A novel strategy of estimating the GNSS ray-tracing tropospheric delay based on the satellite clustering areas

The tropospheric delay is a natural error source that occurs when objects, including radio signals, pass through the lower atmosphere at altitudes below 70 km. It needs to be carefully adjusted in numerous scientific research endeavors, e.g., remote sensing satellite-based Earth observation, aircraft precise positioning, and active radar surveillance system. Global Navigation Satellite System (GNSS) has become a popular tropospheric delay generating technique thanks to its facilitating approach, high precision estimation, and low-cost receiving devices. However, the existing GNSS data processing methodology allows for the estimation of zenith tropospheric delay by considering the physical characteristics of the troposphere and employing a simplified functional model. Therefore, the accuracy of ray-tracing tropospheric delay would be compromised when converting the zenith delay to the slant direction using a mapping function. This manuscript proposes a novel GNSS data processing strategy which can directly estimate the ray-tracing tropospheric delay. The hemispherical observable space is divided into several regions based on different azimuth and elevation angles, and GNSS satellites that belong to the same region are considered to experience the identical tropospheric delay due to the tropospheric physical characteristics. In this case, the ray tracing tropospheric delay can be calculated, and the corresponding estimation precision can be guaranteed, as an individual delay may incorporate observables from multiple satellites. The estimation precision could be further improved once additional GNSS receivers are involved in the data processing. Furthermore, this manuscript analyzes the impact of multiple GNSS receivers on ray-tracing tropospheric delays, considering that the signal paths may not be strictly identical for the receivers. Among which, potential signal blockage impacts, horizontal distance impacts and altitude variation impacts between each receiver are regarded as the main factor in the precision evaluation. This manuscript also considers compensating the tropospheric delay for radar as a case study, and the theoretical precision in regard to GNSS receiver locations are presented. The simulation results reveal theoretical precision can achieve 2.4 mm; while the experimental results achieve an accuracy of 9.98 mm compared to a water vapor radiometer in the usage of 4 GNSS receivers, demonstrating the availability of the proposed tropospheric delay estimation method.