Seasonal forecasting of Bangladesh summer monsoon rainfall using simple multiple regression model

Abstract

In this paper, the development of a statistical forecasting method for summer monsoon rainfall over Bangladesh is described. Predictors for Bangladesh summer monsoon (June-September) rainfall were identified from the large scale ocean-atmospheric circulation variables (i.e., sea-surface temperature, surface air temperature and sea level pressure). The predictors exhibited a significant relationship with Bangladesh summer monsoon rainfall during the period 1961-2007. After carrying out a detailed analysis of various global climate datasets; three predictors were selected. The model performance was evaluated during the period 1977-2007. The model showed better performance in their hindcast seasonal monsoon rainfall over Bangladesh. The RMSE and Heidke skill score for 31 years was 8.13 and 0.37, respectively, and the correlation between the predicted and observed rainfall was 0.74. The BIAS of the forecasts (% of long period average, LPA) was -0.85 and Hit score was 58%. The experimental forecasts for the year 2008 summer monsoon rainfall based on the model were also found to be in good agreement with the observation. © Indian Academy of Sciences.