Ionospheric Disturbance Index for GNSS Precise Positioning: Comparison between GIX and ROTI over China 

Abstract

The spatial gradient of Total Electron Content (TEC) demonstrates a notable correlation with occurrences of ionospheric scintillation and degradation in Global Navigation Satellite Systems (GNSS) performance. To assess the degree of the ionosphere perturbation over China, we utilize two ionospheric disturbance indices, i.e., the Gradient Ionospheric Index (GIX) and Rate of Total Electron Content Index (ROTI). Analyzing data from over 230 GNSS stations in China, we investigated the consistency and differences in ionospheric irregularities as indicated by GIX and ROTI, focusing particularly on the impact of horizontal gradients of vertical TEC (VTEC) on GNSS positioning. Experimental results indicate that the occurrence of ionospheric irregularities and the degradation of kinematic Precise Point Positioning (PPP) performance are more closely aligned with the strong horizontal VTEC gradients indicated by GIX, both in evolutionary characteristics and temporal variations, compared to ROTI. Notably, severe position errors and strong horizontal gradient of VTEC were consistently observed throughout the entire duration of St. Patrick's Day storm, including the prestorm period (06-16 UT on 16th March), the main phase (06-13 UT on 17th March), and the late recovery phase (05-14 UT on March 19th). Furthermore, significant PPP degradation (3-D PPP>1m) occurred not only preceded the intensification of ROTI but also in periods of stable ionospheric conditions indicated by ROTI values below 0.4 TECU/min, particularly on March 19th. The degradation primarily occurred in regions with significant irregularities, especially when the horizontal VTEC gradient surpassed a specific threshold (GIXx, P95+ > 55 mTECU/km, 1 mTECU/km = 10−3 TECU/km). These findings indicate the potential of GIX in effectively capturing the effects of ionospheric disturbances and enhancing the safety and accuracy of GNSS navigation and positioning.