EVALUATION OF THE PERFORMANCE OF MERIS SPECTROMETER DATAIN SNOW COVER MONITORING IN THE BOREAL FOREST BELT

摘要

This paper describes a work plan for the evaluation of the capabilities and performance ofrnMedium Resolution Imaging Spectrometer (MERIS) data for monitoring snow covered landrnsurfaces, specifically in the boreal forest belt. The performance of MERIS over snow in thernnorthern latitudes is mainly estimated by comparison of time series derived from MERIS surfacernreflectance against values derived from Analytical Spectral Devices (ASD) Field Spec Pro JRrndata acquired over validation sites in Finland during the 2005 melting season. The purpose of thisrninvestigation is not to develop a snow-mapping algorithm but rather to identify generallyrnapplicable standard values for seasonally snow covered land cover reflectance and variability forrnimprovement of existing algorithms. Accordingly, the goal is to evaluate and compare MERISrnwith airborne and ground-based validation spectra over typical boreal land cover types withrnvarious snow and weather conditions. Ground-based high accuracy hyperspectral data arernvaluable tools in detecting and specifying land cover spectral features and helping to identifyrnthem from coarser resolution satellite observations. The resulting estimates for temporal andrnspatial variability in spectra can be utilized to reduce errors in snow mapping algorithms in thernNorthern Hemisphere. The main potential for snow monitoring and limitations for the estimationrnof MERIS-derived standard variability for seasonally snow-covered surfaces are discussed inrnconclusion.