基于FY-3C微波辐射成像仪海表温度产品的偏差订正方法研究  

A bias correction method for sea surface temperature based on FY-3C MWRI data

作　　者：廖志宏 徐宾[1] 张雷[1] 师春香[1] 周自江[1] LIAO Zhihong;XU Bin;ZHANG Lei;SHI Chunxiang;ZHOU Zijiang(National Meteorological Information Center,Beijing 100081,China)

机构地区：[1]国家气象信息中心,北京100081

出　　处：《气象学报》2020年第5期840-852,共13页Acta Meteorologica Sinica

基　　金：国家自然科学基金项目(41806213);国家重点研发计划项目(2018YFC1506604)。

摘　　要：提出一种针对FY-3C搭载的微波辐射成像仪(MWRI)海表温度产品的分段回归偏差订正方法,该方法通过引进气候态海表温度数据,建立与关联实测海表温度相匹配的回归模型,并通过对模型中关联变量的误差分析,选择最优样本进行分段回归,以实现对海表温度数据的重新估计。通过对MWRI海表温度数据的偏差订正试验表明,采用分段回归方法获得的订正结果无论在误差指标的空间分布还是时间序列上,都要明显优于采用传统概率密度函数偏差订正方法的结果。其中,采用概率密度函数方法订正后的海表温度产品误差标准差和均方根误差从订正前的0.9—1.0℃,减小到0.8℃左右,而采用分段回归方法获得相应的订正误差仅为0.6℃左右,订正效果有明显改善。A Piece-Wise Regression(PWR) method is proposed for bias correction of sea surface temperature(SST) from the Fengyun-3 C(FY-3 C) MicroWave Radiation Imager(MWRI) products. In this method, a regression model is developed to match the associated in-situ SST with daily climatological SST, and the optimal matchups are selected through the error analysis of the associated variables in the model. SSTs are then recalculated by using these optimal matchups in the Piece-Wise Regression model.Compared with the traditional probability density function(PDF) matching technique for bias correction, the PWR method can better remove biases in the spatial-temporal domain, and the standard deviations(SDs) and RMSEs are decreased from 0.9—1.0℃ to 0.6℃.This result is much better than that from the PDF method, which reduces the SDs and RMSEs from 0.9—1.0℃ to about 0.8℃.

关 键 词：偏差订正 海表温度 FY-3C 微波辐射成像仪 分段回归 

分 类 号：P716.12[天文地球—海洋科学]