Improving arable land heterogeneity information in available land cover products for land surface modelling using MERIS NDVI data

Abstract. Regionalization of physical land surface models requires the supply of detailed land cover information. Numerous global and regional land cover maps already exist but generally, they do not resolve arable land into different crop types. However, arable land comprises a huge variety of different crops with characteristic phenological behaviour, demonstrated in this paper with Leaf Area Index (LAI) measurements exemplarily for maize and winter wheat. This affects the mass and energy fluxes on the land surface and thus its hydrology. The objective of this study is the generation of a land cover map for central Europe based on CORINE Land Cover (CLC) 2000, merged with CORINE Switzerland, but distinguishing different crop types. Accordingly, an approach was developed, subdividing the land cover class arable land into the regionally most relevant subclasses for central Europe using multiseasonal MERIS Normalized Difference Vegetation Index (NDVI) data. The satellite data were used for the separation of spring and summer crops due to their different phenological behaviour. Subsequently, the generated phenological classes were subdivided following statistical data from EUROSTAT. This database was analysed concerning the acreage of different crop types. The impact of the improved land use/cover map on evapotranspiration was modelled exemplarily for the Upper Danube catchment with the hydrological model PROMET. Simulations based on the newly developed land cover approach showed a more detailed evapotranspiration pattern compared to model results using the traditional CLC map, which is ignorant of most arable subdivisions. Due to the improved temporal behaviour and spatial allocation of evapotranspiration processes in the new land cover approach, the simulated water balance more closely matches the measured gauge.