%0 Journal Article %T Optimum configuration of RegCM4. 7 model in prediction of weekly cumulative precipitation during three extreme precipitation events of March-April 2019 %J Journal of Agricultural Meteorology %I Iranian Society of Irrigation and Water Engineering %Z 2345-3419 %A Babaeian, Iman %A Karimian, maryam %A Modirian, Raheleh %D 2021 %\ 11/22/2021 %V 9 %N 2 %P 48-60 %! Optimum configuration of RegCM4. 7 model in prediction of weekly cumulative precipitation during three extreme precipitation events of March-April 2019 %K RegCM4. 7 model %K CFSv. 2 model %K Heavy precipitation %K March 2019 %K Iran %R 10.22125/agmj.2021.287482.1115 %X From March 1 to April 4, 2019, three flash-flood events occurred across different areas of Iran. The purpose of this study is to select the optimum configuration of RegCM4. 7 regional climate model for weekly precipitation prediction several days ahead of these extreme events. For this purpose, the RegCM4. 7 climate model was run using the CFSv. 2 general circulation model boundary layer conditions and cumulus convection scheme. The results were compared with the ERA5 reanalysis using Taylor diagram and Kling Gupta (KGE) index. Accumulated precipitation of each of the three selected weeks was predicted in 6 different lead times. To select the best convection schemes, six of them namely. Tiedtke, Emanuel, Grell, Kian, Kou and MM5 shallow water were evaluated under the conditions of extreme events. In the first stage, the Tiedtke, Emanuel and Grell schemes showed better performance compared to other schemes. In the second stage, the effect of different boundary layer schemes was investigated. It was found that Tiedtke -Holtslag, Tiedtke -UW and Grell-UW are the most appropriate configuration of RegCM4. 7 model in predicting heavy rainfall in the study period, respectively. Although good KGE values was only obtained for one week ahead, but the correlation coefficients were relatively good for hindcast with lead time of 2-4 weeks. Despite of low KGE values for the weeks beyond second week, acceptable correlations coefficients confirms that the RegCM4. 7 model under the Tiedtke -Holtlag configuration is capable of weekly cumulative prediction precipitation well in advance, although it has weak capability in predicting precipitation spatial pattern. %U