A New Hybrid Processing Strategy to Improve Temporal Gravity Field Solution

The way to reach the prelaunch baseline accuracy is still one of crucial issues for Gravity Recovery and Climate Experiment (GRACE) temporal gravity field model determination. In this work, two hybrid processing strategies are developed to improve the accuracy of GRACE temporal solution. One is for kinematic empirical parameters, the other is for arc-specific parameters. These two hybrid processing strategies are compared with two solely processing strategies including filter predetermined strategy (FPS) and pure predetermined strategy (PPS). Assessing with a closed-loop simulation study, the results demonstrate the outperformance of our hybrid processing strategies. Using GRACE Level 1B data, the GRACE monthly gravity field models HUST-FPS-PPS and HUST-Grace2019 are, respectively, determined with the hybrid processing strategies, and they are compared with the solely solutions HUST-FPS and HUST-PPS. The comparison results demonstrate that (1) as for the kinematic empirical parameter, the hybrid processing strategy can retain the temporal signal but also reduce the noise level. For instance, the annual amplitudes over Amazon River Basin are 16.0 cm for HUST-PPS, 19.1 cm for HUST-FPS, and 19.4 cm for HUST-FPS-PPS. In contrast, the cumulative geoid difference of HUST-FPS-PPS reduces by 11% and 26%, when compared to HUST-PPS and HUST-FPS. (2) As for the arc-specific parameters, the hybrid processing strategy also significantly improves the recovered monthly solution. Compare with Center for Space Research RL05 and HUST-FPS-PPS, the cumulative geoid difference of HUST-Grace2019 reduces by 19% and 12%. The comparison results in spatial domain also support our conclusions. The hybrid processing strategy of this work is likely applicable for GRACE Follow-On data.