SMAP卫星海表面亮温仿真及海表面盐度遥感反演

海表面盐度遥感是海洋遥感研究的重要内容之一。针对大气对海表面盐度遥感的影响,根据大气辐射传输理论,对大气影响进行仿真分析及修正,进而通过神经网络模型反演了海表面盐度。研究结果表明:大气对海表面盐度遥感存在一定的影响,需进行校正;当大气温度和压强精度分别达到2℃和10 hPa时,可以去除大气影响;训练样本集选取数量的不同将对神经网络反演精度造成一定的影响;SMAP卫星海表面盐度遥感反演结果相对误差较小,残差基本集中在0.6以内,但在盐度值低于34.4‰的区域误差较大。

SMAP Satellite Sea Surface Brightness Temperature Simulation and Sea Surface Salinity Retrieval

Sea surface salinity remote sensing is one of the important contents of the remote sensing research of the ocean. For the influence of the sea surface salinity remote sensing caused by the atmospheric, according to the theory of atmospheric radiation transfer, the atmospheric radiation effects were simulated and corrected, and then the sea surface salinity was inverted by the neural network model. The result showed that the atmospheric radiation effect was serious, and it needed to be corrected. When the precision of atmospheric temperature and pressure of the earth surface was 2 ℃ and 10 hPa, the atmospheric influence could be removed. The difference in the number of the training sample sets would have a certain impact on the accuracy of neural network inversion. The salinity retrieval relative error of the SMAP satellite was small, and the residual error was basically concentrated within 0.6, but the error was larger in the region where the salinity value is lower than 34.4‰.