Estimating tropospheric delay with GNSS observations to provide a model product

Abstract Traditional global navigation satellite system (GNSS) tropospheric delay modeling involves multi-step processing, including deriving tropospheric delay values through precise point positioning (PPP) and fitting model coefficients. This study proposes a novel approach for direct estimation of tropospheric delay model coefficients with GNSS observations to provide a model product. Utilizing a tropospheric delay model integrated with global or local GNSS observations, this method simplifies the process and provides a tropospheric delay model product suitable for GNSS positioning applications. Data from 105 International GNSS Service (IGS) stations were used to directly estimate the tropospheric delay model coefficients, with 24 IGS stations used for performance validation. The proposed method achieves a 5.41% reduction in mean RMS error compared to the current method, which involves multi-step processing, including deriving tropospheric delay values through PPP and fitting model coefficients. Large errors observed in North America and parts of Asia with the current method are reduced with the proposed method, enhancing overall model reliability. Additionally, PPP accuracy validation demonstrates performance improvements with the proposed method in the North, East, and Up directions for both 30 min and 60 min positioning, compared to the current method. Enhancements are especially pronounced in regions such as Asia and Antarctica, ensuring greater stability and precision over various durations. Overall, the proposed method demonstrated better performance compared to the current method, providing a more reliable and efficient solution for precise GNSS positioning. Notably, the proposed method directly estimates the tropospheric delay model coefficients for a model product, simplifying the process.