from 400 km to
1200 km at this time, which is consistent with recent observations of irregularity
wavelengths in the range 500 - 800 km [Aa et al., 2020]. The spacing also changes
as a function of longitude during the evolution of the irregularities. It is noted that
gravity waves with horizontal wavelength down to ∼ 500 km are properly resolved by
WACCM-X, according to analysis of the kinetic energy spectrum of model results.
Additionally, there is an enhancement of TEC at midlatitude in the post-midnight
sector; this is associated plasmasphere ‘drainage’ into the ionosphere because of the
low pressure during solar minimum conditions at night [Li et al., 2018]. By contrast,
no large-scale bubbles formed for the July case. There are, however, relatively weak
irregularities in the mid-latitude region over western Russia and the southern Indian
ocean.
To emphasize the role of gravity waves in generating the depletions seen in the
bottom panel of Fig. 1 we show results from a high-resolution SAMI3 simulation for
the March case that used the empirical thermospheric models NRLMSISE00 for the
neutral densities and temperature [Picone et al., 2002] and HWM14 for the winds
5
[Drob et al., 2015]. The TEC for this simulation is shown in Fig. 2 and is to be
compared to the bottom panel of Fig. 1. The gross morphology of the ionosphere
is roughly the same but there are clear differences. There are no plasma bubbles or
nighttime mid-latitude enhancements in Fig. 2 as in Fig. 1, and the daytime equa-
torial ionization crests are not as distinct. This suggests that in addition to gravity
waves being important in seeding the post-sunset structure that the difference in ther-
mospheric winds also plays a significant role in the electrodynamics of the ionosphere.
In Fig. 3 we compare 135.6 nm emissions from the simulation for the March case at
23:59 UT (left and middle panels) to GOLD emission observations (right panel) from
geosynchronous orbit [Eastes et al., 2019]. The GOLD results are for October, 2018
which corresponds to equinox conditions at solar minimum, similar to the conditions
of the simulation. For the 135.6 nm emission from OI, the total irradiance (in
Rayleighs) is given by [Melendez-Alvira et al., 1999; England et al., 2008].
IRR =1
4π106
∫α1356(Te)Ne[O
+]ds+β1356k1k24π106
∫Ne[O][O+]
k2[O+] + k3[O]ds. (1)
The coefficients used in Eq. (1) are given in England et al. (2008). The center
panel is on the same color scale as the GOLD data (maximum of 40 Rayleighs); it
shows that the intensity of the 135.6 nm emissions from the model is less than the
data. The left panel reduces the color scale maximum to 12 Rayleighs in order to
highlight the structure in the model results; it shows a remarkable similarity to the
data. The model results capture the extended ionization arcs from the post-sunset
period (eastern South America) to midnight (western Africa) observed in the data.
Moreover, regular plasma striations (bubbles) are also observed in the model as in
the data on similar scale lengths. We note that since the emission rate is roughly
proportional to n2e, the electron density from the simulation is ∼ 1.7 smaller than
needed to agree with the observations.
We show the relationship between F region plasma structure and the generalized
Rayleigh-Taylor instability (GRTI) growth rate (see Appendix). In Figs. 4 (July
6
case) and 5 (March case) we show the electron density as a function of longitude and
altitude (top panel) and the growth rate of the GRTI (bottom panel) at time 17:44
UT. These plots are in the plane of the magnetic equator. In Fig. 4 (July case) the
maximum growth rate is ∼ 7.9 × 10−4 s−1 at ∼ 30 longitude. This corresponds to
an e-folding time τ ∼ 21 min which is consistent with previous estimates [Sultan,
1996]. However, at this time no plasma bubbles have developed that penetrate the
topside F layer; there are undulations in the bottomside F layer which are commonly
observed. On the other hand, in Fig. 5 (March case) the maximum growth rate at
∼ 30 longitude is ∼ 1.0× 10−3 which corresponds to an e-folding time τ ∼ 15 min.
Plasma bubbles are beginning to develop at this time. Moreover, in the longitude
range 35 - 75 several large scale plasma bubbles have developed and penetrate
the topside F layer. The growth rates in these bubbles are ∼ 2.9 × 10−3 s−1 which
corresponds to an e-folding time ∼ 5.8 min.
In Figs. 6 (July case) and 7 (March case) we show the maximum growth rate of
the GRTI in the altitude range 200 - 300 km (left panel) and the electron density
along the magnetic equator at 312 km as a function of local time and longitude. The
GRTI growth rate is positive shortly after sunset (∼ 18:00 UT) for both cases. In
the July case, the growth rate is relatively uniform in longitude at ∼ 19:00 LT with
a maximum of ∼ 1.6× 10−3 s−1 which corresponds to an e-folding time ∼ 10.4 min.
