an ESA DUE initiative


International LST and Emissivity Group

The International Land Surface Temperature and Emissivity Working Group (ILSTE-WG) aims to provide advice and recommendations to the wider scientific and user communities on the best practices for retrieval, validation and exploitation of Land Surface Temperature (LST), Ice Surface Temperature (IST), Lake Surface Water Temperature (LSWT), and Land Surface Emissivity (LSE).

Authors: S. Ermida, C. Jimenez, C. Prigent, I.F. Trigo and C. daCamara


A comparison of land surface temperature (Ts) derived from the Advanced Microwave Scanning Radiometer - Earth observation system (AMSR-E) with infrared Ts is presented. The infrared Tsinclude clear-sky estimates from the MODerate resolution Imaging Spectroradiometer (MODIS), the Spinning Enhanced Visible and Infrared Imager (SEVIRI), the Geostationary Operational Environmental Satellite (GOES) Imager, and the Japanese Meteorological Imager (JAMI). The higher discrepancies between AMSR-E and MODIS are observed over deserts and snow covered areas. The former seems to be associated with Ts underestimation by MODIS, whereas the latter are mostly related to uncertainties in microwave emissivity over snow/ice. Ts differences between AMSR-E and MODIS are significantly reduced after masking out snow and deserts, with a bias change from 2.6/4.6 K to 3.0/1.4 K for day/nighttime, and a standard deviation (STD) decrease from 7.3/7.9 K to 5.1/3.9 K. When comparing with all infrared sensors, the STD of the differences between microwave and infrared Ts is generally higher than between IR retrievals. However, the biases between microwave and infrared Ts are, in some cases, of the same order as the ones observed between infrared products. This is the case for GOES, with daytime biases with respect to AMSR-E and MODIS of 0.45 K and 0.60 K, respectively. While the infrared Ts are clear-sky estimates, AMSR-E also provides Ts under cloudy conditions. For frequently cloudy regions, this results in a large increase of available Ts estimates (>250%), making the microwave Ts a very powerful complement of the infrared estimates.


Spatial distribution of the bias (K) of surface temperature differences between three geostationary sensors (GOES Imager, MSG SEVIRI, and MTSAT-2 JAMI) and MODIS, and between the same geostationary sensors and AMSR-E. The bias is calculated separately for the daytime and nighttime observations of the January-September 2011 period.


Link to the paper

Ermida S., C. Jimenez, C. Prigent, I.F. Trigo, and C. daCamara (2017), Inversion of AMSR-E observations for land surface temperature estimation - Part 2: Global comparison with infrared satellite temperature, J. Geophys. Res., 122, doi:10.1002/2016JD026148.