Mean Sea Surface Model Creation And Application Research Over China Sea And Its Vicinity Area Derived From Multi-satellites Altimeter

Mean sea surface is a stable sea level over a certain period, which contains theundulation of the geoid and the mean dynamic topography information relative to thereference ellipsoid. As a result, it has been widely applied in the research of physicaloceanography, geophysics and other areas. Relied on the advantages of large scale,all-time and all-weather, satellite altimetry provides a new reliable means for theestablishment of mea sea surface model.Based on the combined use of multi-satellite altimeter data in a time sequencemore than two decades which included Geosat, ERS-1, ERS-2, T/P, Jason-1, Envisat,and Jason-2data, these altimeter data are analyzed and validated cycle by cycle toeliminate the exceptional data cycle after ionosphere correction smoothing, dataediting and geophysical correcting. Then the effects of time variation and radial orbiterror in the altimetry data are weakened by the collinear process and crossoveradjustment. After the unification of reference ellipsoid and reference frame andmulti-satellites crossover adjustment, a2’×2’ resolution mean sea surface model overChina sea and its adjacent sea areas(0°N-45°N,100°E-140°E) is obtained by Shepardmethod.For numerical tests, the presented mean sea surface model is compared with thegrid point values of MSS_CNES_CLS10model with the root-mean-square value of9.06cm, while the results of comparison with the along track values ofMSS_CNES_CLS01model in the Jason-1/2GDR data are in the range of7to8cm. This proves that the new mean sea surface model is correct and reliable. Finally, theresearch of the mesoscale eddy detection and the kuroshio detection of flow velocityand flow direction is carried out on the basis of the new mean sea surface model.In the data process of establish the new model, this paper uses almost all thealtimeter data available and improves in three aspects compared to the traditionalmethods. Firstly, all the selected data is validated by the cross points difference cycleby cycle, and the anomaly cycle is picked out and eliminated. Secondly theionosphere correction in the GDR data is smoothed in order to weaken the apparatuseffects. Thirdly, in order to weaken the ocean time variation effects of geodetic data,the T/P original data in the corresponding time range is used to compute the seasurface height anomaly instead of the existed sea surface height anomaly model. Inthis way, the priori information of the mean sea surface included in the exited modelcan be avoided.