The gravity recovery and climate experiment(GRACE) mission and followed missions can provide unconstrained monthly time-varying gravity fields that are polluted by north-south striping noise. This noise is defined as a quasi-periodic oscillation along longitude, propagating continuously along latitude. Aiming at this characteristic, we introduce the moving average smoothing technique to suppress the noise in spatial domain, implementing different degrees of smoothing based on the different prior knowledge between ocean and land. Then, the proposed method is compared with common DDK filters in spatial, frequency, and time domains, showing a comparable filtering effect. The evaluation results show that the noise level of proposed method in ocean is about 20 mm, which is lower than that of DDK5-DDK7(26.20-50.25 mm). Although there are limitations such as the setting of sliding window size depends on experience, the proposed method does not need to know the variance information between signal and noise in advance and can provide a new idea for the design of novel filters.

GRACE  /  striping noise  /  moving average  /  spatial filter