River recharge sources and the partitioning of catchment evapotranspiration fluxes as revealed by stable isotope signals in a typical high-elevation arid catchment

Catchment-scale hydrological cycles are expected to suffer more extremes under a background of climate change. Quantifying hydrological changes in high and remote areas is practically challenging. However, stable isotopes in river water can be seen to vary, dependent upon the combined influence exerted by recharge sources and local climatic conditions; the study of river water stable isotopes can therefore provide a meaningful method for delineating catchment-scale hydrological studies. In this study, we present high-resolution time series of river delta C-18 and d-excess values; additionally, we identify the seasonal dynamics of river recharge sources and major components of the catchment-scale water balance, together with precipitation and groundwater isotopes, and concurrent meteorological data recorded in Magazangbu catchment on the northwestern Tibetan Plateau (TP). Using isotopic analysis, and within a proportional framework, we partitioned the isotopic fractionation (El) or non-fractionation (E2) from soil evaporation fluxes (E-soil) apparent in different processes, using NDVI (Normal Differential Vegetation Index) data collected by MODIS satellites to calculate the vegetation fractional coverage (VFC), and Global Land Data Assimilation System (GLDAS) records to determine evapotranspiration data (ET). Finally, the contributions made by each ET component (E-soil and plant transpiration) to total catchment ET were computed for the high and remote northwestern TP. Our results show that: (1) river 8180 values were high in summer and low in winter, while d-excess values displayed a contrary seasonal cycle; (2) for the monsoon period, precipitation contributed 60.6% to Magazangbu catchment runoff. Deeper groundwater was the main water source for the winter low base flow, and shallow groundwater or high elevation snowmelt was the principal component of the spring thaw and autumn freezing periods; and (3) a substantial proportion of E-soil (96.4% annually; 92.2% during monsoon) was consumed without isotopic fractionation (E2); plant transpiration (T) constituted less than half of total ET (41% annually, 29% during monsoon) in Magazangbu catchment. This calculation of river recharge sources and partitioning of catchment ET components using isotopic signals and MODIS NDVI data or GLDAS ET data provide new methods for hydrological studies in high and remote areas. These results provide important catchment-scale water-balance information which is very useful to climate models conducted in a high-elevation arid environment. (C) 2017 Elsevier B.V. All rights reserved.