Different NSE value calculated from RSWAT

[Sreeraj Sreenivas]

Dear all,

When I ran RSWAT with the best parameter values, I noticed a slight difference between the NSE values calculated by RSWAT and my own R code (which I verified in an Excel workbook). I have observed data for the period 2010 to 2017, with flags kept as 'C' for all dates, as I need to calculate the NSE values for those periods only. I simulated the SWAT from 2004 to 2017, with a two-year warm-up period. Although the SWAT execution covered 14 years, I set my calibration range from 2010 to 2017, resulting in 2922 streamflow values in the output. RSWAT NSE value I got 0.56 but when I checked using my own code I am getting NSE value of 0.62.. Surprisingly, when I checked the NSE results, it showed values for both calibration and validation. Could you please comment on how I got the validation NSE? 

Regards,

Sreeraj