A drought index for soybeans; a contribution to a soybean yield prediction model for Rio Grande do Sul, Brazil

Abstract

The main objective of this paper was the establishment of an index for the assessment of quantitative yield reductions in the soybean (Glycine max (L.) Merrill) crop. Daily and monthly rainfall data were utilized to calculate eleven different drought indices in various calendar month periods troughout the life cycle of the soybean crop in the "Colonial Missões" microregion. The results indicated that the best drought index is the summation of the daily values of 1 - ER/EP in the trimester: December, January and February which showed a 0.91 coefficient of partial correlation to the yield. As a result of the establishment of this index, a weather-yield technology model was developed aiming the soybean yield forecast with acceptable precision and accuracy.