Projection of summer monsoon rains in Southeast Iran based on Ensembel model

Projection of summer monsoon rains in Southeast Iran based on Ensemble model the precipitation factor has a variable and random nature and has a different behavior in terms of space and time. Therefore, the prediction of precipitation has more uncertainty compared to other meteorological variables. In the current study, the data output of Cordex database and CMIP5 models were used using the neural network method to reduce the uncertainty and estimate precipitation properly. The results showed that due to the high correlation of temperature, humidity and air pressure with precipitation, the use of these variables is beneficial in reducing the uncertainty of precipitation forecast. In addition, the non-linear method of artificial neural networks can be used to bias the rainfall data of Cordex database and CMIP5 to forecast the rainfall in the southeast of the country. Another result of this research is the increasing trend of rainfall in the southeast of Iran, especially in the coastal areas. This can be considered as a result of the increase in the level under the influence of rains affected or simultaneous with the southwest monsoon of India. The increasing trend of precipitation in the southern coasts is also related to the increase in the storage capacity of moisture content. The interannual variability of India's monsoon rainfall also shows a steady positive trend under continued global warming. Since both the increase in the duration of monsoon rains and the increase in interannual variability in the future are seen in most models, we can be confident in these predicted trends. Indian summer monsoon rainfall is also predicted to be higher under global warming in the 2050s compared to the baseline.