Introduction

Large area yield forecasting prior to harvest is of interest to government agencies, commodity firms and producers (Boote et al., 1996). Early information on yield and production volume may support these institutions in planning transport activities, marketing of agricultural products or planning food imports. Moreover, at world scale, agricultural market prices are affected by informat ion on the supply or consumption of foodstuffs (Marcus & Heitkemper, 1994). Market price adjustments or change in agricultural supplies in one area of the world often causes price adjustments in other areas far distant.

The Directorate General for Agriculture (DG AGRI) of the EU is responsible for implementation and control of the various EU policies on agriculture. To manage these policies DG AGRI requires detailed information on planted area, crop yield and production volume (De Winne, 1994). The main crops of interest are wheat, barley, oats, grain maize, rice, potato, sugar beet, pulses for human consumption, soybean, oilseed rape, sunflower, tobacco and cotton. Information on land use, land use changes and yields is routinely collected by various national statistical services that convey this information to the statistical office of the European Commission, EUROSTAT. Collection and compilation of these agricultural statistics is time consuming and laborious. In exceptional cases, these statistics are available some months after the season has ended, however, as a rule, it takes one or even two years before this information is available in the EUROSTAT databases. Consequently, at this stage these statistics are of limited use for the timely evaluation of the various policies. Hence, more timely and accurate information is needed.

To support DG AGRI in executing its tasks, in 1988 the Monitoring Agriculture with Remote Sensing (MARS) project was initiated with the objective to generate monthly information on land use, land use changes, exceptional growing conditions such as water stress and expected yields. This information had to be provided for various crops for all member states of the EU. To realise this objective, the MARS project used field surveys, high and low resolution satellite data and a crop growth simulation model.

In order to estimate the expected yields, a crop growth simulation model was combined with a detailed soil map, parameters for the various crops and spatial crop information to create the Crop Growth Simulation System (CGMS). CGMS uses daily meteorological observations to estimate crop status (i.e. water stress, biomass production, etc.) in the course of the growing season and crop yield at the end of the season.