Abstract
This paper aimed to explore the possibility of the use of logistic regression in the prediction of drought in Sudan. In this paper Regression models were used to predict drought in 5 states in Sudan. Results showed that the average prediction error for above-average and below-average years is reported for each station. Area-specific statistics are listed in Table1. Correlations between predictions and observations range from r = −0.31 to r = −0.68, with the model performing slightly better in predicting the number of wet days in drier years. In general, however, the simple linear model displays an average absolute error ranging between 9 and 22 days.
Keywords: Regression, Drought.

تحميل البحث