Regionalization of annual runoff characteristics and its indication of co-dependence among hydro-climate–landscape factors in Jinghe River Basin, China

Abstract

It is a challenge to properly generalize hydrological characteristics under the great heterogeneity of climate and landscape conditions across space because the linkage and interaction among hydro-climate–landscape factors are complicate and ambiguous at regional scale. In this study, multivariate statistical analyses including clustering, correlation and regression analysis were combined with Budyko and L’vovich frameworks to regionalize runoff characteristics over Jinghe River Basin of northwest China. For all 23 sub-basins, the hydrologic factors were quantified using the metrics of mean annual values and intra-annual variability of runoff. The climatic factors are determined from precipitation, potential evapotranspiration and aridity index, and the landscape factors were extracted from topography, soils and vegetation of the sub-basins. Results illustrated that the 23 sub-basins can be classified into two groups, the dry Loess Plateau (LP) and the wet Mountain Region (MR) in the study basin. The runoff metrics of sub-basins in each group present similarity in spatial distribution, intra-annual variations and the dominant influence factors of climate and landscape. But such runoff metrics characteristics and their co-dependence are significantly different between the two clustered sub-basins. Higher runoff and gentler hydrographs were observed in the MR in response to wetter and greater intra-annual variability in climate and greater spatial variability in landscape, whereas lower runoff and sharper hydrograph were seen in response to drier and greater intra-annual variability in climate, and less spatial variability in landscape in the LP. The runoff spatial distribution is more sensitive to climate spatial variation than to landscape in LP as opposed to the MR. Among the landscape factors, forest distribution is the dominant control on the spatial runoff characteristics in LP whereas topography is principal factor in MR. Our results highlight that current measures of reforestation plus marked change in climate in the Loess Plateau could lead to significant change in streamflow.