有效波高数据融合可以弥补单颗卫星高度计数据存在的时间和空间分辨率不足的问题,为有效波高的海洋学应用提供更精确的分析资料。对反距离加权法、克里金插值法和逐步订正法等数据融合方法进行了研究,得到了适合于有效波高数据融合的模型和参数,并利用GFO,Jason-1和Envisat高度计获取的我国海域及其邻近海域有效波高数据,对不同融合方法、滤波器窗口和卫星个数等影响融合结果的因素进行了分析,结果表明:(1)融合后的有效波高分布特征与前人多年统计分析结果基本一致;(2)数据较密集时,不同融合方法的结果差别不大;(3)选择的滤波器窗口大小与时间分辨率有关,对于时间分辨率为10 d的融合而言,采用2.5°×2.5°-3°×3°大小的滤波器窗口所得到的融合结果最合适;(4)参加融合的卫星个数至少为3颗。
Significant wave height (SWH) data from multisatellite altimeters are merged to overcome the problem of insufficient temporal and spatial resolutions of single satellite altimeter data. The fusion data are more accurate for various marine applications. Data fusion methods, such as inverse distance to a power, Kriging and successive correction, are studied in order to find the suitable models and parameters for fusion of SWH data. Using the SWH data from GFO, Jason- 1 and Envisat in China's seas and their adjacent seas, some factors which influence the fusion result are analyzed. These factors include different fusion methods, size of filter window and number of satellites. It is shown that (1) after the effective fusion, the distribution of SWH is almost the same as previous statistical analysis; (2) there is little difference among the results by different methods in case the distribution of altimeter data is sparse; (3) the filter window needs to be approximately 2.5°× 2.5°-3°× 3° for reasonable results when the temporal resolution is 10 d; (4) at least three satellites are needed.