Optimized ensemble single frequency ionospheric TEC model over the low latitude Indian region for GNSS applications

Abstract An ionospheric broadcast model in a low-cost single-frequency GNSS receiver provides ionospheric corrections to improve its positional accuracy. This study evaluates the performance of the Klobuchar, NeQuick-G models, and proposes an optimized ensemble ionospheric model during the minimum (2020) and maximum (2024) phases of Solar Cycle 25 over the Indian low-latitude and equatorial regions. Dual-frequency Total Electron Content (TEC) data from International GNSS Service (IGS) stations over the Indian region, including JMSM, LCK4, JDPR, SHLG, and IISC stations data were analysed to quantify the proposed ionospheric TEC model accuracy under both quiet and disturbed geomagnetic conditions. The year 2024 exhibited strong solar activity, with Dst values reaching −412 nT and Kp index up to 9 during the May 10-11 geomagnetic storm. Model errors were computed as the difference between observed and predicted Vertical TEC (VTEC) values. The proposed optimized ensemble model, based on performance-weighted integration of Klobuchar and NeQuick-G outputs, consistently achieved the lowest RMSE values, ranging from 1.66 TECU (winter 2020) to 6.98 TECU (May 2024 storm), compared with the NeQuick-G (up to 30.07 TECU) and Klobuchar (up to 40.02 TECU). The analysis of proposed optimized ensemble ionospheric model revealed clear seasonal behaviour capture, with VTEC maxima during the equinoxes (50–100 TECU) and minima during solstices (20–40 TECU). Results demonstrate that the optimized model significantly enhances ionospheric prediction accuracy under both quiet and storm-time conditions, offering a robust solution for single-frequency GNSS applications in low-latitude regions strongly affected by ionospheric variability.