基于奇异谱分析的地基GPS可降水汽含量变化规律分析OACSTPCD

We used GAMIT/GLOBK 10.7 to process the 10-year data of 20 GPS stations in Japan from 2010 to 2019,and evaluated their accura-cy in this paper.Firstly,we used singular spectrum analysis(SSA)and fast Fourier transform(FFT)to detect the periodic term and trend term of precipitable water vapor(PWV),and proposed a PWV change rate estimation method by combining SSA and least square fitting,which could ef-fectively avoid the influence of periodic term and noise on PWV change rate.Then,we deeply analyzed and studied the temporal variation char-acteristics and spatial distribution characteristics of PWV in Japan.The results show that the accuracy of time series is 3.35 mm by GAMIT/GLOBK.The PWV in Japan shows an obvious upward trend from 2010 to 2019,with an upward rate of 0.13 mm/a,accompanied by obvious pe-riodic changes.Spatially,PWV in Japan is high in the coastal area,low in the inland area,high in the north and south,and low in the central area.The research results can provide some basic references for climate change research.