Variation of hmF2 and NmF2 deduced from DPS-4 over Multan (Pakistan) and their comparisons with IRI-2012 & IRI-2016 during the deep solar minimum between cycles 23 & 24

We report the results of ionospheric measurements from DPS-4 installed at Multan (Geog coord. 30.18°N, 71.48°E, dip 47.4°). The variations in F2-layer maximum electron density NmF2 and its peak height hmF2 are studied during the deep solar minimum between cycles 23 & 24 i.e 2008–2009 with comparisons conducted with the International Reference Ionosphere (IRI) versions 2012 & 2016. We find that the hmF2 observations peak around the pre-sunrise and sunrise hours depending on the month. Seasonally, the daytime variation of NmF2 is higher in the Equinox and Summer, while daytime hmF2 are slightly higher in the Equinox and Winter. High values of hmF2 around midnight are caused by an increase of upward drifts produced by meridional winds. The ionosphere over Multan, which lies at the verge of low and mid latitude, is affected by both $E\times B$ drifts and thermospheric winds as evident from mid-night peaks and near-sunrise dips in hmF2. The results of the comparison of the observed NmF2 and hmF2 for the year 2008–2009 with the IRI-2012 (both NmF2 and hmF2) and IRI-2016 (only hmF2) estimates indicate that for NmF2, IRI-2012 with Consultative Committee International Radio (CCIR) option produces values in better agreement with observed data. Whereas, for hmF2, IRI-2016 with both International Union of Radio Science (URSI) and CCIR SHU-2015 options, predicts well for nighttime hours throughout the year. However, the IRI-2012 with CCIR option produces better agreement with data during daytime hours. Furthermore, IRI-2012 with CCIR option gives better results during Equinox months, whereas, IRI-2016 with both URSI and CCIR SHU-2015 options predict well for Winter and Summer.     

Ionospheric variation at Thailand equatorial latitude station: Comparison between observations and IRI-2001 model predictions

In this paper, the F2-layer critical frequency (foF2) and peak height (hmF2) measured by the FM/CW ionosonde at Thailand equatorial latitude station, namely Chumphon (10.72°N, 99.37°E, dip 3.22) are presented. The measurement data during low solar activity from January 2004 to December 2006 are analyzed based on the diurnal, seasonal variation. The results are then compared with IRI-2001 model predictions. Our study shows that: (1) In general, both the URSI and CCIR options of the IRI model give foF2 close to the measured ones, but the CCIR option produces a smaller range of deviation than the URSI option. The agreement during daytime is generally better than during nighttime. Overestimation mostly occurs in 2004 and 2006, while underestimation is during pre-sunrise hours in June solstice in 2005. The peak foF2 around sunset is higher during March equinox and September equinox than the other seasons, with longer duration of maximum levels in March equinox than September equinox. Large coefficients of variability foF2 occur during pre-sunrise hours. Meanwhile, the best agreement between the observed foF2 and the IRI model is obtained in June solstice. (2) In general, The IRI (CCIR) model predicts the observed hmF2 well during daytime in June solstice from 2004–2006, but it overestimates during March equinox, September equinox and December solstice. For nighttime, the model overestimates hmF2 values for all seasons especially during March equinox and September equinox. However, the model underestimates hmF2 values during September equinox and for some cases during June solstice and December solstice at pre-sunrise. The agreement between the IRI model and the hmF2(M3000_OBS) is worst around noontime, post-sunset and pre-sunrise hours. All comparative studies give feedback for new improvements of CCIR and URSI IRI models.     

