Time constants in the ionosphere from neural network models

Neural network (NN) models for the low latitude and the polar ionosphere from the D- to the F-region were developed which are based on incoherent scatter radar data from Arecibo and EISCAT Svalbard, respectively. The various geophysical input parameters defining the NN are not only the ones that represent the time one wants to predict, but also the geophysical conditions prior to the time of the prediction. The optimum length of these preceding periods are derived for the two models are different, but a period of 60 days is a compromise acceptable for both latitudes. Furthermore from the Arecibo data time constants of electron density decay after sundown are derived which – arguably – are also relevant elsewhere, including the polar latitudes. Whereas at all altitudes the electron densities decay exponentially after sundown, below 300 km there is an additional variation with solar zenith angle.     

The variability and predictability of the IRI B0, B1 parameters over Grahamstown,South Africa

The International Reference Ionosphere (IRI) parameters B₀ and B₁ provide a representation of the thickness and shape, respectively, of the F2 layer of the bottomside ionosphere. These parameters can be derived from electron density profiles that are determined from vertical incidence ionograms. This paper aims to illustrate the variability of these parameters for a single mid latitude station and demonstrate the ability of the Neural Network (NN) modeling technique for developing a predictive model for these parameters. Grahamstown, South Africa (33.3°S, 26.5°E) was chosen as the mid latitude station used in this study and the B₀ and B₁ parameters for an 11 year period were determined from electron density profiles recorded at that station with a University of Massachusetts Lowell Center for Atmospheric Research (UMLCAR) Digisonde. A preliminary single station NN model was then developed using the Grahamstown data from 1996 to 2005 as a training database, and input parameters known to affect the behaviour of the F2 layer, such as day number, hour, solar and magnetic indices. An analysis of the diurnal, seasonal and solar variations of these parameters was undertaken for the years 2000, 2005 and 2006 using hourly monthly median values. Comparisons between the values derived from measured data and those predicted using the two available IRI-2001 methods (IRI tables and Gulyaeva, T. Progress in ionospheric informatics based on electron density profile analysis of ionograms. Adv. Space Res. 7(6), 39–48, 1987.) and the newly developed NN model are also shown in this paper. The preliminary NN model showed that it is feasible to use the NN technique to develop a prediction tool for the IRI thickness and shape parameters and first results from this model reveal that for the mid latitude location used in this study the NN model provides a more accurate prediction than the current IRI model options.     

Response of low latitude D-region ionosphere to the total solar eclipse of 22 July 2009 deduced from ELF/VLF analysis

Response of the D-region of the ionosphere to the total solar eclipse of 22 July 2009 at low latitude, Varanasi (Geog. lat., 25.27° N; Geog. long., 82.98° E; Geomag. lat. = 14° 55’ N) was investigated using ELF/VLF radio signal. Tweeks, a naturally occurring VLF signal and radio signals from various VLF navigational transmitters are first time used simultaneously to study the effect of total solar eclipse (TSE). Tweeks occurrence is a nighttime phenomena but the obscuration of solar disc during TSE in early morning leads to tweek occurrence. The changes in D-region ionospheric VLF reflection heights (h) and electron density (n_e: 22.6–24.6 cm⁻³) during eclipse have been estimated from tweek analysis. The reflection height increased from ∼89 km from the first occurrence of tweek to about ∼93 km at the totality and then decreased to ∼88 km at the end of the eclipse, suggesting significant increase in tweek reflection height of about 5.5 km during the eclipse. The reflection heights at the time of totality during TSE are found to be less by 2–3 km as compared to the usual nighttime tweek reflection heights. This is due to partial nighttime condition created by TSE. A significant increase of 3 dB in the strength of the amplitude of VLF signal of 22.2 kHz transmitted from JJI-Japan is observed around the time of the total solar eclipse (TSE) as compared to a normal day. The modeled electron density height profile of the lower ionosphere depicts linear variation in the electron density with respect to solar radiation as observed by tweek analysis also. These low latitude ionospheric perturbations on the eclipse day are discussed and compared with other normal days.     

