An Autoregressive Integrated Moving Average (ARIMA) Based Forecasting of Ionospheric Total Electron Content at a low latitude Indian Location

Abstract

Ionospheric total electron content (TEC) plays an important role in introducing delay errors in space-based navigation and communication signals and requires early forecasting of the plausible impacts on the relying systems. In the present work, an autoregressive integrated moving average (ARIMA) is implemented in the time series analysis to forecast the TEC at an Indian low latitude location (KL University, Guntur; Geographic 16.37°N, 80.37°E) during the quiet (5–9 January 2021) and disturbed (3–7 March 2022) geomagnetic conditions. The performance of the model is evaluated from the biases, root mean square error (RMSE), and correlation coefficients between model forecast and observed TEC. The results show that bias remains between -3 to +3 TECU and +2 to-4 during quiet and disturbed days, respectively. The corresponding RMSE values are within a limit of 5 TECU and 6 TECU. The occurrence of plasma irregularities is also verified by analyzing the scintillation indices during the period. A further analysis refinement of the model is aimed to improve the forecasting accuracy over the region.