Ionospheric behavior of the F2 peak parameters foF2 and hmF2 at Hainan and comparisons with IRI model predictions

Monthly median values of foF2, hmF2 and M(3000)F2 parameters, with hourly time interval resolution for the diurnal variation, obtained with DPS-4 digisonde observations at Hainan (19.4°N, 109.0°E) are used to study the low latitude ionospheric variation behavior. The observational results are compared with the International Reference Ionospheric Model (IRI) predictions. The time period coverage of the data used for the present study is from March 2002 to February 2005. Our present study showed that: (1) In general, IRI predictions using CCIR and URSI coefficients follow well the diurnal and seasonal variation patterns of the experimental values of foF2. However, CCIR foF2 and URSI foF2 IRI predictions systematically underestimate the observed results during most time period of the day, with the percentage difference ΔfoF2 (%) values changing between about −5% and −25%, whereas for a few hours around pre-sunrise, the IRI predictions generally overestimate the observational ones with ΔfoF2 (%) sometimes reaching as large as ∼30%. The agreement between the IRI results and the observational ones is better for the year 2002 than for the other years. The best agreement between the IRI results and the observational ones is obtained in summer when using URSI coefficients, with the seasonal average values of ΔfoF2 (%) being within the limits of ±10%. (2) In general, the IRI predicted hmF2 values using CCIR M(3000)F2 option shows a poor agreement with the observational results. However, when using the measured M(3000)F2 as input, the diurnal variation pattern of hmF2 given by IRI2001 has a much better agreement with the observational one with the detailed fine structures including the pre-sunrise and post-sunset peaks reproduced reasonably well. The agreement between the IRI predicted hmF2 values using CCIR M(30,000)F2 option and the observational ones is worst for the afternoon to post-midnight hours for the high solar activity year 2002. During daytime hours the agreement between the hmF2 values obtained with CCIR M(30,000)F2 option and the observational ones is best for summer season. The discrepancy between the observational hmF2 and that obtained with CCIR M(30,000)F2 option stem from the CCIR M(3000)F2 model, which does not produce the small scale structures observed in the measured M(3000)F2.     

The response of African equatorial foF2 to geomagnetic storms: Comparison between observations and IRI-2007 predictions

This study examines the response of the African equatorial ionospheric foF2 to different levels of geomagnetic storms, using the foF2 hourly data for the year 1989 from Ouagadougou (12.4°N, 1.5°W, dip: 2.8°N). The study also compares the observed data for the selected storm periods with the latest IRI model (IRI-2007). The foF2 values (both observed and predicted) show typical features of daytime peak and post-midnight minimum peak. The response of the ionospheric foF2 over Ouagadougou to storms events, during the night-time and post-midnight hours indicates negative responses of the ionospheric foF2, while that of the daytime hours indicates positive responses. For the investigation on the variability of the observed foF2 with respect to IRI-2007 model, with the exception of the analysis of the 20–22, October, 1989 data, where a midday peak was also observed on the first day, this study reveals two characteristic daily foF2 variability peaks: post-midnight and evening peaks. In addition, for all the geomagnetic storms considered, the URSI and CCIR coefficients of the IRI-2007 model show reasonable correspondence with each other, except for some few discrepancies. For instance, the event of 28–30 August, 1989 shows comparatively higher variability for the URSI coefficient, and at the foF2 peak values, the event of 20–22 October, 1989 shows that the CCIR coefficient is more susceptible to foF2 variability than the URSI coefficient. This study is aimed at providing African inputs for the future improvement of the IRI model.     

Study the behavior of the ionospheric frequencies at Karachi during three consecutive solar minima of cycle 21, 22 & 23 and their comparison with IRI-2016

