Determination Of Deep Drainage Rates On Hillslopes Of The Chinese Loess Plateau Using Environment Tracers

Loess Plateau located in northwestern China, wherein water plays a dominant role for plant growth due to limited precipitation and arid climate. The grain for green project which has initiated since 1999, has improved local environment and soil erosion effectively. However, irrational land use conversion has induced new environmental problems, such as soil desiccation and groundwater recharge decrease, which have threatened the sustainability of grain for green project. Thus quantify groundwater recharge is vital for evaluation of the sustainability of the project. Previous study focused more on hydrological process on tableland and only few of them paid attention to water process on hillslope scale. The aim of this study is to clarify hydrological processes at hillslope scales and the research can be divided into three aspects:(1) To quantify groundwater recharge at different slope positions using environmental traces;(2) To characterize soil water consumption at hillslope scales under different land use type; and(3) to explore soil profile water isotope distribution patterns under different land use type on loess hillslope. The main findings were congregated as follows:(1)Water consumption characteristic on the tableland and hillslopes under different land use type. Compared with shallow rooted plant(grass and maize), deep rooted plant(apple tree and apricot) appeared water deficit after several planting years. On tableland, water consumption depth reached 15 m in 17-year-old apple orchard. For the 23-year-old apple orchard in upper hillslope, water consumption depth reached 22.6 m, but soil water deficit amount was limited due to ageing and poor management. For apricot in middle and down slope, the water consumption depth ranged between 6 and 7 m.(2)Average infiltration groundwater recharge rate on loess hillslope. Clear and complete tritium peaks were found at different slope position. This indicated that piston flow is the main water flow mechanism. Using the tritium peak method, groundwater recharge rates were calculated as 20 mm y-1, 22 mm y-1 and 27 mm y-1 for upper, middle and down hillslope, respectively. The recharge accounted for 3%-5% of annual precipitation. The recharge rate in the middle and down hillslope was 1.1 and 1.35 times of that in upper hillslope. This can be attributed to precipitation induced surface runoff that provided more water to the lower position and enhanced groundwater recharge at lower slope positions. Using modified chloride mass balance method, groundwater recharge rate were calculated as 19-23 mm y-1, 23-29 mm y-1 and 37-45 mm y-1 for upper, middle and down hillslope, respectively. The recharge accounted for 3.3%-7.7% of annual precipitation. In summary, modified chloride mass balance method resulted in larger groundwater recharge rates as compared with the tritium method. This main reason is that Cl- is the anion exclusion. Compared with grass land, tritium peaks were shallower and peak derived infiltration rate were 6.0 cm y-1,7.8 cm y-1,10.0 cm y-1 for upper, middle and down slope position. Slow infiltration rate can be attributed to deep rooted plant water uptake from deeper soil layer and inhibit vertical flow.(3) Using environmental tracers to characterize groundwater recharge. As affected by soil evaporation and precipitation, soil water isotope depleted in soil surface and tends to stable in deep soil layer. Combined with tritium profile, we concluded that piston flow was the main recharge pattern in grass hillslope. Compared with grassland, forest on the hillslope accompanied with water isotope depletion. In addition Cl- was diluted in upper hillslope and tritium profile fluctuated dramatically. The main reasons for this phenomenon can be attributed to rapid flow path caused by root decay that leads to precipitation infiltrated quickly to deep soil layer through these channels.