Methane in China - 2003 to 2010 from AIRS hyper spectral infrared observations

摘要

Spaceborne measurements by Atmospheric Infrared Sounder (AIRS) on the EOS/Aqua satellite provide a global view of methane (CH4) distribution in the mid-upper troposphere (MUT-CH4) since September, 2002. In the AIRS retrieval algorithm, a modified Optimal Estimation Model (OEM) is utilized based on Singular Value Decomposition (SVD) technique, which minimizes the dependence of the solution to the first-guess field and the first-guess error characteristics and can get an accurate CH4 retrieval at the maximum sensitive layer in the mid-upper troposphere. Nevertheless, in the less sensitive layer, the CH4 retrieval is close to the first-guess. Therefore, a better first-guess can improve the accuracy and efficiency of CH4 vertical profile retrieval. In this paper, a stable inversion model based on Empirical Orthogonal Functions (EOF) of computational efficiency and acceptable accuracy was developed to solve the above problems. Based on RTTOV 10.2 (a high precision atmospheric radiative transfer model for AIRS), an empirical regression model based on orthogonal function was formulized to retrieve CH4 vertical profile from spacebonre hyperspectral infrared observations. Validation of AIRS CH4 total column amount retrieved from EOF versus FTIR measurements shows that its Person correlative coefficient is bigger than 0.85. The CH4 volume mixing ratio (VMR) near surface retrieved from EOF versus Mt. Waliguan, China (WLG, 36.2879°N 100.8964°E, 3810 m) shows that the RMS error is less than 1.5%. The regression model based on EOF can retrieve CH4 profile from spaceborne hyperspectral infrared observations in a rapid and stable way.