Infrared instrumentation on geostationary satellites is now rapidly approaching the spectral quality and accuracy of modern sensors flying on polar platforms. Currently, the core of EUMETSAT geostationary meteorological programme is the Meteosat Second Generation (MSG). However, EUMETSAT is preparing for the Meteosat Third Generation (MTG). The capability of geostationary satellites to resolve the diurnal cycle and hence to provide time-resolved sequences or times series of observations is a source of information which could suitably constrain the derivation of geophysical parameters.
Nowadays, also because of lack of time continuity, when dealing with observations from polar platforms, the problem of deriving geophysical parameters is normally solved by considering each single observation as independent of past and future events. For historical reason, the same approach is currently pursued with geostationary observations, which are still being dealt with as they were with polar observations.
In this study we show some preliminary results on emissivity and surface temperature retrieval for SEVIRI observations, using the Kalman filter methodology (KF) and compare the retrievals with those obtained using IASI observations co-localized with SEVIRI ones using the times accumulation approach (Optimal Estimation OE). The Sahara desert was chosen as target area, and both SEVIRI and IASI data (infrared radiances and cloud mask) were acquired. The time period considered is that of July 2010 (the whole month). ECMWF analyses for the same date and target area have also been acquired, which comprise Ts, T(p), O(p), Q(p) for the canonical hours 0:00, 6:00, 12:00 and 18:00. Moreover, for the purpose of developing a suitable background for emissivity, the Global Infrared Land Surface Emissivity database developed at CIMSS, University of Wisconsin, derived by MODIS observations was used and was available from the year 2003 till 2011.
Concerning the performance of the two methodologies, the retrieval of skin temperature is almost equivalent. The agreement between OE and KF is fairly good if compared with ECMWF analysis for sea surface, while for land surface, OE and KF agree fairly well with ECMWF during the night, but at midday ECMWF shows a cold bias of 10 K and more. For emissivity the comparison with the UW/BFEMIS database for the same date and location is fairly good for both methods.
EARSeL European Association of Remote Sensing Laboratories, Strasbourg, France