The behavior of critical frequencies of ionospheric E and F2 layers (foE & foF2) along with minimum ionospheric frequency (fmin) is studied for solar minima of cycle 21 (1986), 22 (1996) and 23 (2008) over Karachi (24.95$°$N, 67.13$°$E), Pakistan. The station is located at the crest of equatorial ionization anomaly region. Beside seasonal differences, pronounced change in the values of frequencies is noted from one solar minimum to another solar minimum. A strong and direct correlation of foF2 with Smoothed Sunspot Number (SSN) and F10.7?cm solar flux is observed. In the minimum of cycle 23, reduction in foF2 is noted due to reduction of solar EUV as compared to other minima. Also disappearance of semi-annual variations in foF2 is noted in cycle 23 minimum. Unexpectedly higher values of foE and fmin are observed in minimum of cycle 23 as compared to other minima. It is difficult to explain this unusual behavior of fmin and foE along with disappearance of semi-annual variation in foF2. It is possible that during very low solar activity, thermospheric conditions are changed which in turn altered the ionosphere. Further investigation of atmosphere-ionosphere coupling is required to understand this complex behavior. On comparison of observed values with IRI-2016, higher deviations are observed in foE before noon hours while in case of foF2, large deviations are noted during daytime. The absence of foF2 semi-annual variation in cycle 23 is not reproduced by IRI-2016. It is suggested that IRI-2016 need some modification for extremely low solar activity condition.     

Evaluation of the improvement of IRI-2016 over IRI-2012 at the India low-latitude region during the ascending phase of cycle 24

This paper investigated the performance of the latest International Reference Ionosphere model (IRI-2016) over that of IRI-2012 in predicting total electron content (TEC) at three different stations in the Indian region. The data used were Global Positioning System (GPS) data collected during the ascending phase of solar cycle 24 over three low-latitude stations in India namely; Bangalore (13.02°N Geographic latitude, 77.57°E Geographic longitude), Hyderabad (17.25°N Geographic latitude, 78.30°E Geographic longitude) and Surat (21.16°N Geographic latitude, 72.78°E Geographic longitude). Monthly, the seasonal and annual variability of GPS-TEC have been compared with those derived from International Reference Ionosphere IRI-2016 and IRI-2012 with two different options of topside electron density: NeQuick and IRI01-corr. It is observed that both versions of IRI (i.e., IRI-2012 and IRI-2016) predict the GPS-TEC with some deviations, the latest version of IRI (IRI-2016) predicted the TEC similar to those predicted by IRI-2012 for all the seasons at all stations except for morning hours (0500 LT to 1000?LT). This shows that the effect of the updated version is seen only during morning hours and also that there is no change in TEC values by IRI-2016 from those predicted by IRI-2012 for the rest of the time of the day in the Indian low latitude region. The semiannual variations in the daytime maximum values of TEC are clearly observed from both GPS and model-derived TEC values with two peaks around March-April and September-October months of each year. Further, the Correlation of TEC derived by IRI-2016 and IRI-2012 with EUV and F10.7 shows similar results. This shows that the solar input to the IRI-2016 is similar to IRI 2012. There is no significant difference observed in TEC, bottom-side thickness (B0) and shape (B1) parameter predictions by both the versions of the IRI model. However, a clear improvement is visible in hmF2 and NmF2 predictions by IRI-2016 to that by IRI-2012. The SHU-2015 option of the IRI-2016 gives a better prediction of NmF2 for all the months at low latitude station Ahmedabad compare to AMTB 2013.     

Comparison of peak characteristics of F2 ionospheric layer over Tehran region at a low solar activity period with IRI-2001 and IRI-2007 models predictions

In the present work values of peak electron density (NmF2) and height of F2 ionospheric layer (hmF2) over Tehran region at a low solar activity period are compared with the predictions of the International Reference Ionosphere models (IRI-2001 and IRI-2007). Data measured by a digital ionosonde at the ionospheric station of the Institute of Geophysics, University of Tehran from July 2006 to June 2007 are used to perform the calculations. Formulations proposed by  and  are utilized to calculate the hmF2. The International Union of Radio Science (URSI) and International Radio Consultative Committee (CCIR) options are employed to run the IRI-2001 and IRI-2007 models. Results show that both IRI-2007 and IRI-2001 can successfully predict the NmF2 and hmF2 over Tehran region. In addition, the study shows that predictions of IRI-2007 model with CCIR coefficient has closer values to the observations. Furthermore, it is found that the monthly average of the percentage deviation between the IRI models predictions and the values of hmF2 and NmF2 parameters are less than 10% and 21%, respectively.     