Response of the mid-latitude D-region ionosphere to the total solar eclipse of 22 July 2009 studied using VLF signals in South Korean peninsula

In this paper, we analyze VLF signals received at Busan to study the the D-region changes linked with the solar eclipse event of 22 July 2009 for very short (∼390 km) transmitter–receiver great circle path (TRGCP) during local noon time 00:36–03:13 UT (09:36–12:13 KST). The eclipse crossed south of Busan with a maximum obscuration of ∼84%. Observations clearly show a reduction of ∼6.2 dB in the VLF signal strength at the time of maximum solar obscuration (84% at 01:53 UT) as compared to those observed on the control days. Estimated values of change in Wait ionospheric parameters: reflection height (h′) in km and inverse scale height parameter (β) in km⁻¹ from Long Wave Propagation Capability (LWPC) model during the maximum eclipse phase as compared to unperturbed ionosphere are 7 km and 0.055 km⁻¹, respectively. Moreover, the D-region electron density estimated from model computation shows 95% depletion in electron density at the height of ∼71 km. The reflection height is found to increase by ∼7 km in the D-region during the eclipse as compared to those on the control days, implying a depletion in the Lyman-α flux by a factor of ∼7. The present observations are discussed in the light of current understanding on the solar eclipse induced D-region dynamics.     

Studies of electron and ion temperatures at 500 km altitude during sunrise using Indian SROSS C2 satellite

Solar dependence of electron and ion temperatures (T_e and T_i) in the ionosphere is studied using RPA data onboard SROSS C2 at an altitude of ∼500 km and 77°E longitude during early morning hours (04:00–07:00 LT) for three solar activities: solar minimum, moderate and maximum during winter, summer and equinox months in 10°S–20°N geomagnetic latitude. In winter the morning overshoot phenomenon is observed around 06:00 LT (T_e enhances to ∼4000 K) during low-solar activity and to T_e ∼ 3800 K, during higher solar activity. In summer, it is observed around 05:30 LT, but the rate of T_e enhancement is higher during moderate solar activity (∼2700 K/hr) than the low-solar activity (∼1700 K/hr). During equinox, this phenomenon is delayed and is observed around 06:00 LT (∼4200 K) during all three activities.     

The electron formations under the radiation belts at L-shells 1.2–1.9

We continue to analyze the distribution of electron fluxes with energy 30–500 keV under the radiation belts at low and middle latitudes (L = 1.2–1.9) using experimental data obtained onboard ACTIVE satellite. Special attention is given to altitudinal distribution of electron fluxes and detailed analysis of these electron formations. We observe three main regions of electron flux registration that seem to exist constantly under the radiation belts. These regions are: magneto-conjugated to SAA region (in the north hemisphere), local zone of low intense electron flux accumulation to the west of SAA, and extensive region in the north hemisphere to the east. The analysis of experimental data obtained from ACTIVE satellite (orbit height 500–2500 km) shows that electron fluxes are registered in the wide altitude range up to 1100 km. It is shown that these formations have complicated initial structure with two regions of flux maximums: at L = 1.3 and L = 1.6–1.8. We compare particle data with low frequency (LF) data (ARIEL-4 satellite) and high frequency (HF) data (CORONAS-I satellite). Also we discuss the possible mechanisms of the appearance of these formations under the radiation belts.     

