Synergetic integration of SWAT and multi-objective optimization algorithms for evaluating efficiencies of agricultural best management practices to improve water quality

Abstract

Nitrate is one of the most complicated forms of nitrogen found in aquatic surface systems, which results in the eutrophication of the water. During the last few decades, due to agriculture and animal husbandry activities, as well as urban development, a significant amount of pollutants have accumulated in the Jajrood river in northern Iran. In this research, we simulated nitrate load in a rural watershed to assess the outlet stream’s qualitative status and evaluate the influence of best management practices (BMPs). To accomplish this, we prepared, processed, and integrated different datasets, including land-use land-cover (LULC) maps, physiographic layers, and hydrological and agricultural datasets. In the modeling section, the Soil and Water Assessment Tool (SWAT) was used to simulate nitrate load over 28 years (1991–2019). Additionally, the multi-objective optimization algorithm (MOPSO) was implemented to reduce the intended objective functions, including the number of best management practices and the nitrate concentration considering different scenarios. The calibration of the basin’s discharge and nitrate indicated that the SWAT model performed well in simulating the catchment’s streamflow (R2 = 0.71) and nitrate (R2 = 0.69). The recommended BMPs for reducing nutrient discharge from the basin are using vegetated filter strips on river banks and fertilizer reduction in agricultural activities. According to the results from this investigation, the integrated model demonstrates a strong ability to optimally determine the type, size, and location of BMPs in the watershed as long as the reduction criteria change. In a situation of water scarcity, the studies reported here could provide useful information for policymakers and planners to define water conservation policies and strategies.