Investigating the effect of time scale variations of hydro climate parameters on the accuracy of LARS-WG6 climate change model

In this research, in order to investigate the effect of short and long term observational data on the quality of predicting climate parameters, the LARS-WG6 statistical downscaling model is performed for EC-EARTH model and the optimistic scenario RCP4.5 and the pessimistic scenario RCP8.5. Also the HadGEM2-ES model with the optimistic scenario RCP2.6 and the moderate scenario RCP4.5 and the optimistic scenario RCP8. 5 are investigated. Forecasts are applied on the next three 20-year periods (2021-2040, 2041-2060, 2061-2061). Three daily statistical periods of the Drudzen Dam Synoptic Station including the 30-year period (1988-2017), the 20-year period (1998- 2017) and the 18-year period (1988-2005), were used. The 18-year period was determined by examining the 30-year trend. A fractal based procedure is employed for extracting a complete pattern of repeating weather events including rainfall, minimum and maximum temperatures. The modelchr(chr('39')39chr('39'))s capability in the simulation was measured according to the base period and the accuracy of the anticipations was evaluated using MAD, MSE and RMSE statistical tests. Studying the results clearly shows the effect of time scale changes on the accuracy of predictions. Accordingly the 20 year time period showed the least error, especially for the precipitation parameter. In general, it was revealed that this model predicts minimum and maximum temperatures with high accuracy and the rainfall prediction holds less accuracy. The prediction of models under different scenarios and different time intervals showed an increase in minimum and maximum temperatures. Moreover, the model indicates while the average annual precipitation increases in the coming years, the pattern of precipitation will change. In April, January and December, we will see an increase in precipitation and a decrease will be experienced in February.