Longitudinal distribution of thermospheric densities and ionization rates in the upper atmosphere of Mars at mid-latitude using MGS ACC data

The Accelerometer Experiment (ACC) onboard Mars Global Surveyor (MGS) measured 1600 density profiles in the upper atmosphere of Mars during aerobraking. These measurements reveal large-scale and small-scale structure in the thermosphere of Mars. Here, the measurements of mass density for 115 orbits (#P0670–P0789) from November 1 to 30, 1998, under spring equinox and medium solar activity conditions (average F_10.7 ∼ 137) during phase 2 of the aerobraking in the thermosphere of Mars at different altitudes and longitudes are presented for northern mid-latitude (17–42°N) in the dayside atmosphere using ACC onboard MGS. From these mass densities, the neutral densities of different gases are derived from their mixing ratios. Using these neutral densities, the longitudinal distribution of photoionization rates and photoelectron impact ionization rates are calculated at wavelength range 1–102.57 nm due to EUV and soft X-ray radiation under photochemical controlled region using Analytical Yield Spectrum approach (AYS). These conditions are appropriate for MGS Phase 2 aerobraking period from which the accelerometer data is used. Under the photochemical equilibrium condition, the electron density near the peak varies as the square root of the total peak ionization rate. Using this fact, an attempt is being made to estimate the mean primary and secondary peak electron density by averaging the longitudinal variations of total peak ionization rates in the northern mid-latitude (17–42°N) ionosphere of Mars, as there is no radio science measurement at this latitude region by MGS.     

Deformation of the ionosphere structure during the space weather events on October–November 2003

We present the spatial maps of the ionosphere–plasmasphere slab thickness τ (ratio of the vertical total electron content, TEC, to the F-region peak electron density, NmF2) during the intense ionospheric storms of October–November 2003. The model-assisted technology for estimate of the upper boundary of the ionosphere, hup, from the slab thickness components in the bottomside and topside ionosphere – eliminating the plasmasphere contribution of τ – is applied at latitudes 35° to 70°N and longitudes −10° to 40°E, from the data of 20 observatories of GPS-TEC and ionosonde networks, for selected days and hours of October and November 2003. The daily–hourly values of NmF2, hmF2 and TECgps are used as the constrained parameters for the International Reference Ionosphere extended to the plasmasphere, IRI-Plas, during the ionospheric quiet days, positive and negative storm phases for estimate of τ and hup. Good correlation has been found between the slab thickness and the upper boundary of the ionosphere for the intense ionospheric storms at October–November 2003. During the negative phase of the ionospheric storm, when the ionospheric plasma density is exhausted, the nighttime upper boundary of the ionosphere is greatly uplifted towards the magnetosphere tail, while the daytime upper boundary of the ionosphere is reduced below 500 km over the Earth.     

Imaging Global Electron Content backwards in time more than 160 years ago

The Global Electron Content, GEC, represents the total number of electrons in the spherical layer over the Earth restricted by orbit of Global Positioning Satellite system (20,200 km). GEC is produced from Global Ionospheric Map of Total Electron Content, GIM-TEC, transformed to the electron density varying with height using the International Reference Ionosphere and Plasmasphere model, IRI-Plas. The climatologic GEC model is developed from GIM-TEC maps for a period 1999–2012 including the solar activity, annual and semiannual cycles as the most important factors affecting daily GEC variation. The proxy R_zp of the international sunspot numbers, R_i, is used as a measure of solar activity composed of 3 day smoothed R_i, 7 day and 81 day backwards mean of R_i scaled to the range of 1–40 proxy units, p.u. The root mean square error of the GEC climatologic model is found to vary from 8% to 13% of GEC. Taking advantage of a long history of sunspot numbers, the climatologic GEC model is applied for GEC reconstruction backwards in time for more than 160 years ago since 1850. The extended set of GEC values provides the numerical representation of the ionosphere and plasmasphere electron content coherent with variations of solar activity as a potential proxy index driving the ionosphere models.     

