University of Tehran Press Iranian Journal of Soil and Water Research 2008-479X 51 8 2020 10 22 Development of Monthly Ensemble Precipitation Forecasting System in Sefidrud Basin, IRAN Development of Monthly Ensemble Precipitation Forecasting System in Sefidrud Basin, IRAN 1881 1893 74188 10.22059/ijswr.2019.284942.668251 FA Hossein Dehban Irrigation and Reclamation Engineering Department, Faculty of Agricultural Engineering & Technology, College of Agriculture & Natural Resources, University of Tehran, Karaj, Iran Kumars Ebrahimi Irrigation and Reclamation Engineering Department, Faculty of Agricultural Engineering & Technology, College of Agriculture & Natural Resources, University of Tehran, Karaj, Iran 0000-0002-9914-4383 Shahab Araghinejad Irrigation and Reclamation Engineering Department, Faculty of Agricultural Engineering & Technology, College of Agriculture & Natural Resources, University of Tehran, Karaj, Iran Javad Bazrafshan Irrigation and Reclamation Engineering Department, Faculty of Agricultural Engineering & Technology, College of Agriculture & Natural Resources, University of Tehran, Karaj, Iran 0000-0002-6721-8990 Journal Article 2019 07 20 Precipitation forecasting is one of the most important tools in water resources planning and management. Recently, new methods called atmospheric dynamic models have been used to predict many hydro-climate variables including precipitation. Before using the predictions of these models in planning and decision making, the accuracy of the mentioned predictions and their bias correction should be evaluated. Therefore the objective of this study is to ascertain the biases and to combine the results of precipitation forecasting with a set of global dynamic forecasting models. To achieve this aim, firstly the precipitation forecast results of each model were compared separately with the regional recorded precipitation data in the period of 1982 to 2017. Using this approach, the systematic errors were removed and corrected, i.e. using the quantile mapping method. This work was done for different forecast periods and also for different months. Furthermore, based on the accuracy of each model, a hybrid/multi-model prediction system was developed using Bayesian averaging method (BMA). The results showed that after the bias correction, using the quantitative mapping, at least one model among 78 prediction models have a relatively high correlation value of about 0.7. This result was recorded for the next one-month horizon. This correlation was increased to more than 0.8, by combining 78 predictive members, using Bayesian averaging method. Therefore, the accuracy of the predicted precipitation increases significantly using bias correction in tandem with combining the prediction models. Precipitation forecasting is one of the most important tools in water resources planning and management. Recently, new methods called atmospheric dynamic models have been used to predict many hydro-climate variables including precipitation. Before using the predictions of these models in planning and decision making, the accuracy of the mentioned predictions and their bias correction should be evaluated. Therefore the objective of this study is to ascertain the biases and to combine the results of precipitation forecasting with a set of global dynamic forecasting models. To achieve this aim, firstly the precipitation forecast results of each model were compared separately with the regional recorded precipitation data in the period of 1982 to 2017. Using this approach, the systematic errors were removed and corrected, i.e. using the quantile mapping method. This work was done for different forecast periods and also for different months. Furthermore, based on the accuracy of each model, a hybrid/multi-model prediction system was developed using Bayesian averaging method (BMA). The results showed that after the bias correction, using the quantitative mapping, at least one model among 78 prediction models have a relatively high correlation value of about 0.7. This result was recorded for the next one-month horizon. This correlation was increased to more than 0.8, by combining 78 predictive members, using Bayesian averaging method. Therefore, the accuracy of the predicted precipitation increases significantly using bias correction in tandem with combining the prediction models. https://ijswr.ut.ac.ir/article_74188_ffa95d1d93019c10ff2ff05e97f2d73d.pdf