Improving the process-based simulation of growth heterogeneities in agricultural stands through assimilation of earth observation data

Complex process-based land surface models require detailed input parameters. While the process-describing parameters mostly are well known from laboratory research, the spatial parameters, such as terrain, land use or soil maps, often are available at coarse spatial resolutions only. The mass and energy balance of the land surface is strongly determined by plant growth. Depending on the application of the model, especially the spatial distribution of growth influencing site characteristics, such as soil properties for example, therefore is of major importance. This study demonstrates, how Earth Observation data can be used to overcome the lack of spatial detail, applying the land surface model PROMET to a precision farming task, i.e. site specific cereal yield modelling on a farm in northeast Germany. Maps of photosynthetically active leaf area, generated from RapidEye and Landsat TM data, were assimilated and the model output was successfully validated against measured yield maps for the summer of 2010.