Principles of Crop Growth Modelling

Weather variables are derived from weather stations and are interpolated to the locations where the crop model is applied. Although the density of weather stations in many areas is often quit high, many stations do not report in near-real time making them unsuitable for near-real time crop monitoring applications. Due to the limited density of weather stations, a considerable uncertainty is often present in gridded weather products derived from weather stations. Weather variables can be provided by NWP models such as those are applied by the European Centre for Medium Range Weather Forecasts ECMWF and The National Centers for Environmental Prediction NCEP together with National Center for Atmospheric Research NOAA. NWP models long time suffered from poor spatial resolution as the grid resolution of the model is often in the order of a 0.5 to 2 degrees. A more subtle problem of NWP model output was inconsistency in the time-series due to incremental upgrades of NWP model itself. Therefore, any biases in the time-series caused by NWP model upgrades will distort the analysis of historic time-series of simulated and reported yields. This problem has been recognized by the NWP community and has resulted in the reanalysis project such as ECMWFERA-ARTEMIS and NCEP-NOAA reanalysis from European and USA. Based on the research by de Wit et.al 2010, NWP from ERA-INTERIM can be used to replace weather variables in the implementation of regional crop yield forecasting in Europe. Meteorological satellites MeteoSat such as the NOAA-AVHRR series or the MeteoSat series are capable of providing timely and reliable meteorological variable for crop yields forecast Roebeling et.al, 2004. MeteoSat particularly provides good opportunities with its 30 min. revisit interval and relatively compared to NWP models high spatial resolution. MeteoSat imagery can be used for deriving estimates of global and net radiation derived from cloud cover and albedo, daily minimum and maximum temperature derived from day and night surface temperature, and potential evapotranspiration derived from available radiation. Moreover, opportunities for rainfall estimates can be made available by integrating MeteoSat cloud cover estimates with rain gauge products.