Evaluating climate changes and land use changes on water resources using hybrid Soil andWater Assessment Tool-DEEP optimized by metaheuristics

摘要

The objective of this research is to investigate the effect of land-use and climate change on the volume of runoff produced and its impact on water resources. It is clear that the more accurate the assessment of runoff in the future, the better the evaluation of the impact of land-use and climate changes on water resources. Thus, to more accurately predict water sources in the future, runoff is estimated using two Soil and Water Assessment Tool (SWAT) and SWAT-DEEP-IFSO models. The results showed that the SWAT-DEEP-IFSO model has a better estimation than the SWAT method. Because Deep Belief Network and an improved optimization algorithm called Improved Fluid Search Optimization (IFSO) algorithm have been utilized in optimizing the hybrid SWAT model that can reduce the error of runoff simulation. After selecting the best model for estimating runoff, the effect of land-use and climate change is evaluated using different scenarios. Results show that the simulated runoff under the influence of climate and land-use changes can impact on the average yearly and monthly runoff. These factors have a significant effect on the average yearly runoff changes and are predicted an increasing trend for average yearly runoff. But climate and land-use changes have a different impact on the different months and for some months of the year the runoff has an increasing trend and for some months the runoff has a decreasing trend. Based on land-use scenarios, it is found that due to the decrease in rangelands and increase in agricultural lands, there is a possibility of severe floods and drought periods in the region, which reduces the water resources of the study area.