Mozambique is located along the east coast of southern Africa, with economy primarily based on agriculture. Maize (Zea mays L.) is the most important crop, cultivated without irrigation, with yields dependent mostly on weather conditions. The objective of the study was to develop a spectral-agrometeorological model for maize yields forecast, in Manica province. The study area involved the districts of Gondola, Manica, Mossurize and Sussundenga, responsible for more than 80% of corn production in Manica province from 2000 to 2009. The meteorological variables ETr / ETm and ISNA were used in the model, after being generated from meteorological data in the ECMWF model. The spectral variables were EVI and NDVI, from product MOD13Q1, and LSWI was calculated from combinations of infrared and shortwave reflectance bands. The spectral agrometeorological model, adjusted by multiple linear regressions, had spectral and meteorological indices as independent variables, and maize yield (average and relative) as dependent variable. As results, the regional model which used yields from the districts of Gondola, Manica and Sussundenga, and the relative yield, was the most suitable for maize yields forecast in the region, with r² = 0.762 and RMSE of 9.46%.
MODIS; multiple linear regressions; NDVI; EVI; LSWI