Theoretical simulation of electron density and temperature distribution at Indian equatorial and low latitude ionosphere

The electron density and temperature distribution of the equatorial and low latitude ionosphere in the Indian sector has been investigated by simultaneously solving the continuity, momentum and energy balance equations of ion and electron flux along geomagnetic field lines from the Northern to the Southern hemisphere. Model algorithm is presented and results are compared with the electron density and electron temperature measured in situ by Indian SROSS C2 satellite at an altitude of ∼500 km within 31°S–34°N and 75 ± 10°E that covers the Indian sector during a period of low solar activity. Equatorial Ionization Anomaly (EIA) observed in electron density, morning and afternoon enhancements, equatorial trough in electron temperature have been simulated by the model within reasonable limits of accuracy besides reproducing other normal diurnal features of density and temperature.     

Large-scale traveling ionospheric disturbances of auroral origin according to the data of the GPS network and ionosondes

The intensity of large-scale traveling ionospheric disturbances (LS TIDs), registered using measurements of total electron content (TEC) during the magnetic storms on October 29–31, 2003, and on November 7–11, 2004, had been compared with that of local electron density disturbances. The data of TEC measurements at ground-based GPS receivers located near the ionospheric stations and the corresponding values of the critical frequency of the ionospheric F region f_oF2 were used for this purpose. The variations of TEC and f_oF2 were similar for all events mentioned above. The previous assumption that the ionospheric region with vertical extension from 150 to 200 km located near the F-layer maximum mainly contributes to the TEC variations was confirmed for the cases when the electron density disturbance at the F region maximum was not more than 50%. However, this region probably becomes vertically more extended when the electron density disturbance in the ionospheric F region is about 85%.     

Coordinated in situ measurements of plasma irregularities and ground based scintillation observations at the crest of equatorial anomaly

First comparison of in situ density fluctuations measured by the DEMETER satellite with ground based GPS receiver measurements at the equatorial anomaly station Bhopal (geographic coordinates (23.2°N, 77.6°E); geomagnetic coordinates (14.29°N, 151.12°E)) for the low solar activity year 2005, are presented in this paper. Calculation of the diurnal maximum of the strength of the equatorial electrojet, which can serve as precursor to ionospheric scintillations in the anomaly region is also done. The Langmuir Probe experiment and Plasma Analyzer onboard DEMETER measure the electron and ion densities respectively. Irregularities in electron density distribution cause scintillations on transionospheric links and there exists a close relationship between an irregularity and scintillation. In 40% of the cases, DEMETER detects the irregularity structures (dNe/Ne ? 5% and dNi/Ni (O⁺) ? 5%) and GPS L band scintillations (S4 ? 0.2) are also observed around the same time, for the low solar activity period. It is found that maximum irregularity intensity is obtained in the geomagnetic latitude range of 10–20° for both electron density and ion density. As the GPS signals pass through this irregularity structure, scintillations are recorded by the GPS receiver installed at the equatorial anomaly station, Bhopal it is interesting to note that in situ density fluctuations observed on magnetic flux tubes that pass over Bhopal can be used as indicator of ionospheric scintillations at that site. Many cases of density fluctuations and associated scintillations have been observed during the descending low solar activity period. The percentage occurrence of density irregularities and scintillations shows good correspondence with diurnal maximum of the strength of electrojet, however this varies with different seasons with maximum correspondence in summer (up to 66%) followed by equinox (up to 50%) and winter (up to 46%). Also, there is a threshold value of EEJ strength to produce density irregularities ((dNe/Ne)_max ? 5%) and for moderate to strong scintillations (S4 ? 0.3) to occur. For winter this value is found to be ∼40 nT whereas for equinox and summer it is around 50 nT.     

