A regional GNSS-VTEC model over Nigeria using neural networks: A novel approach

Okoh, Daniel; Owolabi, Oluwafisayo; Ekechukwu, Christopher; Folarin, Olanike Olufunmilayo; Arhiwo, Gila; Agbo, Joseph N.; Bolaji, O. S.; Rabiu, Babatunde

doi:10.1016/j.geog.2016.03.003

Cited by 53 publications

(45 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This indicates that IRI model is not as good as MIDAS and ANN techniques in making accurate storm‐time TEC reconstructions. A similar observation was highlighted in previous works that compared MIDAS reconstructions with IRI predictions (Chartier et al, ; Giday et al, ), and ANN estimations with IRI predictions (Habarulema et al, , ; Okoh et al, ; Watthanasangmechai et al, ). The underestimation of TEC by IRI model compared to MIDAS and ANN can be attributed to difference in altitude ranges TEC is estimated (Chartier et al, ; Habarulema & Ssessanga, ; Kenpankho et al, ).…”

Section: Resultssupporting

confidence: 84%

“…Recently, when ANN and empirical orthogonal function (EOF) models were applied to storm-time TEC modeling for a midlatitude station (Sutherland [SUTH]: 32.38 ∘ S, 20.81 ∘ E; South Africa), findings showed that ANN results agree with observations better than EOF predictions (Uwamahoro & Habarulema, 2015). Overall, ANN model has been reported to perform better than other storm-time TEC models such as IRI and EOF (Habarulema et al, 2007(Habarulema et al, , 2009Okoh et al, 2016;Uwamahoro & Habarulema, 2015;Watthanasangmechai et al, 2012).…”

Section: 1029/2017rs006499mentioning

confidence: 99%

“…Most of these studies reported a good agreement between the actual data and ANN model predictions. Comparative TEC results in low latitudes showed superior performance of ANN models over the IRI model (Okoh et al, ; Watthanasangmechai et al, ). A similar observation was reported for the African midlatitude region when evaluating the performance of ANNs and IRI during geomagnetically quiet conditions (Habarulema et al, , ) and geomagnetic storms (Habarulema et al, ).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Reconstruction of Storm‐Time Total Electron Content Using Ionospheric Tomography and Artificial Neural Networks: A Comparative Study Over the African Region

et al. 2018

View full text Add to dashboard Cite

The work presented here aims to evaluate the capabilities of Multi‐Instrument Data Analysis System (MIDAS) compared with artificial neural networks (ANNs) to reconstruct storm‐time total electron content (TEC) over the African low‐latitude and midlatitude regions. For MIDAS, the inversion was done based on the Global Positioning System (GPS) measurements from receiver stations extending from −30∘ to 36∘ in latitude and 30∘ to 44∘ in longitude while for ANNs, individual storm‐time models based on historical GPS data from receivers within the same region covered by MIDAS were used. Based on the minimum Dst index reached during the storm period, moderate (−50 nT ≥Dst> −100 nT), strong (−100 nT ≥Dst>1em−200 nT), and severe (−200 nT ≥Dst>1em−350 nT) storms were used for validation. MIDAS and ANNs results were compared with IRI‐2016 predictions and validated with real GPS TEC observations. A statistical analysis revealed that MIDAS and ANNs provide comparable results in storm‐time TEC reconstruction with average mean absolute errors of 4.81 and 4.18 TECU respectively. However, MIDAS performed better compared to ANNs in following TEC enhancements and depletions as well as short‐term features observed during the selected storm periods. In terms of latitude, it was found that on average, MIDAS performs 13% better than ANNs in the African midlatitude, while ANN model performs 24% better than MIDAS in low latitudes. Furthermore, comparisons with IRI predictions showed that both MIDAS and ANNs produce more accurate estimations of the storm‐time TEC than IRI model.

show abstract

Section: Resultssupporting

confidence: 84%

Section: 1029/2017rs006499mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Reconstruction of Storm‐Time Total Electron Content Using Ionospheric Tomography and Artificial Neural Networks: A Comparative Study Over the African Region

et al. 2018

View full text Add to dashboard Cite

show abstract

“…Neural network forecasts are derived from nonlinear, statistical algorithms that determine and model complex relationships between inputs and outputs to find patterns in the data that can be extrapolated. Some researches (e.g., Habarulema, ; Okoh et al, ) have demonstrated that neural networks are efficient for modeling ionospheric variations that depend mostly on the Sun's activities. For parameters like SSN that have shown variations in the SC properties (e.g., the value of SSN at the cycle peaks and the cycle durations), neural networks can be guided in techniques that combine other methods to increase their accuracies.…”