However, no density irregularities are evident at 312 km at this time. On the other
hand, for the March case, the growth rate is non-uniform in longitude at ∼ 19:00 LT
with a maximum of ∼ 2.4 × 10−3 s−1 which corresponds to an e-folding time ∼ 7.0
min. Large scale density irregularities (plasma bubbles) are evident at 312 km that
correspond to regions of fast GRTI growth.
4 SUMMARY
We have presented the first high-resolution (latitude/longitude resolution ≲ 70
km), global simulation study of the ionosphere/thermosphere system that captures
7
the onset and nonlinear development of large-scale equatorial spread F plasma
bubbles. The coupled model comprises the ionospheric code SAMI3 and the
atmosphere/themosphere code WACCM-X. Thermospheric data from WACCM-X
(neutral densities, temperature, and winds) are used as inputs to SAMI3. Two cases
were modeled for different seasons and geophysical conditions: the March case (low
solar activity: F10.7 = 70) and the July case (high solar activity: F10.7 = 170). We
found that equatorial plasma bubbles formed and penetrated the topside F layer for
the March case but not the July case. The spacing between the irregularities is in the
range 400 - 1200 km; this is consistent with the scale of the resolved gravity waves.
However, a more detailed examination of the relationship between gravity waves
and irregularity scale lengths is needed. We emphasize that the bubbles formed
self-consistently (i.e., no artificial perturbations were introduced in the simulations)
which suggests that gravity waves were the dominant seed mechanism [Kelley et al.,
1981; Aa et al., 2020]. To support this hypothesis we performed a high resolution
SAMI3 simulation using empirical thermospheric models and no plasma bubbles
occurred. Additionally the calculation of the GRTI growth rate shows that the
e-folding time was shorter for the March case (growth time τ ∼ 7 min) than the July
case (growth time τ ∼ 12 min).
Although only two cases are considered, these results are consistent with observations
of strong longitiudinal and seasonal dependences for the observation of equatorial
spread F irregularities [Yizengaw and Groves, 2018]. Also, even though large scale
plasma bubbles did not develop for the July case, there were large-scale undulations
of the bottomside F layer which are observed [Woodman and LaHoz, 1976; Hysell,
2000]. Future work will focus on a wider range of geophyscial conditions and a clearer
identification of the underlying thermospheric and ionospheric conditions responsible
for the generation of equatorial spread F irregularities. For example, one area to
investigate in detail is the relationship between post-sunset rise of the F layer and
upwelling growth in controlling the onset of equatorial plasma bubbles as discussed
in Tsunoda et al. (2018).
8
5 APPENDIX
We have derived the flux-tube integrated growth rate, similar to Haerendel (1973)
and Sultan (1996), in dipole coordinates and have included ion inertia. The complex
frequency is given by (ω = ωr + iγ)
ω =2C
−B ± (B2 − 4AC)1/2(2)
where
A =
∫σHi0
1
Ωi
G(s) ds B = i
∫σP0G(s) ds C =
∫1
Ln
F (s)G(s) ds
and
F (s) = σP
( c
BE0ϕ − Vn0p
)+ σH
( c
BE0p + Vn0ϕ
)− σHi0
gpΩi
and
G(s) =1
bs
r0 sin3 θ
∆
and conductivities
σP =nec
B
[νin/Ωi
1 + ν2in/Ω
2i
+νen/Ωe
1 + ν2en/Ω
2e
]σH =
nec
B
[ν2in/Ω
2i
1 + ν2in/Ω
2i
− ν2en/Ω
2e
1 + ν2en/Ω
2e
]σPi =
nec
B
νin/Ωi
1 + ν2in/Ω
2i
σHi =nec
B
1
1 + ν2in/Ω
2i
and n is the density, e is the electric charge, c is the speed of light, B is the magnetic
field strength, ∆ = (1 + 3 cos2 θ)1/2, ναn is the neutral collision frequency, Ωα is the
ion cyclotron frequency, gp = −g p · r, bs = (r30/r3)∆, and
1
Ln
G(s) =1
r0
∆
sin3 θ
1
n0
∂n0
∂p
1
bs
r0 sin3 θ
∆=
1
bs
1
n0
∂n0
∂p
Acknowledgments: This research was supported by NASA (JDH:
NNH17ZDA001N; HLL: 80NSSC17K0007), NSF (JDH: AGS-1931415; HLL:
9
AGS-1552153) and AFOSR (HLL: FA9550-16-1-0050). National Center for
Atmospheric Research is a major facility sponsored by the National Science
Foundation under Cooperative Agreement No. 1852977. CESM/WACCM-
X source code is available at http://www.cesm.ucar.edu. Data available at
http://doi.org/10.5281/zenodo.3735040. We thank Dr. Astrid Maute for a critical
reading of the manuscript and Dr. Scott England for providing the code to calculate
the 135.6 nm emissions.
