A Model for Prediction of Annual Precipitation in Kerman Province

Prediction of annual precipitation usually guarantees success in dry-farming and promises a better pasture management. In this research, the long-term observed daily precipitation for 28 different stations in Kerman province was analyzed. Starting from the first day of autumn, a good linear relationship is obtained between the days with 47.5 mm precipitation (t47.5, day), and the amount of annual precipitation (pa, mm). Furthermore, to increase the correlation coefficient of this relation-ship, long-term mean annual precipitation (Pma, mm), elevation, longitude, and latitude for each station were also used in the multiple regression analysis. The results showed that none of these factors (i.e., elevation, longitude and latitude) could improve the correlation coefficient. The independent variables were Pma and t47.. As another independent variable, temper-ature of the water in Persian-Gulf (south of region) in the water-air frontage were added. The three–month cumulative surface temperatures in autumn (Tau, ºC), improved the correlation coefficient of the multiple regression. These final equations were used to predict the annual rainfall in Baft-soltanee station. The results showed that the both simple models could predict wet and dry years sufficiently well but the error of estimation in the second model was lower.