Introduction
One of the most important consequences of climate change in recent years is the increase in droughts, floods, and especially land use change caused by droughts in Iran. Azarakhshi et al. (2017) Investigation of the Effect of Precipitation Changes and Land Use on Sediment Production) Case Study: Watershed Senobar- Torbat  Heydarieh). It was concluded that an increase in droughts, resulting in a decrease in river discharge and conversion of arable lands to rangelands, led to a decrease in the suspended sediment content of the poplar watershed. Salmani et al. (2018) evaluated the hydrological response of Til Abad watershed during future periods due to predicted land use change. The results show that monthly runoff simulation is more accurate than sedimentation and the effects of changes in 2040 are more than 2025p.
Materials and Methods
In this study, the impact of land use and drought on runoff of Tuyserkan watershed in central Zagros during three periods of 1991, 2003 and 2015 (25 years) was investigated using hydrological modeling of rainfall runoff relationship (IHACRES). Firstly, for data acquisition and survey of land use, digital data of TM, ETM + sensors and Arc Gis 9,3 software, as well as for drought calculation of SPI index with DIP model and annual time period as a common hydrological cycle has been used. After the runoff simulation stage in the IHACRES model, the simulated data were extracted for the whole study period and validated by regression method and R2, APRE and BIAS indices were used for evaluation in this study..
Discussion and Results
The results show that land use in the studied period of 25 years has area changes in certain classes and shows that natural resource areas in Tuyserkan basin are decreasing and most land use is affected by drought. In addition, even in 2015, with the annual change in land use, it has reduced the areas of natural resources, especially conversion of rangelands to dryland and increased runoff. Studies show that drought is a climate phenomenon that is affected by precipitation and temperature and has a cross-sectional trend in the Tuyserkan Basin and has severe fluctuations. In general, drought severity decreased in the study period. According to the results of 1991, the humidity was mild and hydrologically low in the 1991-1993 crop year, although drought severity decreased. The year 2003 is relatively humid and the crop year 2002-2003 and 2003-2004 are humid and there is no drought which is the main cause of increased rainfall. In 2015, the drought index assessment showed a moderate moisture index, the main cause of which is likely to be increased fine dust and rainfall and drought has not been very evident, and four months of this wet season have been reported. During the study period, runoff increased with increasing rainfall, indicating that runoff was subject to changes in rainfall and drought severity did not have a significant effect on it. The results of statistical indices show that the factor P value is lower than the calibration period, which indicates more model uncertainty at this stage. The coefficients of NS and R2 at this stage were higher than the calibration stage, and the factor R value was less than one, indicating a relatively good discharge simulation at this stage. The RMSE value indicates a model error and the low values of this criterion represent better performance of the IHACRES model in the runoff simulation.
Conclusions
Land use changes and drought fluctuations affect the runoff of the watershed. So that the reduction of natural areas due to land use change over time increases the volume of runoff. Climate change drought also reduces runoff volume and increases runoff with wetter season. The results of 3 and 12 month drought index analysis over a 25 year period show that most of the years are classified into three classes of mild wet, mild and very severe drought, with severe hydrological fluctuations. Surveys show that most land use change over a period of 25 years has occurred in natural resource areas, especially in rangelands affected by drought and decreasing this area along with drought fluctuations over time has increased runoff volume has been reduced. The results obtained from calibration of simulation runoff in rainfall and runoff model indicate acceptable model performance of the model and the basin hydrograph study at three time periods of 1991, 2003 and 2015 (25 years) shows that runoff in the basin has an increasing trend. This is because the drought is not so noticeable, significant land use change and increased rainfall in some years. Therefore, it is suggested to develop a comprehensive plan for control and prevention of land use change and land preparation in the watershed in order to manage water and soil resources.