A new global F2 peak electron density model for the International Reference Ionosphere (IRI)

A new neural network (NN) based global empirical model for the F2 peak electron density (NmF2) has been developed using extended temporal and spatial geophysical relevant inputs. Measured ground based ionosonde data, from 84 global stations, spanning the period 1995 to 2005 and, for a few stations from 1976 to 1986, obtained from various resources of the World Data Centre (WDC) archives (Space Physics Interactive Data Resource SPIDR, the Digital Ionogram Database, DIDBase, and IPS Radio and Space Services) have been used for training a NN. The training data set includes all periods of quiet and disturbed magnetic activity. A comprehensive comparison for all conditions (e.g., magnetic storms, levels of solar activity, season, different regions of latitudes, etc.) between foF2 value predictions using the NN based model and International Reference Ionosphere (IRI) model (including both the International Union of Radio Science (URSI) and International Radio Consultative Committee (CCIR) coefficients) with observed values was investigated. The root-mean-square (RMS) error differences for a few selected stations are presented in this paper. The results of the foF2 NN model presented in this work successfully demonstrate that this new model can be used as a replacement option for the URSI and CCIR maps within the IRI model for the purpose of F2 peak electron density predictions.     

Comparison of GPS-TEC measurements with NeQuick2 and IRI model predictions in the low latitude East African region during varying solar activity period (1998 and 2008–2015)

This paper examines the performances of NeQuick2, the latest available IRI-2016, IRI-2012 and IRI-2007 models in describing the monthly and seasonal mean total electron content (TEC) over the East African region. This is to gain insight into the success of the various model types and versions at characterizing the ionosphere within the equatorial ionization anomaly. TEC derived from five Global Positioning System (GPS) receivers installed at Addis Ababa (ADD, 5.33°N, 111.99°E Geog.), Asab (ASAB, 8.67°N, 116.44°E Geog.), Ambo (ABOO, 5.43°N, 111.05°E Geog.), Nairobi (RCMN, ?4.48°N, 108.46°E Geog.) and Nazret (NAZR, 4.78°N, 112.43°E Geog.), are compared with the corresponding values computed using those models during varying solar activity period (1998 and 2008–2015). We found that different models describe the equatorial and anomaly region ionosphere best depending on solar cycle, season and geomagnetic activity levels. Our results show that IRI-2016 is the best model (compared to others in terms of discrepancy range) in estimating the monthly mean GPS-TEC at NAZR, ADD and RCMN stations except at ADD during 2008 and 2012. It is also found that IRI-2012 is the best model in estimating the monthly mean TEC at ABOO station in 2014. IRI show better agreement with observations during June solstice for all the years studied at ADD except in 2012 where NeQuick2 better performs. At NAZR, NeQuick2 better performs in estimating seasonal mean GPS-TEC during 2011, while IRI models are best during 2008–2009. Both NeQuick2 and IRI models underestimate measured TEC for all the seasons at ADD in 2010 but overestimate at NAZR in 2009 and RCMN in 2008. The periodic variations of experimental and modeled TEC have been compared with solar and geomagnetic indices at ABOO and ASAB in 2014 and results indicate that the F10.7 and sunspot number as indices of solar activity seriously affects the TEC variations with periods of 16–32?days followed by the geomagnetic activity on shorter timescales (roughly periods of less than 16?days). In this case, NeQuick2 derived TEC shows better agreement with a long term period variations of GPS-TEC, while IRI-2016 and IRI-2007 show better agreement with observations during short term periodic variations. This indicates that the dependence of NeQuick2 derived TEC on F10.7 is seasonal. Hence, we suggest that representation of geomagnetic activity indices is required for better performance over the low latitude region.     

