Hydrological and water quality modeling of Küçük Menderes River Basin and management scenarios

This study aims to investigate the water quantity and quality in the Küçük Menderes Basin, where agricultural activities are intense, and which observe the effects of different agricultural practices. We used the SWAT model. The model outputs, which were simulated between 2001-2019 with a 5-year warm-up period, were evaluated with monthly flow, sediment, NO2, NO3, NH4, and TN variables. We calibrated the model simultaneously for flow, sediment, and nutrient variables, using the SWAT-CUP SUFI-2 algorithm. 1500 simulations and 114 parameters were used in the calibration. Although the calibration and validation intervals vary for each station and the calibration period ranges between 2006 and 2014 and the validation period ranges between 2009 and 2018. Ten observation stations were used for flow, two for sediment, and one for quality. Calibration results were evaluated using R2, NSE, and PBIAS objective functions, p-factor and r-factor uncertainty analyses, global sensitivity analysis, and graphical visualizations. Satisfactory results were obtained for flow calibration, with an overall NSE score of 0.65, whereas sediment calibration failed. Good values were obtained from total nitrogen for quality, but the number of data was insufficient. Validation results also vary between stations with 0.5 - 0.8 for R2. A scenario analysis was carried out with three different agricultural applications: a 30 % reduction in irrigation amount, a 30 % reduction in fertilizer amount, and the application of both. The scenario results showed a reduction of 18.8 % for the annual average nitrate load when fertilizer amounts were reduced by 30 % and a reduction of 17.2 % for the annual average nitrate load when irrigation and fertilizer amounts were reduced. No major differences were seen for other variables.