The asymmetrical features in electron density during extreme solar minimum

The variations of plasma density in topside ionosphere during 23rd/24th solar cycle minimum attract more attentions in recently years. In this analysis, we use the data of electron density (Ne) from DEMETER (Detection of Electromagnetic Emissions Transmitted from Earthquake Regions) satellite at the altitude of 660–710 km to investigate the solstitial and equinoctial asymmetry under geomagnetic coordinate system at LT (local time) 1030 and 2230 during 2005–2010, especially in solar minimum years of 2008–2009. The results reveal that ΔNe (December–June) is always positive over Southern Hemisphere and negative over northern part whatever at LT 1030 or 2230, only at 0–10°N the winter anomaly occurs with ΔNe (December–June) > 0, and its amplitude becomes smaller with the declining of solar flux from 2005 to 2009. The ΔNe between September and March is completely negative during 2005–2008, but in 2009, it turns to be positive at latitudes of 20°S–40°N at LT 1030 and 10°S–20°N at LT 2230. Furthermore, the solstitial and equinoctial asymmetry index (AI) are calculated and studied respectively, which all depends on local time, latitude and longitude. The notable differences occur at higher latitudes in solar minimum year of 2009 with those in 2005–2008. The equinoctial AI at LT 2230 is quite consistent with the variational trend of solar flux with the lowest absolute AI occurring in 2009, the extreme solar minimum, but the solstitial AI exhibits abnormal enhancement during 2008 and 2009 with bigger AI than those in 2005–2007. Compared with the neutral compositions at 500 km altitude, it illustrates that [O/N₂] and [O] play some roles in daytime and nighttime asymmetry of Ne at topside ionosphere.     

Global characteristics of ionospheric electric fields and disturbances during the first hours of magnetic storms

The ionospheric plasma density can be significantly disturbed during magnetic storms. In the conventional scenario of ionospheric storms, the negative storm phases with plasma density decreases are caused by neutral composition changes, and the positive storm phases with plasma density increases are often related to atmospheric gravity waves. However, recent studies show that the global redistribution of the ionospheric plasma is dominated primarily by electric fields during the first hours of magnetic storms. In this paper, we present the measurements of ionospheric disturbances by the DMSP satellites and GPS network during the magnetic storm on 6 April 2000. The DMSP measurements include the F region ion velocity and density at the altitude of ∼840 km, and the GPS receiver network provides total electron content (TEC) measurements. The storm-time ionospheric disturbances show the following characteristics. The plasma density is deeply depleted in a latitudinal range of ∼20° over the equatorial region in the evening sector, and the depletions represent plasma bubbles. The ionospheric plasma density at middle latitudes (20°–40° magnetic latitudes) is significantly increased. The dayside TEC is increased simultaneously over a large latitudinal range. An enhanced TEC band forms in the afternoon sector, goes through the cusp region, and enters the polar cap. All the observed ionospheric disturbances occur within 1–5 h from the storm sudden commencement. The observations suggest that penetration electric fields play a major role in the rapid generation of equatorial plasma bubbles and the simultaneous increases of the dayside TEC within the first 2 h during the storm main phase. The ionospheric disturbances at later times may be caused by the combination of penetration electric fields and neutral wind dynamo process.     

Modeling and observations of the north–south ionospheric asymmetry at low latitudes at long deep solar minimum

Using the physics based model SUPIM and FORMOSAT-3/COSMIC electron density data measured at the long deep solar minimum (2008–2010) we investigate the longitude variations of the north–south asymmetry of the ionosphere at low latitudes (±30° magnetic). The data at around diurnal maximum (12:30–13:30 LT) for magnetically quiet (Ap ? 15) equinoctial conditions (March–April and September–October) are presented for three longitude sectors (a) 60°E–120°E, (b) 60°W–120°W and (c) 15°W–75°W. The sectors (a) and (b) have large displacements of the geomagnetic equator from geographic equator but in opposite hemispheres with small magnetic declination angles; and sector (c) has large declination angle with small displacement of the equators; vertical E × B drift velocities also have differences in the three longitude sectors. SUPIM investigates the importance of the displacement of the equators, magnetic declination angle, and E × B drift on the north–south asymmetry. The data and model qualitatively agree; and indicate that depending on longitudes both the displacement of the equators and declination angle are important in producing the north–south asymmetry though the displacement of the equators seems most effective. This seems to be because it is the displacement of the equators more than the declination angle that produces large north–south difference in the effective magnetic meridional neutral wind velocity, which is the main cause of the ionospheric asymmetry. For the strong control of the neutral wind, east–west electric field has only a small effect on the longitude variation of the ionospheric asymmetry. Though the study is for the long deep solar minimum the conclusions seem valid for all levels of solar activity since the displacement of the equators and declination angle are independent of solar activity.     

