Soil water balance in response to rainfall variability over a dammed valley farmland in the Chinese Loess Plateau

Abstract

The dammed valley farmlands created by a mega land consolidation project are crucial in combating land degradation and sustaining food security in the Chinese Loess Plateau. Understanding the response of soil water balance to rainfall variability is essential for the effective water resources management in water scarcity regions facing global climate change. Based on continuous in situ 3-m soil water profile beneath a dammed valley maize farmland, the Hydrus-1D model was calibrated and then employed to simulate the soil water movement and water budgets during two maize growing seasons with contrasting hydrological conditions in 2019 (drought year) and 2020 (normal year). Results showed that the Hydrus-1D model could accurately depict the soil water dynamics at different depths during the two growing seasons. The simulated soil water storage changes (ΔS) and actual evapotranspiration (ET) also showed good agreement with the ΔS from soil water monitoring data and ET from eddy covariance method, respectively. The simulated data indicated that the maximum infiltration depth of rainfall (P) was 100 cm and 150 cm in 2019 and 2020, respectively. The soil water recharge from groundwater by capillary rise varied significantly with crop phenology and water availability, but with close sums during the two growing seasons (45.24 mm and 34.15 mm). ET dominated the soil water consumption for both years, with the ratio of ET to P of 0.98 and 0.73 during the growing seasons of 2019 and 2020, respectively. The simulated ΔS was pronouncedly lower in the drought year (51.79 mm) compared with the normal year (147.53 mm), representing 16.5 % and 30.3 % of the total growing-season P, respectively. The simulated deep drainage was 20.18 mm in 2020, but this value was reduced to zero in 2019. This research could deepen our understanding on the impacts of rainfall variability on soil water balance in the water-limited agricultural regions.