Agroclimatc multiple linear regression model for upland crop productivity forecast

Abstract

Linear Multiple-Regression was used to determine a relationship between monthly rainfall X_j and annual crop yield (Y) of wheat, barley and potatoes. The mathematical expression used in the models were expressed as: Y = a + b1X1 + b2X2 + b3X3 + b4X4 where the parameters a, b1, b2, b3 and b4 were found through multiple linear regression. The models allow yield forecast three months before harvest. The wheat models are the best and the barley's are better than those of the potatoes.