TEC variability near northern EIA crest and comparison with IRI model

Monthly median values of hourly total electron content (TEC) is obtained with GPS at a station near northern anomaly crest, Rajkot (geog. 22.29°N, 70.74°E; geomag. 14.21°N, 144.9°E) to study the variability of low latitude ionospheric behavior during low solar activity period (April 2005 to March 2006). The TEC exhibit characteristic features like day-to-day variability, semiannual anomaly and noon bite out. The observed TEC is compared with latest International Reference Ionosphere (IRI) – 2007 model using options of topside electron density, NeQuick, IRI01-corr and IRI-2001 by using both URSI and CCIR coefficients. A good agreement of observed and predicted TEC is found during the daytime with underestimation at other times. The predicted TEC by NeQuick and IRI01-corr is closer to the observed TEC during the daytime whereas during nighttime and morning hours, IRI-2001 shows lesser discrepancy in all seasons by both URSI and CCIR coefficients.     

Comparison of ionospheric F2 peak parameters foF2 and hmF2 with IRI2001 at Hainan

Monthly median values of foF2, hmF2 and M(3000)F2 parameters, with quarter-hourly time interval resolution for the diurnal variation, obtained with DPS4 digisonde at Hainan (19.5°N, 109.1°E; Geomagnetic coordinates: 178.95°E, 8.1°N) are used to investigate the low-latitude ionospheric variations and comparisons with the International Reference Ionosphere (IRI) model predictions. The data used for the present study covers the period from February 2002 to April 2007, which is characterized by a wide range of solar activity, ranging from high solar activity (2002) to low solar activity (2007). The results show that (1) Generally, IRI predictions follow well the diurnal and seasonal variation patterns of the experimental values of foF2, especially in the summer of 2002. However, there are systematic deviation between experimental values and IRI predictions with either CCIR or URSI coefficients. Generally IRI model greatly underestimate the values of foF2 from about noon to sunrise of next day, especially in the afternoon, and slightly overestimate them from sunrise to about noon. It seems that there are bigger deviations between IRI Model predictions and the experimental observations for the moderate solar activity. (2) Generally the IRI-predicted hmF2 values using CCIR M(3000)F2 option shows a poor agreement with the experimental results, but there is a relatively good agreement in summer at low solar activity. The deviation between the IRI-predicted hmF2 using CCIR M(3000)F2 and observed hmF2 is bigger from noon to sunset and around sunrise especially at high solar activity. The occurrence time of hmF2 peak (about 1200 LT) of the IRI model predictions is earlier than that of observations (around 1500 LT). The agreement between the IRI hmF2 obtained with the measured M(3000)F2 and the observed hmF2 is very good except that IRI overestimates slightly hmF2 in the daytime in summer at high solar activity and underestimates it in the nighttime with lower values near sunrise at low solar activity.     

Ionospheric variability at taiwan low latitude station: Comparison between observations and IRI-2001 model

This paper presents the observed ionospheric F-region critical frequency, foF2, and peak height, hmF2, at northern crest of equatorial ionization anomaly (EIA) area station, namely Chung-Li (24.9°N, 121.1°E, dip 35°), and to be compared with International Reference Ionosphere model (IRI-2001) predictions for the period from 1994 to 1999, corresponding to half of the 23rd solar cycle. The diurnal and seasonal variation of foF2 and hmF2 are analyzed for different solar phases, respectively. The result shows the largest discrepancies were observed during nighttime for foF2 and hmF2, respectively. The value of foF2 both CCIR and URSI selected in the IRI model produced a good agreement during the daytime and underestimated during the noon time for high solar activities. The underestimation at noon time is mainly caused by the fountain effect from equator. Further, the peak height hmF2 shows a larger variability around the midnight than daytime in the equinox and winter seasons and reserved in summer, respectively. The study shows that the monthly median values of observed hmF2 is somewhat lower than those predicated by the IRI model, at night time in all the seasons except the period of 04:00–06:00 LT and reverse at daytime in summer. In general the IRI model predictions with respect to the observed in hmF2 is much better than foF2. The percentage deviation of the observed foF2 (hmF2) values with respect to the IRI model varies from 5% to 80% (0–25%) during nighttime and 2–17% (0–20%) at daytime, respectively. In general, the model generates good results, although some improvements are still necessary to implement in order to obtain better simulations for ionospheric low-latitudes region.     