Jovian periodicities (∼10 h, ∼40 min) on Ulysses’ Distant Jupiter Encounter observations around the Halloween CIR events

We analyzed data from four different instruments (HI-SCALE, URAP, SWOOPS, VHM/FGM) onboard Ulysses spacecraft (s/c) and we searched for possible evidence of Jovian emissions when the s/c approached Jupiter during the times of Halloween events (closest time approach/position to Jupiter: February 5, 2004/R = 1683 R_J,θ = ∼49°). In particular, we analyzed extensively the low energy ion measurements obtained by the HI-SCALE experiment in order to examine whether low energy ion/electron emissions show a symmetry, and whether they are observed at north high latitudes upstream from the jovian bow shock, as is known to occur in the region upstream from the south bow shock as well ( Marhavilas et al., 2001). We studied the period from October 2003 to March 2004, as Ulysses moved at distances 0.8–1.2 AU from the planet at north Jovicentric latitudes <75°, and we present here an example of characteristic Jovian periodicities in the measurements around a CIR observed by Ulysses on days ∼348–349/2003 (R = 1894 R_J,θ = 72°). We show that Ulysses observed low energy ion (∼0.055–4.7 MeV) and electron (>∼40 keV) flux and/or spectral modulation with the Jupiter rotation period (∼10 h) as well as variations with the same period in solar wind parameters, radio and magnetic field directional data. In addition, characteristic strong ∼40 min periodic variations were found superimposed on the ∼10 h ion spectral modulation. Both the ∼10 h and ∼40 min ion periodicities in HI-SCALE measurements were present in several cases during the whole period examined (October 2003 to March 2004) and were found to be more evident during some special conditions, for instance during enhanced fluxes around the start (forward shock) and the end (reverse shock) of CIRs. We infer that the Jovian magnetosphere was triggered by the impact of the CIRs, after the Halloween events, and it was (a) a principal source of forward and reverse shock-associated ion flux structures and (b) the cause of generation of ∼10 h quasi-periodic magnetic field and plasma modulation observed by Ulysses at those times.     

3D Imaging of the OH mesospheric emissive layer

A new and original stereo imaging method is introduced to measure the altitude of the OH nightglow layer and provide a 3D perspective map of the altitude of the layer centroid. Near-IR photographs of the OH layer are taken at two sites separated by a 645 km distance. Each photograph is processed in order to provide a satellite view of the layer. When superposed, the two views present a common diamond-shaped area. Pairs of matched points that correspond to a physical emissive point in the common area are identified in calculating a normalized cross-correlation coefficient (NCC). This method is suitable for obtaining 3D representations in the case of low-contrast objects. An observational campaign was conducted in July 2006 in Peru. The images were taken simultaneously at Cerro Cosmos (12°09′08.2″ S, 75°33′49.3″ W, altitude 4630 m) close to Huancayo and Cerro Verde Tellolo (16°33′17.6″ S, 71°39′59.4″ W, altitude 2272 m) close to Arequipa. 3D maps of the layer surface were retrieved and compared with pseudo-relief intensity maps of the same region. The mean altitude of the emission barycenter is located at 86.3 km on July 26. Comparable relief wavy features appear in the 3D and intensity maps. It is shown that the vertical amplitude of the wave system varies as exp (Δz/2H) within the altitude range Δz = 83.5–88.0 km, H being the scale height. The oscillatory kinetic energy at the altitude of the OH layer is comprised between 3 × 10⁻⁴ and 5.4 × 10⁻⁴ J/m³, which is 2–3 times smaller than the values derived from partial radio wave at 52°N latitude.     

