小波变换融合双源GNSS-R数据提升海面高精度方法OA

To address the issues of insufficient accuracy and resolution in sea surface height retrieval from single-source GNSS-R data,this study leverages the high-spatiotemporal-resolution and dense observation capability of CYGNSS and the global coverage characteristics of FY-3E,employs a wavelet transform method to fuse the dual-source data,and carries out the construction of a high-precision sea surface height model for Chinese seas.The fusion results show that,compared with a single satellite,the fused result from CYGNSS(8 satellites)reduces the mean absolute error(MAE)from 3.80 m to 1.72 m,an improvement of 54.7%;after further fusion with FY-3E,the MAE drops to 1.51 m,achieving an overall improvement of 58.7%.Further analysis indicates that as the data duration increases from 1 month to 5 months,the specular point density rises from 0.12 points/km² to 0.40 points/km²,and the MAE decreases to 0.93 m,yielding accuracy superior to that of single-source retrieval methods.The GNSS-R multi-source data fusion method used in this study can effectively improve the model accuracy of single-source GNSS-R data,and the research results have the potential to support ocean dynamic environment monitoring and the construction of high-precision mean sea surface heights.