References
Aa, E., S. Zou, R. Eastes, D.K. Karan, S.-R. Zhang, P.J. Erickson, and
A.J Coster, Coordinated groundbased and spacebased observations of equato-
rial plasma bubbles, J. Geophys. Res.: Space Physics, 125, e2019JA027569,
https://doi.org/10.1029/2019JA027569, 2020.
Booker, H.G. and H.G. Wells, Scattering of radio waves by the F -region of the iono-
sphere, Terr. Mag. Atmos. Elec. 43, 249, 1938.
Drob, D. P., et al., An empirical model of the Earth’s horizontal wind fields: HWM07,
J. Geophys. Res., 113, A12304, doi:10.1029/2008JA013668, 2008
Drob, D. P., J. T. Emmert, J. W. Meriwether, J. J. Makela, E. Doornbos, M. Conde,
G. Hernandez, J. Noto, K. A. Zawdie, S. E. McDonald, J. D. Huba, and J. H.
Klenzing, An update to the Horizontal Wind Model (HWM): The quiet time ther-
mosphere, Earth and Space Science, doi:10.1002/2014EA000089, 2015.
Eastes, R. W., S.C. Solomon, R.E. Daniell, D.N. Anderson, A.G. Burns, S.L. England,
et al., Globalscale observations of the equatorial ionization anomaly, Geophysical
Research Letters, 46, 9318, https://doi.org/10.1029/2019GL084199, 2019.
England, S., T. Immel, and J.D. Huba, Modeling the longitudinal variation in the
post-sunset far-ultraviolet O+ airglow using the SAMI2 model, J. Geophys. Res.
113, A01309, doi:10.1029/2007JA012536, 2008.
Haerendel, G., Theory of equatorial spread F , preprint, Max-Planck Inst. fur Ex-
traterr. Phys., Munich, Germany, 1974.
10
Hain, K., The partial donor cell method, J. Comput. Phys. 73, 131, 1987.
Huba, J.D., G. Joyce, and J.A. Fedder, SAMI2 (Sami2 is Another Model of the
Ionosphere): A New Low-Latitude Ionosphere Model J. Geophys. Res., 105, 23,035,
2000.
Huba, J.D., A Tutorial on Hall Magnetohydrodynamics, in Space Simulations, eds.
M. Scholer, C.T. Dum, and J. Buchner (Springer, New York) p. 170, 2003.
Huba, J.D., G. Joyce, and J. Krall, Three-dimensional equatorial spread F modeling,
Geophys. Res. Lett. 35, L10102, doi:10.1029/2008GL033509, 2008.
Huba, J.D. and G. Joyce, Global modeling of equatorial plasma bubbles, Geophys.
Res. Lett. 37, L17104, doi:10.1029/2010GL044281, 2010.
Hurrell, J. W., Holland, M. M., Gent, P. R., Ghan, S., Kay, J. E., Kushner,
P. J., et al, The community Earth system model: A framework for collab-
orative research. Bulletin of the American Meteorological Society, 94, 1339,
https://doi.org/10.1175/BAMSD1200121.1, 2013.
Hysell, D.L., An overview and synthesis of plasma irregularities in equatorial spread
F , J. Atmos. Sol. Terr. Phys. 62, 1037, 2000.
Kelley, M. C., M.F. Larsen, C. LaHoz, and J.P¿ McClure, Gravity wave ini-
tiation of equatorial spread F: A case study, J. Geophys. Res., 86, 9087,
doi:10.1029/JA086iA11p09087, 1981.
Li, Q., Y. Hao, D. Zhang, and Z. Xiao, Nighttime enhancements in the midlatitude
ionosphere and their relation to the plasmasphere, J. Geophys. Research: Space
Physics, 123, 7686, https://doi.org/10.1029/2018JA025422, 2018.
