This threshold value was selected so as to best capture the variability of drainage densities among

the studied catchments. Four variables representing mean drainage directions were calculated, namely South, Southwest, West and Northwest. A value of 1 (or 0) means that the catchment is draining toward the named direction (or opposite to the named direction). The geographic coordinates of the flow gauging stations (latitude and longitude) were selected as two additional candidate explanatory variables (Table 2). Two soil characteristics, likely to control Crizotinib in vitro hydrological processes, were selected from the MRC soil database (MRC, 2011): soil depth and top soil texture. A four-unit scale suggested by MRC was used for

quantification (Table 1). Averaged values for each soil characteristics and each catchment were averaged by weighting each scale unit by the respective area covered in the catchment. Three land-cover types, likely to alter hydrology, were selected as candidate explanatory variables: forest, bunded rainfed lowland rice paddy fields, the majority of which is never irrigated, and wetlands, including marsh and swamp. The percentage of surface area covered by each land-cover type in each catchment was computed using the digitized 2003 land cover map of the Lower Mekong Basin prepared by MRC (2011). Forest cover was produced by merging four forest types available as separate land-cover classes in the published map: “coniferous forest”, “deciduous forest”, “evergreen forest” and “forest plantation”. The two other land-cover types were directly available since they RG7422 mw correspond to distinct land cover classes on the published map. Table 3 presents the results of the multiple regression analyses for the 14 flow metrics listed in column 1. Column 2 provides the value of the intercept term β0. Columns 3–11 provide the coefficients βt associated with each explanatory variable Xi included in the power-law models (cf. Eq. (1)). Units of the explanatory variables are indicated in Table 2. Values of the explanatory variable “Padd” and of the flow metrics 0.50, 0.60, 0.70, 0.80, 0.90,

0.95 and Min ( Table 3) should be incremented by 1 for inclusion in Eq. (1) (cf. Section 2). As examples, Eqs. (6) and (7) show how to predict the 0.95 flow percentile (Q0.95) Oxymatrine and mean annual flow (Qmean) using the coefficients provided in Table 3. equation(6) Q0.95=exp−27.857×Rain2.698×Peri1.436×Elev0.966×Lati−1.291×(Padd+1)−0.285−1Q0.95=exp−27.857×Rain2.698×Peri1.436×Elev0.966×Lati−1.291×(Padd+1)−0.285−1 equation(7) Qmean=exp−18.989×Rain2.543×Area0.883×Drai1.089Qmean=exp−18.989×Rain2.543×Area0.883×Drai1.089 In order to make the power-law models usable by a broad range of users, Table 3 presents, for each of the 14 flow metrics, an equation including climatic, geomorphologic and/or geographic explanatory variables only, exclusive of other catchment characteristics.