Section: Introductionmentioning

confidence: 99%

A Hybrid Regression‐Neural Network (HR‐NN) Method for Forecasting the Solar Activity

et al. 2018

Self Cite

View full text Add to dashboard Cite

The Sun is the major driver of space weather events, and as a result, most applications requiring modeling/forecasting of space weather phenomena depend largely on the activities of Sun. Accurate modeling of solar activity parameters like the sunspot number (SSN) is therefore considered significant for the quantitative modeling of space weather phenomena. Sunspot number forecasts are applied in ionospheric models like the International Reference Ionosphere model and in several other projects requiring prediction of space weather phenomena. A method called Hybrid Regression‐Neural Network that combines regression analysis and neural network learning is used for forecasting the SSN. Considering the geomagnetic Ap index during the end of the previous cycle (known as the precursor Ap index) as a reliable measurement, we predict the end of solar cycle 24 to be in March 2020 (±7 months), with monthly SSN 5.4 (±5.5). Using an estimated value of precursor Ap index as 5.6 nT for solar cycle 25, we predict the maximum SSN to be 122.1 (±18.2) in January 2025 (±6 months) and the minimum to be 6.0 (±5.5) in April 2031 (±5 months). We found from the model that on changing the assumed value of precursor Ap index (5.6 nT) by ±1 nT, the predicted peak of solar cycle 25 changes by about 11 sunspots for every 1‐nT change in the assumed precursor Ap index.

show abstract

“…During quiet conditions, the results showed that the predicted TEC 10.1029/2018JA025455 is in good agreement with the actual data (Cander, 1998;Leandro & Santos, 2007;Ratnam et al, 2012). Comparisons of IRI and neural network (NN) predictions during geomagnetically quiet conditions revealed that NN makes more accurate TEC predictions than IRI (Habarulema et al, 2007;Habarulema, McKinnell, Cilliers, et al, 2009;Okoh et al, 2016;Watthanasangmechai et al, 2012). Results have, however, shown that, although NN model can follow TEC dynamics during geomagnetic storms, the accuracy is still low and improvements are required (Habarulema et al, 2010(Habarulema et al, , 2011Uwamahoro & Habarulema, 2015;Watthanasangmechai et al, 2012).…”

Section: Introductionmentioning

confidence: 99%

Storm Time Total Electron Content Modeling Over African Low‐Latitude and Midlatitude Regions

2018

View full text Add to dashboard Cite

This paper presents storm time total electron content (TEC) modeling results based on artificial neural networks, for both low‐latitude and midlatitude African regions. The developed storm time TEC models were based on Global Positioning System (GPS) observations from GPS receiver stations selected in low latitude, Northern and Southern hemisphere midlatitude regions of the African sector. GPS data selection was based on a storm criterion of Dst≤ −50 nT and storm data sets used to develop the models were within the periods 2001–2015, 2000–2015, and 1998–2015, for African low latitude, Northern and Southern Hemisphere midlatitude regions, respectively. For the first time in storm time TEC modeling, the meridional wind velocity was introduced as an additional input to the well‐known TEC modeling inputs (diurnal variation, seasonal variation, solar activity, and geomagnetic activity representations) to take into account the effect of neutral winds in moving ionization within the ionosphere along the magnetic field lines. Results showed that the use of meridional wind as an additional input leads to overall percentage improvements of about 5%, 10%, and 5% for the low‐latitude, Northern and Southern Hemisphere midlatitude regions, respectively. High‐latitude storm‐induced winds and the interhemispheric blows of the meridional winds from summer to winter hemisphere may be associated with these improvements.

show abstract

A regional GNSS-VTEC model over Nigeria using neural networks: A novel approach

Cited by 53 publications

References 9 publications

Reconstruction of Storm‐Time Total Electron Content Using Ionospheric Tomography and Artificial Neural Networks: A Comparative Study Over the African Region

Reconstruction of Storm‐Time Total Electron Content Using Ionospheric Tomography and Artificial Neural Networks: A Comparative Study Over the African Region

A Hybrid Regression‐Neural Network (HR‐NN) Method for Forecasting the Solar Activity

Storm Time Total Electron Content Modeling Over African Low‐Latitude and Midlatitude Regions

Contact Info

Product

Resources

About