Suppressing stripe noise in GRACE/GFO level-2 products with dual low-pass filtering

The results obtained through GRACE/GRACE Follow-On spherical harmonic coefficient model exhibit noticeable north–south strip noise. While applying low-pass filtering along the latitude circle direction can effectively eliminate the north–south strip error, it unavoidably introduces east–west strip noise. This paper introduces a method that utilizes the cosine function to determine the cutoff frequency for the low-pass filtering in the latitude direction. Additionally, a low-pass filtering step is added in the meridian circle direction. This combined approach successfully mitigates the north–south strip error and simultaneously suppresses the east–west strip noise. The analysis results highlight the improvement in signal to noise ratio of global surface quality anomaly estimation achieved by utilizing the cosine function to determine the cutoff frequency. The use of data from five institutions reveals that the dual low-pass filtering (DLP) improves the signal to noise ratio by more than 2% compared to the traditional efficient low-pass-filtering (ELP). When analyzing the uncertainty in the Yangtze River Basin using the three-cornered hat method, it is observed that the uncertainties of the DLP solution, traditional ELP solution, and mascon solution are comparable. Furthermore, the DLP solution exhibits the smallest uncertainty, measured at 4.89 cm. In terms of root mean square error, the DLP almost consistently yields the lowest values across various regions, with the difference from the smallest root mean square error value remaining within 2 mm in certain regions.