The Nash–Sutcliffe model efficiency coefficient have been introduced as one additional performance indicator that will also be possible to use for selection of output. Note that coefficient is base on the summation of the squared sum of residuals between simulated and measured and the corresponding squared summation of the residual between observed values and the mean of the observations. This means that a simple selection based on RMSE and a selection based on the Nash-Sutcliffe coefficient will give the same results. However, the Nash-Sutcliffe coefficient is normalized with an upper limit of 1 whereas the RMSE will be directly dependent on the units and magnitude of the variable considered.
Statistics of the various performance indicators will be shown after selecting the right click option of the validation sheet of a completed CoupModel multirun document.