A watershed scale Groundwater-Land-Surface Model

Weather and climate models rely on land surface models (LSMs) to represent land-surface processes. Subsurface waters, however, are not well described in most LSMs. Coupled models of the atmospheric boundary layer, land surface and subsurface, which incorporate groundwater components into LSMs and couple the deeper subsurface with the atmosphere, may yield significant improvements in both short-term climate forecasting and flood/drought forecasting. A groundwater-land-surface model system has been developed from the Penn State Integrated Hydrologic Model (PIHM). The land-surface scheme is mainly adapted from the Noah LSM, which is widely used in mesoscale atmospheric models and has undergone extensive testing. Because PIHM is capable of simulating lateral water flow and has deep groundwater, the new model is able to represent some of the land-surface heterogeneity caused by topography. At the same time, the robust land-surface scheme incorporated into PIHM provides accurate sensible heat flux and evapo-transpiration rates. The closely-coupled hydrologic and land-surface schemes guarantee mass conservation throughout domain and energy balance at land surface, and provide physical constrains to surface heat fluxes and groundwater. The new model has been implemented and calibrated for the Shale Hills watershed in central Pennsylvania and run from 0000 UTC 1 May to 0000 UTC 1 June 2009. Parameters including soil hydraulic conductivity, macropore hydraulic conductivity, soil porosity, van Genuchten soil parameters, and minimum stomatal resistance are optimized to fit the model outputs with in-situ measurements. The model is driven by North American Regional Reanalysis (NARR) atmospheric forcing data. The model reproduces realistic topographically-induced distributions of water table, soil moisture, and skin temperature. The sensible and latent heat fluxes simulated by the new LSM compare well with the eddy-covariance flux measurements on most days. Errors are relatively large when NARR fails to provide realistic radiation data, e.g. on 8, 11, 26 and 27 May. The model is able to capture the fluctuation of water table reasonably well. It therefore provides more realistic soil moisture variations which benefits the simulation of the surface energy balance. The simulated sensible heat flux, soil evaporation, canopy transpiration, and skin temperature are correlated with water table depth. The sensitivity of these land-surface variables to ground water is small when water table depth is smaller than 0.1 m, large when water table depth is between 1.5 m and 2.0 m, and intermediate when water table depth is between 0.1m and 1.5 m. Coupled models such as this have the ability to advance our understanding of subsurface-land surface-atmosphere interactions by providing better representations of the Earth surface memories present in groundwater dynamics.