“…For state variables many authors use mean (absolute) error (M(A)E), root mean 260 square error (RMSE), index of agreement (IoA;Willmott, 1982), and less often the Nash-Sutcliffe modelling 261 efficiency (NSE; Nash-Sutcliffe, 1970) (e.g., Donatelli et al, 2004;Gribb et al, 2009;Herbst et al, 2005; 262 Khodaverdiloo et al, 2011;Patil and Rajput, 2009;Ritter et al, 2003;Vereecken et al, 2010). For rate variables 263 authors generally use MAE, mean difference (MD), absolute maximum error (AME), RMSE, IoA, NSE, 264 coefficient of determination (R 2 ), percentage of error (PE), percentage of bias (P bias ) (e.g., Akkal-Corfini et al, 265 2010;Ale et al, 2012;Dawson et al, 2007Dawson et al, , 2010Jabro et al, 2012;Jachner et al, 2007; Kersebaum et al, 266 2007; Krause et al, 2005;Moriasi et al, 2007;Qi et al, 2012;Reusser et al, 2009;Stumpp et al, 2009; Van der 267 Laan et al, 2011;Wang et al, 2006;Willmott et al, 1985).…”