Thermal Land-surface Variables From METEOSAT-IR Data

Abstract

There are two series of satellites that provide long and continuous series of data for the Mediterranean Basin: NOAA and METEOSAT. While NOAA/AVHRR has 5 window channels (1 visible, I near IR, I water vapour (WV; 3.7 mum), and 2 terrestrial IR (8-13 mum), METEOSAT is limited to 3 channels (I visible, 1 WV, and 1 terrestrial IR). On the other hand, METEOSAT resolves dynamic processes with 48 measurements per day while AVHRR performs 4 measurements per day for a given location. In the framework of climate analyses of surface properties the vegetation cover and the albedo must be derived from AVHRR (spectral capabilities), while the thermal properties must be derived from METEOSAT (temporal resolution). Therefore, only a combined evaluation of many years of AVHRR and METEOSAT can reveal climatic effects. The large amount of data and the high degree of automation that is required poses a technical challenge, while the development of adequate algorithms and their application are scientific challenges. In order to respond to these challenges best, the Forschungszentrum Karlsruhe and the FU-Berlin agreed to join forces and to share the tasks according to their respective foci of research. The FU-Berlin provides archived satellite data and derives the Normalized Difference Vegetation Index (NDVI) and albedo from AVHRR. The Forschungszentrum Karlsruhe - IMK determines thermal land surface properties from METEOSAT IR data and provides access to an automatic mass storage system. The determination of thermal surface parameters is designed for METEOSAT's temporal and spectral capabilities and consists of the following, components: The IR measurements are calibrated and satellite instruments are intercalibrated. Cloud covered pixels are detected using dynamic thresholds. A neural network is used to determine the atmospheric influence on IR measurements at satellite level. The network is about 5.000 times faster than MODTRAN and uses ECMWF atmospheric data as input. The processing includes the inversion of satellite brightness temperatures to thermodynamic land surface temperatures (Gottsche and Olesen, 2002). The spatial interpolation from a limited number of locations with known atmospheric profiles (ECMWF grid cells) to each pixel (Shepard algorithm; Schroedter et al., 2001). The temporal interpolation from 4 times per day, for which the atmospheric situation is known, to all 48 METEOSAT slots (Schadlich et al., 2001). A model consisting of a cosine (daytime) and an exponential decay (night-time) is fitted to 10 day or monthly composites of cloud free data in order to derive thermal surface parameters, e.g. minimum temperature, diurnal amplitude, and the time of the maximum temperature (Gottsche and Olesen, 2001). The merging of METEOSAT and AVHRR IR data to time series of 2 kin spatial and 30 minutes temporal resolution. All the above components were developed at the IMK. The two most recent components, the atmospheric corrections using neural networks and the model for the derivation of thermal surface parameters, are described in more detail below. In combination with work carried out by the FU-Berlin this will result in one of the most complete evaluations of long term satellite data.