Ionospheric TEC Forecasting over an Indian Low Latitude Location Using Long Short-Term Memory (LSTM) Deep Learning Network

The forecasting of ionospheric electron density has been great interest to the research scientists and engineers’ community as it significantly influences satellite-based navigation, positioning, communication applications under influence space weather. Hence, present paper adopts a long short-term memory (LSTM) deep learning network model forecast total content (TEC) by exploiting global positioning system (GPS) observables, at low latitude Indian location in Bangalore (IISC; Geographic 13.03° N 77.57° E), during 24th solar cycle. proposed uses about eight years GPS-TEC data (from 2009 2017) for training validation, whereas 2018 was used independent testing TEC. Apart from input TEC parameters, considers sequential geophysical indices realize effects. performance is evaluated comparing forecasted values with observed empirical ionosphere (international reference ionosphere; IRI-2016) through set validation metrics. analysis results test period showed that LSTM output closely followed relatively minimal root mean square error (RMSE) 1.6149 highest correlation coefficient (CC) 0.992, compared IRI-2016. Furthermore, day-to-day validated year 2018, inferring outcomes are better than IRI-2016 considered location. Implementation other latitudinal locations region suggested an efficient regional across region. work complements efforts towards establishing delays irregularities, which responsible degrading static, well dynamic, space-based navigation performances.