I have successfully run SWAT model in my study area, but I met problems in calibration by SWATCUP. My best result is that R square equals 0.50 and NS equals 0.46, which was low and far from criteria. Every time I increased iterations, R square increased slightly while the NSE value decreased intensely.What is the reason for this? How can I solve this problem? Please give me some advice. Thanks a lot!