عنوان مقاله [English]
IntroductionIn the present study, to monitor droughts, the RCP4.5 scenario of the CanESM2 model of the Fifth IPCC Report and the most appropriate distribution functions of drought indicators were used to assess the current climate change and drought conditions in the present and future. Since the drought in an area can be affected by various climatic parameters, in this study, in addition to using SPI as a practical index, the important SSI index was also used to assess drought.Materials and MethodsIn the present study, the following steps were performed to monitor, evaluate, and inform the occurrence of droughts in Tehran province.1.Quality control of precipitation and water flow parameters during the period 1986-20182.Prediction of these parameters during the period 2020-2050 based on daily output data of CanESM2 model under the RCP4.5 scenario using SDSM model3.Selecting the most appropriate distribution function with time series for both SPI and SSI index4.Drought detection and simulation using SPI and SSI drought characteristics during the next period (2050-2050).Results and discussionThe results of predicting the time series of precipitation and water flow using the DSM modelIn evaluating this model, two RMSE and MSE criteria were used, the results are given in Table 1 Mehr abad(rainfall)Latian(rainfall)Namrod(rainfall)Ahar(rainfall)Latian(Water flow)Firoz koh(Water flow)Namrod(Water flow)Jajrud (Water flow)RMSE0.350.390.390.353.70.361.062.67MSE-0.020.050.0040.051.030.06-0.0011.4 According to the results of Table 1, all eight stations had acceptable errors and it can be claimed that the SDSM model is more successful in predicting precipitation than Water flow.Selecting the most appropriate cumulative distribution functionsTables (2) and (3) show the ranking results of the studied functions for precipitation and forecasted data of meteorological and hydrometric stations. Table 2. Statistical characteristics of Smirnov Kolmogorov test according to annual (precipitation) data of meteorological stationsMehr abadLatianNamrodAharDistribution functionsrankP-ValuerankP-ValuerankP-ValuerankP-Value30.92620.91430.82630.732GAMMA40.85640.81240.64840.518Normal20.92310.93320.93320.848weibull 3p10.97630.87710.98510.932Fatigue life 3p0.9260.9580.9400.962R2-3.76-4.51-4.05-4.72ME3.924.654.185.03RMSE As shown in Table (2), at Mehrabad, Nimrud, and Ahar stations, the Fatigue life function was selected, and at the Latian station, the Wibble function was selected as the best cumulative distribution function.Table 3. Statistical Characteristics of Smirnov Klumography Test Based on Annual Data (Water flow)LatianFiroz kohNamrodJajrudDistribution functionsrankP-ValuerankP-ValuerankP-ValuerankP-Value10.98740.95240.96030.943Fatigue life 3p40.86910.97220.97510.965normal30.94220.96210.98920.949weibull 3p20.97130.96130.97040.926GAMMA0.9250.9080.9760.950R2-6.644-1-3.898-7.634ME6.4541.3304.0027.560RMSE Using the Kolmogorov Smirnov test and according to the P-Value, Normal distribution function shows a better fit for Firoozkooh and Jajroud stations The Weibull function also shows the best fit for the Namrod station, and the Fatigue life function shows the most suitable fit for the Latin station.Matching SSI and SPI drought indicators for the next periodThe results show that due to the use of distribution functions, the drought situation has had similar results based on two indicators with two different quantities. This means that the use of proposed distribution functions has greatly reduced the percentage of predictive errorConclusionThe results for future showed that Sharifabad station has the highest drought index (-2.74) based on SSI, and according to SPI, the highest drought index (-2.17) is for Latian station. It should be noted that the matching of the two indicators at Namroud and Latian stations was also studied and the results showed that the difference in the numerical values of these two quantities did not fit well for a 5 year period.