Abstract. The interactions between shallow groundwater and land surface processes, mediated by capillary rise processes from groundwater, may play an important role in the ecohydrology of riparian zones in both humid and semi-arid ecosystems. Some recent land surface models (LSM) incorporate the contribution of groundwater to land surface processes with varying levels of complexity. In this paper, we examine the sensitivity of evapotranspiration at the land surface to the depth of groundwater using three models with different levels of complexity, two widely used representative soil hydraulic parameter sets, and four soil textures. The selected models are Hydrus-1D, which solves the Richards equation, the Integrated Biosphere Simulator (IBIS), which uses a multi-bucket approach with interactions between buckets, and a single-bucket model coupled with a classic simple capillary rise flux approximation. These models are first corroborated with field observations of soil moisture and groundwater elevation data from a site located in south-central Nebraska, USA. We then examine the sensitivity of the Richards equation to node spacing, as well as the relationship between groundwater depth and the ratio of actual to potential evapotranspiration (ET) for various soil textures and water table depths. The results show that selecting one representative soil parameter set over another may result in up to a 70% difference in actual ET (relative to the potential ET) when the depth to water table is in 0–5 m depending on the soil type. Moreover, solution type of the Richards equation and node spacing have also effect on surface ET up to 50% and 30% respectively depending on the depth-to-groundwater and node spacing. Therefore, further studies are needed to understand the sensitivities of land surface and atmospheric models to the existence of saturated layers, including studies with more field validation in regions with different climates and land cover types.
Land Surface Model and Particle Swarm Optimization Algorithm Based on the Model-Optimization Method for Improving Soil Moisture Simulation in a Semi-Arid Region
