基于AMSR2微波辐射计的近海面气温反演算法研究

摘要

基于多元线性回归方法,利用2013-01-06的AMSR2辐射计亮温数据和红外-微波融合SST数据产品,开展了近海面气温反演算法研究,并用TAO,RAMA和PIRATA等浮标实测数据对近海面气温的反演结果进行检验.近海面气温反演结果误差情况:均方根误差为0.66℃,偏差为0.02℃,相关系数R为0.91,该误差结果表明所建立近海面气温反演算法较好的反映在60°S~60°N纬度范围内的近海面气温分布情况;同时为进一步确定不同纬度近海面气温反演的误差分布,将近海面气温反演结果与ECMWF再分析数据进行了对比分析,结果表明,从赤道起算,纬度每升高或降低1°,反演均方根误差约增大0.1℃.%In this paper,based on multiple linear regression method and AMSR2 brightness temperature da-ta,the inversion algorithm for near sea surface air temperature is developed,and validated with TAO,RA-MA and PIRATA buoy measured data.The results show the root mean square error of 0.66 ℃,standard deviation of 0.02 ℃,and correlation coefficient of 0.91.The proposed algorithm reproduces the sea surface air temperature distribution of the low latitudes (60°S~60°N).Further comparison with ECMWF data in-dicates that when the latitude increases by 1°,the root mean square error of retrieval increases by about 0.1 ℃.