Sensitivity analysis of an XCO2 retrieval algorithm for high-resolution short-wave infrared spectra

摘要

A sensitivity analysis, based on the Greenhouse-gases Monitoring Instrument (GMI) on Chinese Gaofen 5 satellite, is described considering four input parameters that affect the retrievals of the dry air column-averaged mole fraction of CO2 (XCO2). The analysis is based on synthetic spectra (centered around 0.765 and 1.575 mu m) produced by SCIATRAN. The input parameters considered include the profiles of temperature and pressure, as well as the surface albedo and aerosols. The main findings are temperature errors of 1 K and pressure errors of 2 hPa lead to XCO2 retrieval errors of less than 0.25%. The uncertainties in albedo (ranging from 0.03 to 1.0) or surface types lead to less than 2.0% retrieval error. Considering three aerosol types (urban, rural and maritime), we investigated the retrieval errors in XCO2 that were induced by incorrect assumptions on the aerosol optical thickness (AOT). For AOT values ranging from 0.02 to 1.0, the retrieval errors are less than 6.0% if aerosols are completely neglected in the retrieval. The presented results provide a basis for rational utilization of the a priori database and are conducive to increasing the inversion accuracy.