A simplified daily soil moisture budget model (SDSMBM), which considers the major soil and plant processes involving water, and that uses estimates of Penman's potential evapotranspiration for Brazilian geographical factors (latitude, altitude and minimum distance from the Atlantic Ocean), is described, using a crop-stage water use factor to convert to maximum crop evapotranspiration. Other necessary input data to the model are daily rainfall, the month of normal phenological crop events and values of maximum available soil water capacity, for the crop rotation and region considered.
If water stress occurs, evapotranspiration is reduced, with the degree of reduction depending on available soil moisture. A daily water stress index, based on the assumption that yield reduction is proportional to the reduction in evapotranspiration from the maximum, is calculated. The index is summed for one or more different calendar-month periods, according to the crop and region considered, and determined from statistical regression analysis involving regional crop yields and a technological change index.
The water stress index has been successfully applied to the southern region of Brazil, responsible for 88% of the national production of corn and 95% of the national production of soybeans, where the corn and soybean yields were related to weather and soil conditions.
A non-linear time trend, of corn and soybean yields in the south of Brazil can be observed, due to improved techniques for crop production such as high-yielding varieties, effective fertilizer, agricultural machines and effective chemicals, but the yearly deviation of crop yield is chiefly because of weather fluctuations.
The weather-technology models developed can be used to forecast crop yields 60 days before harvest time, as well as to develop yield probability projections. Severe droughts are observed to occur in 30% of the years with some droughts having a duration of more than 12 months and 2 or 3 crop seasons may be affected. The observation of summer drought probabilities in the southern region of Brazil, and its adverse effects on grain production, recommends the development of a food reserve program in Brazil, as well as research programs for several strategies and technologies designed to minimize drought effects.