NeQuick 2 and IRI Plas VTEC predictions for low latitude and South American sector

Using vertical total electron content (VTEC) measurements obtained from GPS satellite signals the capability of the NeQuick 2 and IRI Plas models to predict VTEC over the low latitude and South American sector is analyzed. In the present work both models were used to calculate VTEC up to the height of GPS satellites. Also, comparisons between the performance of IRI Plas and IRI 2007 have been done. The data correspond to June solstice and September equinox 1999 (high solar activity) and they were obtained at nine stations. The considered latitude range extends from 18.4°N to ?64.7°N and the longitude ranges from 281.3°E to 295.9°E in the South American sector. The greatest discrepancies among model predictions and the measured VTEC are obtained at low latitudes stations placed in the equatorial anomaly region. Underestimations as strong as 40?TECU [1?TECU?=?10¹⁶?m^?2] can be observed at BOGT station for September equinox, when NeQuick2 model is used. The obtained results also show that: (a) for June solstice, in general the performance of IRI Plas for low latitude stations is better than that of NeQuick2 and, vice versa, for highest latitudes the performance of NeQuick2 is better than that of IRI Plas. For the stations TUCU and SANT both models have good performance; (b) for September equinox the performances of the models do not follow a clearly defined pattern as in the other season. However, it can be seen that for the region placed between the Northern peak and the valley of the equatorial anomaly, in general, the performance of IRI Plas is better than that of NeQuick2 for hours of maximum ionization. From TUCU to the South, the best TEC predictions are given by NeQuick2.The source of the observed deviations of the models has been explored in terms of CCIR foF2 determination in the available ionosonde stations in the region. Discrepancies can be also related to an unrealistic shape of the vertical electron density profile and or an erroneous prediction of the plasmaspheric contribution to the vertical total electron content. Moreover, the results of this study could be suggesting that in the case of NeQuick, the underestimation trend could be due to the lack of a proper plasmaspheric model in its topside representation. In contrast, the plasmaspheric model included in IRI, leads to clear overestimations of GPS derived TEC.     

Monthly mean foF2 model for an African low-latitude station and comparison with IRI

We have employed the hourly values of the ionospheric F-region critical frequency (foF2) obtained from Ouagadougou ionosonde, Burkina Faso (geographic coordinates 12° N, 1.8° W) during the interval of 1985–1995 (solar cycle 22) and solar radio flux of 10 cm wavelength (F10.7) to develop a local model (LM) for the African low-latitude station. The model was developed from regression analysis method, using the two-segmented regression analysis. We validated LM with foF2 data from Korhogo observatory, Cote d’Ivorie (geographical coordinates 9.3° N, 5.4° W). LM as well as the International Reference Ionosphere (IRI) agrees well with observations. LM gave some improvement on the IRI-predicted foF2 values at the sunrise (06 LT) at all solar flux levels and in all seasons except June solstice. The performance of the models at the representing the salient features of the equatorial foF2 was presented. Considering daytime and nighttime performances, LM and IRI are comparable in low solar activity (LSA), LM performed better than IRI in moderate solar activity (MSA), while IRI performed better than LM in high solar activity (HSA). CCIR has a root mean square error (r.m.s.e), which is only 0.10 MHz lower than that of LM while LM has r.m.s.e, which is about 0.05 MHz lower than that of URSI. In general, our result shows that performance of IRI, especially the CCIR option of the IRI, is quite comparable with the LM. The improved performance of IRI is a reflection of the numerous contributions of ionospheric physicists in the African region, larger volume of data for the IRI and the diversity of data sources, as well as the successes of the IRI task force activities.     

