基于快速局域线性回归的IRAS/FY-3B大气温湿廓线反演

A fast locally linear regression (FLLR) algorithm is proposed for retrieving the atmospheric temperature and humidity profiles from the infrared observational data of the Infrared Atmospheric Sounder (IRAS) carried on the Fengyun-3B satellite (FY - 3B). The observational samples required by the algorithm are the spatio - temporal matching data of the infrared observational values of IRAS/FY - 3B L1 data and the AIRX2RET V5 product. Taking 2011 as an example, the observational samples are obtained in the research area from 180°W to 180°E and 60°N to 60°S according to the conditions that the absolute difference of the observation time is less than 15 minutes and the absolute difference of the observation angle is less than 2°, and the samples are evaluated. Based on the matching observational samples, a comparative analysis is carried out...

提出一种快速的局域线性回归（Fast Locally Linear Regression，FLLR）算法，用于从搭载在风云三号B星（FY-3B）上的红外大气探测仪（IRAS）红外观测数据反演大气温湿廓线。算法所需的观测样本为IRAS／FY-3B L1数据红外观测值与AIRX2RET V5产品的时空匹配数据，以2011年为例，在180°W~180°E、60°N~60°S的研究区域内按照观测时间绝对差小于15 min和观测角度绝对差小于2°的条件获取观测样本，并对样本进行了评价。在匹配观测样本的基础上比较分析