Liu, H.L., Foster, B. T., Hagan, M. E., McInerney, J. M., Maute, A.,
Qian, L., et al., Thermosphere extension of the Whole Atmosphere Com-
munity Climate Model, Journal of Geophysical Research, 115, A12302.
https://doi.org/10.1029/2010JA015586, 2010.
Liu, H.L., Bardeen, C. G., Foster, B. T., Lauritzen, P., Liu, J., Lu, G., Wang, W.,
Development and validation of the Whole Atmosphere Community Climate Model
with thermosphere and ionosphere extension (WACCMX 2.0), Journal of Advances
in Modeling Earth Systems, 10, 381, https://doi.org/10.1002/2017MS001233, 2018.
11
Melendez-Alvira, D.J., R.R. Meier, J.M. Picone, P.D. Feldman, and B.M. McLaugh-
lin, Analysis of the oxygen nightglow measured by the Hopkins Ultraviolet Tele-
scope: Implications for ionospheric partial radiative recombination rate coefficients,
J. Geophys. Res., 104, 14,901, 1999.
Ossakow, S.L., Spread F theories: A review, J. Atmos. Terr. Phys., 43, 437, 1981.
Picone, J.M., A.E. Hedin, D.P. Drob, and A.C. Aikin, NRLMSISE-00 empirical model
of the atmosphere: Statistical comparisons and scientific issues, J. Geophys. Res.
107, doi: 10.1029/2002JA009430, 2002.
Richmond, A., Ionospheric electrodynamics using magnetic apex coordinates, J. Ge-
omag. Geoelec. 47, 191, 1995.
Scannapieco, A.J. and S.L. Ossakow, Nonlinear equatorial spread F , Geophys. Res.
Lett. 3, 451, 1976.
Sultan, P.J., Linear theory and modeling of the Rayleigh-Taylor instability leading
to the occurrence of equatorial spread F , J. Geophys. Res. 101, 26,875, 1996.
Tsunoda, R., S. Saito and T.T. Nguyen, Post-sunset rise of equatorial F
layeror upwelling growth?, Progress in Earth and Planetary Science, 5:22,
https://doi.org/10.1186/s40645-018-0179-4, 2018.
Weimer, D.R., Predicting surface geomagnetic variations using ionospheric electrody-
namic models, J. Geophys. Res. 110, A12307, doi:10.1029/2005JA011270, 2005.
Woodman, R. F., and C. La Hoz, Radar observations of F region equatorial irregu-
larities, J. Geophys. Res., 81, 5447, doi:10.1029/JA081i031p05447, 1976
Yizengaw, E. and K.M. Groves, Longitudinal and seasonal variability of equa-
torial ionospheric irregularities and electrodynamics, Space Weather, 16, 946,
https://doi.org/10.1029/2018SW001980, 2018.
Yokoyama, T., H. Shinagawa, and H. Jin, Nonlinear growth, bifurcation and pinching
of equatorial plasma bubble simulated by threedimensional highresolution bubble
model, J. Geophys. Res. Space Physics, 119, 10,474, doi:10.1002/2014JA020708,
2015.
Yokoyama, T., A review of the numerical simulation of equatorial plasma bubbles
12
toward scintillation evaluation and forecasting, Progress in Earth and Planetary
Science, 4:37, doi:10.1186/s40645-017-0153-6, 2017.
13
Figure 1: The TEC at 23:59 UT for the July case (F10.7 = 170) (top panel) and the
March case (F10.7 = 70) (bottom panel).
14
Figure 2: The TEC at 00:00 UT from a high-resolution SAMI3 simulation using
the empirical thermosphere models NRLMSISE00 and HWM14 for the March case
conditions.
Figure 3: Comparison of 135.6 nm emissions from the simulation for the March
case (left and middle panels) and GOLD emission data (right panel) observed from
geosynchronous orbit [Eastes et al., 2019].
15
Figure 4: The electron density as a function of longitude and altitude (top panel) and
the growth rate of the GRTI (bottom panel) along the magnetic equator at 17:44 UT
for the July case.
Figure 5: The electron density as a function of longitude and altitude (top panel) and
the growth rate of the GRTI (bottom panel) along the magnetic equator at 17:44 UT
for the March case.
16
Figure 6: The maximum growth rate of the GRTI in the altitude range 200 - 300
km (left panel) and the electron density along the magnetic equator at 312 km as a
function of local time and longitude (July case).
Figure 7: The maximum growth rate of the GRTI in the altitude range 200 - 300
km (left panel) and the electron density along the magnetic equator at 312 km as a
function of local time and longitude (March case).
17