Validation of the STORM model in IRI-2001 at a high latitude station

The International Reference Ionosphere IRI-2001 model contains geomagnetic activity dependence based on an empirical storm time ionospheric correction (STORM model). An extensive validation of the STORM model for the middle latitude region has been performed. In this paper the ability of the STORM model to predict foF2 values at high latitudes is analyzed. For this, ionosonde data obtained at Base Gral. San Martin (68.1°S, 293°E) are compared with those obtained by the IRI-2001 model with or without storm correction during four geomagnetic storms that occurred in 2000 (Rz₁₂ = 117) and 2001 (Rz₁₂ = 111). The results show that predicted values with the STORM model follow the behaviour of foF2 experimental data better than without the STORM model. The relative deviation between measured and predicted foF2 reaches values of up to 24% and 43% with and without the STORM model in IRI-2001, during the main phase of the storms. In order to explain increases of electron density that occurred prior to the storm onset and also decreases of electron density observed during the first part of the recovery of the storm, possible physical mechanisms are discussed.     

Comparison between observed ionospheric foF2 and IRI-2001 predictions over periods of severe geomagnetic activities at Grahamstown,South Africa

The observed ionospheric F2 critical frequency (foF2) values over a South Africa mid-latitude station, Grahamstown, (geographic coordinates: 33.3°S, 26.5°E), were analysed and compared with International Reference Ionosphere (IRI) model, using the CCIR (Comite´ Consultatif International des Radio communications) and URSI (Union Radio-Scientifique Internationale) coefficients, during four geomagnetically disturbed days in the year 2000. These days are April 5, May 23, August 10 and September 15. The data were analysed for five days around the storm day. Comparisons between the IRI-2001 predicted foF2 values, using both CCIR and URSI coefficients and the observed values are shown with their root-mean-square error (RMSE) and the relative deviation module mean (rdmm) for the various storm periods. The CCIR option performed more accurately than the URSI option.     

A prediction model of short-term ionospheric foF2 based on AdaBoost

In this paper, the AdaBoost-BP algorithm is used to construct a new model to predict the critical frequency of the ionospheric F2-layer (foF2) one hour ahead. Different indices were used to characterize ionospheric diurnal and seasonal variations and their dependence on solar and geomagnetic activity. These indices, together with the current observed foF2 value, were input into the prediction model and the foF2 value at one hour ahead was output. We analyzed twenty-two years’ foF2 data from nine ionosonde stations in the East-Asian sector in this work. The first eleven years’ data were used as a training dataset and the second eleven years’ data were used as a testing dataset. The results show that the performance of AdaBoost-BP is better than those of BP Neural Network (BPNN), Support Vector Regression (SVR) and the IRI model. For example, the AdaBoost-BP prediction absolute error of foF2 at Irkutsk station (a middle latitude station) is 0.32 MHz, which is better than 0.34 MHz from BPNN, 0.35 MHz from SVR and also significantly outperforms the IRI model whose absolute error is 0.64 MHz. Meanwhile, AdaBoost-BP prediction absolute error at Taipei station from the low latitude is 0.78 MHz, which is better than 0.81 MHz from BPNN, 0.81 MHz from SVR and 1.37 MHz from the IRI model. Finally, the variety characteristics of the AdaBoost-BP prediction error along with seasonal variation, solar activity and latitude variation were also discussed in the paper.     

