关键词: 改进基线法, 短弧法, GRACE-FO, 时变重力场

Abstract:

To address the challenge of suppressing high-frequency errors in high-precision time-variable gravity field inversion, this paper investigates and develops the baseline method. An improved baseline inversion model is constructed by introducing a decorrelation method to optimize observation error processing and adopting a short-term time-variable parameter estimation strategy to enhance time-variable signal capture capability. The study focuses on comparing its performance with the traditional short-arc method in terms of spectral response and noise control. Based on 2021 GRACE-FO real data, monthly gravity field models up to degrees 60, 96, and 120 are derived. Results indicate that in the low-frequency band (below degree 30), both methods achieve comparable accuracy. However, in the mid-to-high frequency range (above degree 30), the improved baseline method significantly outperforms the short-arc method; notably, its cumulative geoid height error at degree 120 is reduced by approximately 8 cm, demonstrating superior capability in suppressing high-frequency noise. Furthermore, the model derived from the improved baseline method (referred to as the Baseline model) is validated against the CSR RL06.3 model released by Center for Space Research, University of Texas at Austin, the HUST-Grace2024 model released by Huazhong University of Science and Technology, and the Tongji-Grace2022 model released by Tongji University. The comparison reveals that the Baseline model achieves better overall accuracy above degree 60. The global equivalent water height derived from the Baseline model exhibits a more reasonable spatial distribution with less striping noise. Specifically, its RMSE over the ocean is significantly lower than that of the CSR RL06.3, HUST-Grace2024, and Tongji-Grace2022 models, and its signal-to-noise ratio is superior to these models, demonstrating optimal high-frequency noise suppression capability. The study concludes that the improved baseline method, through the combined enhancements of decorrelation processing and short-term time-variable parameterization, effectively improves the accuracy and stability of gravity field models in mid-to-high frequency bands, providing an effective approach for constructing high-resolution and high-reliability time-variable gravity field products.

Key words: improved baseline method, short-arc method, GRACE-FO, time-variable gravity field