Reconstruction of topside density profile by using the topside sounder model profiler and digisonde data

To improve the accuracy of the real time topside electron density profiles given by the Digisonde software a new model-assisted technique is used. This technique uses the Topside Sounder Model (TSM), which provides the plasma scale height (H_s), O⁺–H⁺ transition height (H_T), and their ratio Rt = H_s/H_T, derived from topside sounder data of Alouette and ISIS satellites. The Topside Sounder Model Profiler (TSMP) incorporates TSM and uses the model quantities as anchor points in construction of topside density (Ne) profiles. For any particular location, TSMP calculates topside Ne profiles by specifying the values of foF2 and hmF2. In the present version, TSMP takes the F2 peak characteristics – foF2, hmF2, and the scale height at hmF2 – from the Digisonde measurements. The paper shows results for the Digisonde stations Athens and Juliusruh. It is found that the topside scale height used in Digisonde reconstruction is less than that extracted from topside sounder profiles. Rough comparison of their bulk distributions showed that they differ by an average factor of 1.25 for locations of Athens and Juliusruh. When the Digisonde scale heights are adjusted by this factor, the reconstructed topside profiles are close to those provided by TSM. Compared with CHAMP reconstruction profiles in two cases, TSMP/Digisonde profiles show lower density between 400 and 2000 km.     

An empirical model of electron density in low latitude at 600 km obtained by Hinotori satellite

An empirical model of electron density (Ne) was constructed by using the data obtained with an impedance probe on board Japanese Hinotori satellite. The satellite was in circular orbit of the height of 600 km with the inclination of 31 degrees from February 1981 to June 1982. The constructed model gives Ne at any local time with the time resolution of 90 min and between −25 and 25 degrees in magnetic latitude with its resolution of 5 degrees in the range of F_10.7 from 150 to 250 under the condition of K_p < 4. Spline interpolations are applied to the functions of day of year, geomagnetic latitude and solar local time, and linear interpolation is applied to the function of F_10.7. Longitude dependence of Ne is not taken into account. Our density model can reproduce solar local time variation of electron density at 600 km altitude better than current International Reference Ionosphere (IRI2001) model which overestimates Ne in night time and underestimates Ne in day time. Our density model together with electron temperature model which has been constructed before will enable more understanding of upper ionospheric phenomenon in the equatorial region.     

Testing the three-dimensional IRI-SIRMUP-P mapping of the ionosphere for disturbed periods

This paper describes the three-dimensional (3-D) electron density mapping of the ionosphere given as output by the assimilative IRI-SIRMUP-P (ISP) model for three different geomagnetic storms. Results of the 3-D model are shown by comparing the electron density profiles given by the model with the ones measured at two testing ionospheric stations: Roquetes (40.8°N, 0.5°E), Spain, and San Vito (40.6°N, 17.8°E), Italy. The reference ionospheric stations from which the autoscaled foF2 and M(3000)F2 data as well as the real-time vertical electron density profiles are assimilated by the ISP model are those of El Arenosillo (37.1°N, 353.3°E), Spain, Rome (41.8°N, 12.5°E), and Gibilmanna (37.9°N, 14.0°E), Italy. Overall, the representation of the ionosphere made by the ISP model is better than the climatological representation made by only the IRI-URSI and the IRI-CCIR models. However, there are few cases for which the assimilation of the autoscaled data from the reference stations causes either a strong underestimation or a strong overestimation of the real conditions of the ionosphere, which is in these cases better represented by only the IRI-URSI model. This ISP misrepresentation is mainly due to the fact that the reference ionospheric stations covering the region mapped by the model turn out to be few, especially for disturbed periods when the ionosphere is very variable both in time and in space and hence a larger number of stations would be required. The inclusion of new additional reference ionospheric stations could surely smooth out this concern.