Use of stable isotopes to understand run‐off generation processes in the Red River Delta

This paper presents the use of stable isotopes of water for hydrological characterization and flow component partitioning in the Red River Delta (RRD), the downstream section of the Red River. Water samples were collected monthly during 2015 from the mainstream section of the river and its right bank tributaries flowing through the RRD. In general, O-18 and H-2 river signatures were depleted in summer-autumn (May-October) and elevated in winter-spring (November-April), displaying seasonal variation in response to regional monsoon air mass contest. The Pacific equatorial-maritime air mass dominates in summer and the northern Asia continental air mass controls in winter. Results show that water of the RRD tributaries stems solely from local sources and is completely separated from water arriving from upstream subbasins. This separation is due to the extensive management of the RRD (e.g., dykes and dams) for the purposes of irrigation and inundation prevention. Mainstream river section O-18 and H-2 compositions range from -10.58 and -73.74 parts per thousand to -6.80 and -43.40 parts per thousand, respectively, and the corresponding ranges inside the RRD were from -9.35 and -64.27 parts per thousand to -2.09 and -15.80 parts per thousand. A combination of data analysis and hydrological simulation confirms the role of upstream hydropower reservoirs in retaining and mixing upstream water. River water inside the RRD experienced strong evaporation characterized by depleted d-excess values, becoming negative in summer. On the other hand, the main stream of the Red River has d-excess values around 10 parts per thousand, indicating moderate evaporation. Hydrograph separation shows that in upstream subbasins, the groundwater fraction dominates the river flow composition, especially during low flow regimes. Inside the RRD, the river receives groundwater during the dry season, whereas groundwater replenishment occurs in the rainy season. Annual evaporation obtained from this hydrograph separation computation was about 6.3% of catchment discharge, the same order as deduced from the difference between subbasin precipitation and discharge values. This study shows the necessity to re-evaluate empirical approaches in large river hydrology assessment schemes, especially in the context of climate change.