Brazil wheat yield covariance model

Abstract

A model based on multiple regression was developed to estimate wheat yields for the wheat growing states of Rio Grande do Sul, Parana, and Santa Catarina in Brazil. The meteorological data of these three states were pooled and the years 1972 to 1979 were used to develop the model since there was no technological trend in the yields during these years. Predictor variables were derived from monthly total precipitation, average monthly mean temperature, and average monthly maximum temperature.