Impact of Lower Boundary Condition of Richards' Equation on Water, Energy, and Soil Carbon Based on Coupling Land Surface and Biogeochemical Models

Soil moisture has a significant influence on water, energy, and carbon biogeochemical cycles. A numerical method for solving Richards' equation is usually used for simulating soil moisture. Selection of a lower boundary condition for Richards' equation will further affect the simulation results for soil moisture, water cycle, energy balance, and carbon biogeochemical processes. In this study, the soil water movement dynamic sub-model of a hydrologically based land surface model, the variable infiltration capacity (VIC) model, was modified using the finite difference method (FDM) to solve a mixed form of Richards' equation. In addition, the VIC model was coupled with a terrestrial biogeochemical model, the Carnegie Ames Stanford Approach model of carbon, nitrogen, and phosphorus (CASACNP model). The no-flux boundary (NB) and free-drainage boundary (FB) were selected to investigate their impacts on simulations of the water, energy, and soil carbon cycles based on the coupling model. The NB and FB had different influences on the water, energy, and soil carbon simulations. The water and energy simulations were more sensitive, while the soil carbon simulation was less sensitive to FB than to NB. Free-drainage boundary could result in lower soil moisture, evaporation, runoff, and heterotrophic respiration and higher surface soil temperature, sensible heat flux, and soil carbon content. The impact of the lower boundary condition on simulation would be greater with an increase in soil permeability. In the silt loam soil case, evaporation, runoff, and soil respiration of FB were nearly 16%, 13%, and 1% smaller, respectively, compared to those of NB.