In crop modeling weather is used as an input. The available data ranges from one second to one month at different sites where crop-modeling work in the world is going on. Different curve fitting techniques, interpolation, extrapolation functions etc., are being followed to use weather data in the model operation. Agrometeorological variables are especially subject to variations in space. It is reported that, as of now, anything beyond daily data proved unworthy as they are either over-estimating or under-estimating the yield in simulation. Stochastic weather models can be used as random number generators whose input resembles the weather data to which they have been fit. These models are convenient and computationally fast, and are useful in a number of applications where the observed climate record is inadequate with respect to length, completeness, or spatial coverage. These applications include simulation of crop growth, development and impacts of climate change. In 1995 JW Jones and Thornton described a procedure to link a third-order Markov Rainfall model to interpolated monthly mean climate surfaces. The constructed surfaces were used to generate daily weather data (rainfall and solar radiation). These are being used for purposes of system characterization and to drive a wide variety of crop and live stock production and ecosystem models. The present generation of crop simulation models particularly DSSAT suit of models have proved their superiority over analytical, statistical, empirical, combination of two or all etc., models so far available. In the earliest crop simulation models only photosynthesis and carbon balance were simulated. Other processes such as vegetative and reproductive development, plant water balance, micronutrients, pest and disease, etc., are not accounted for as the statistical models use correlative approach and make large area yield prediction and only final yield data are correlated with the regional mean weather variables. This approach has slowly been replaced by the present simulation models by these DSSAT models. When many inputs are added in future the models become more complex. The modelers who attempt to obtain input parameters required to add these inputs look at weather as their primary concern. They may have to adjust to the situation where they develop capsules with the scale level at which the input data on weather are available.
Was this article helpful?