Climatology
Mahnaz Saber; Bromand Salahi; Roghayeh Maleki Meresht
Abstract
In this research, the water balance components of the Aras basin area were simulated in the SWAT model for a period of 28 years (1987-2014). For this purpose, the efficiency and capability of the SWAT model by SWAT CUP using the SUFI2 algorithm and based on the observed discharge data in the selected ...
Read More
In this research, the water balance components of the Aras basin area were simulated in the SWAT model for a period of 28 years (1987-2014). For this purpose, the efficiency and capability of the SWAT model by SWAT CUP using the SUFI2 algorithm and based on the observed discharge data in the selected hydrometric station of Aras basin (Bdoy) with 70% of the data (1987-2006) and 30% of the rest (2007-2014) was validated. Based on the raster data input to the model, this basin was divided into 68 subbasins with 1264 hydrological response units (HRUs) and calculations were performed on their level. SWAT model calibration was done by using 14 important parameters that were selected from several parameters based on the comparison of sensitivity analysis results. In the sensitivity analysis stage of the model, parameters related to monthly temperature, air temperature, and soil evaporation factor from.bsn,.wgn, and.hru files were identified as the most effective parameters in simulating the flow discharge of the selected hydrometric station of Aras Basin. By running 300 times of calibration, finally, the best round of simulation based on the target criteria was identified and the output data was evaluated. The efficiency and accuracy of the model in the calibration period (1987-2006) based on the evaluation criteria of NS, P-Factor, R-Factor, and R2 were calculated as 0.64, 0.71, 0.27, and 0.79 respectively, which show the satisfactory performance of the model. In the water balance simulation, it is Aras Basin. The values of these criteria in the validation period were calculated as 0.7, 0.78, 0.3, and 0.68 respectively.
Climatology
Roghayeh Maleki Meresht; Bromand Salahi; Mahnaz Saber
Abstract
The current research was carried out to analyze the changes in precipitation in northwest Iran during the coming decades based on GCM models. For this purpose, first, the precipitation of 1985-2014 was trended based on the Mann-Kendall test. Then, the daily precipitation data for each of the studied ...
Read More
The current research was carried out to analyze the changes in precipitation in northwest Iran during the coming decades based on GCM models. For this purpose, first, the precipitation of 1985-2014 was trended based on the Mann-Kendall test. Then, the daily precipitation data for each of the studied stations was simulated in SDSM6.1 software for 1985-2014. Then, under the scenarios (SSP2-4.5) and (SSP5-8.5) of CanEsm5 and MPI-ESMI-2HR models, the precipitation of 2015-2043 was predicted. To evaluate the performance of CMIP6 models and compare the basic and predicted values, MSE, RMSE, and MAE statistical measures were used. According to the results of the Man-Kendal test, the precipitation of the base period in the stations of Tabriz, Ardabil, Urmia, Takab, and Maragheh has a decreasing trend and in the stations of Meshginshahr, Sardasht, Mako, Khalkhal, Sarab, Jolfa, and Parsabad it has an increasing trend. Among the 12 investigated stations, only the Maragheh station had a significant decreasing trend. In other stations, precipitation trends were not significant. According to the predictions made based on the mentioned models, under the medium scenario (SSP 2-4.5), the precipitation will decrease in late winter and early spring. In other months, especially summer and autumn months, the percentage of precipitation will be higher. Based on the SSP5-8.5 scenario, the highest percentage of precipitation decrease in the MPI model was predicted by 33% in Jolfa, Sardasht, and Maragheh stations, and in the CanESM5 model, about 33-35% in Jolfa, Takab, and Urmia stations. According to the results, although both models predicted precipitation with a relatively high error, the MPI model had a lower error and more accuracy in predicting precipitation than the CanESM5 model.