Bottom side profiles for two close stations at the southern crest of the EIA: Differences and comparison with IRI-2012 and NeQuick2 for low and high solar activity

Bottom side electron density profiles for two stations at the southern crest of the Equatorial Ionization Anomaly (EIA), São José dos Campos (23.1°S, 314.5°E, dip latitude 19.8°S; Brazil) and Tucumán (26.9°S, 294.6°E, dip latitude 14.0°S; Argentina), located at similar latitude and separated by only 20° in longitude, have been compared during equinoctial, winter and summer months under low (year 2008, minimum of the solar cycle 23/24) and high solar activity (years 2013–2014, maximum of the solar cycle 24) conditions. An analysis of parameters describing the bottom side part of the electron density profile, namely the peak electron density NmF2, the height hmF2 at which it is reached, the thickness parameter B₀ and the shape parameter B₁, is carried out. Further, a comparison of bottom side profiles and F-layer parameters with the corresponding outputs of IRI-2012 and NeQuick2 models is also reported. The variations of NmF2 at both stations reveal the absence of semi-annual anomaly for low solar activity (LSA), evidencing the anomalous activity of the last solar minimum, while those related to hmF2 show an uplift of the ionosphere for high solar activity (HSA). As expected, the EIA is particularly visible at both stations during equinox for HSA, when its strength is at maximum in the South American sector. Despite the similar latitude of the two stations upon the southern crest of the EIA, the anomaly effect is more pronounced at Tucumán than at São José dos Campos. The differences encountered between these very close stations suggest that in this sector relevant longitudinal-dependent variations could occur, with the longitudinal gradient of the Equatorial Electrojet that plays a key role to explain such differences together with the 5.8° separation in dip latitude between the two ionosondes. Furthermore at Tucumán, the daily peak value of NmF2 around 21:00 LT during equinox for HSA is in temporal coincidence with an impulsive enhancement of hmF2, showing a kind of “elastic rebound” under the action of the EIA. IRI-2012 and NeQuick2 bottom side profiles show significant deviations from ionosonde observations. In particular, both models provide a clear underestimation of the EIA strength at both stations, with more pronounced differences for Tucumán. Large discrepancies are obtained for the parameter hmF2 for HSA during daytime at São José dos Campos, where clear underestimations made by both models are observed. The shape parameter B₀ is quite well described by the IRI-2012 model, with very good agreement in particular during equinox for both stations for both LSA and HSA. On the contrary, the two models show poor agreements with ionosonde data concerning the shape parameter B₁.     

Progress towards a new global foF2 model for the International Reference Ionosphere (IRI)

A new neural network (NN) based global empirical model for the foF2 parameter, which represents the peak electron density has been developed using extended temporal and spatial geophysical relevant inputs. The first results from this new model were presented at the International Reference Ionosphere (IRI) 2006 workshop in Buenos Aires, Argentina, and showed that this new model would be a suitable replacement for the URSI and CCIR maps currently used within the IRI model for the purpose of F2 peak electron density predictions. Measured ground based ionosonde data, from 85 global stations, spanning the period 1995–2005 and, for a few stations from 1976 to 1986, obtained from various resources of the World Data Centre (WDC) archives (Space Physics Interactive Data Resource SPIDR, the Digital Ionogram Database, DIDBase, and IPS Radio and Space Services) have been used for training a NN.     

Calibration of IRI–ITU-R peak density and height over the oceans with topside sounding data

Accuracy of IRI electron density profile depends on the F2 layer peak density and height converted by empirical formulae from the critical frequency and M3000F2 factor provided by the ITU-R (former CCIR). The CCIR/ITU-R maps generated from ground-based ionosonde measurements suffer from model assumptions, in particular, over the oceans where relatively few measurements are available due to a scarcity of ground-based ionosondes. In the present study a grid-point calibration of IRI/ITU-R maps for the foF2 and hmF2 over the oceans is proposed using modeling results based on the topside true-height profiles provided by ISIS1, ISIS2, IK-19 and Cosmos-1809 satellites for the period of 1969–1987. Topside soundings results are compared with IRI and the Russian standard model of ionosphere, SMI, and grouped to provide an empirical calibration coefficient to the peak density and height generated from ITU-R maps. The grid-point calibration coefficients maps are produced in terms of the solar activity, geodetic latitude and longitude, universal time and season allowing update of IRI–ITU-R predictions of the F2 layer peak parameters.