The importance of weather data in crop growth simulation models and assessment of climatic change effects.

Yields of agricultural crops are largely determined by the weather conditions during the growing season. Weather data are therefore important input variables for crop growth simulation models. In practice, these data are accepted at their face value. This is not realistic. Like all measured values, are weather data subject to inaccuracies. Crop growth simulation models are sensitive to weather data used as input, so inaccuracies in weather data can affect the simulation results. The errors in weather data were estimated and their effects on the simulation results of a spring wheat crop growth simulation model were determined. Inaccuracies in weather data caused deviations in simulated yields of 10-15 %.In most weather data sets missing values occur and since crop growth models require daily data the values of the missing data have to be estimated. Several methods to estimate missing values were discussed and their effects on simulation results were studied. Large differences in quality of the estimation methods were found. Some of them resulted in deviations in simulated yields up to 30 %.Daily weather data are not always available and often average weather data are used instead. The effects of using average weather data on simulation results were studied for three sites in different climates. For all sites large deviations in simulation results were found.The increasing CO ₂ concentration is affecting agricultural production in two ways: via a climatic change and via effects on assimilation and transpiration rates. The spring wheat model was used to study the overall effects of higher CO ₂ levels on wheat yields in Western Europe. A temperature rise of 3 °C resulted in a yield decline, doubled CO ₂ concentration in a yield increase and the combination of both in a yield increase of about 2 ton ha <sup>-1