干旱区生态环境知识资源中心

Land surface temperature and emissivity spectra are essential variables for improving models of the earth surface atmosphere interaction or retrievals of atmospheric variables such as thermodynamic profiles, chemical composition, cloud and aerosol characteristics, and so on. In most cases, emissivity spectral variations are not correctly taken into account in climate models, leading to potentially significant errors in the estimation of surface energy fluxes and temperature. Satellite infrared observations offer the dual opportunity of accurately estimating these properties of land surfaces as well as allowing a global coverage in space and time. Here, high-spectral-resolution observations from the Infrared Atmospheric Sounder Interferometer (IASI) over the tropics (30 degrees N-30 degrees S), covering the period July 2007-March 2011, are interpreted in terms of 1 degrees X 1 degrees monthly mean surface skin temperature and emissivity spectra from 3.7 to 14 mu m at a resolution of 0.05 mu m. The standard deviation estimated for the surface temperature is about 1.3 K. For the surface emissivity, it varies from about 1%-1.5% for the 10.5-14- and 5.5-8-mu m windows to about 4% around 4 mu m. Results from comparisons with products such as Moderate Resolution Imaging Spectroradiometer (MODIS) low-resolution emissivity and surface temperature or ECMWF forecast data (temperature only) are presented and discussed. Comparisons with emissivity derived from the Airborne Research Interferometer Evaluation System (ARIES) radiances collected during an aircraft campaign over Oman and made at the scale of the IASI field of view offer valuable data for the validation of the IASI retrievals.