Providing high-resolution surface conditions using a coupled land-surface groundwater model: effects on atmospheric boundary layer simulations over Owens Valley, CA

Land surface conditions such as soil moisture and soil temperature affect flow in the atmospheric boundary layer through surface heat fluxes. Surface data to initialize high-resolution simulations of boundary layer flow are often interpolated from coarse grids because sufficiently high-resolution data are unavailable. In Owens Valley, a region of complex topography, the coarse grid soil moisture data available from both North American Regional Reanalysis (NARR 32 km) as well as the North American Mesoscale Model (NAM 12km) are approximately a factor of three greater than observations of soil moisture from the Terrain-Induced Rotor Experiment (T-REX, March and April, 2006). Coarse adjustments to the soil moisture field have shown strong sensitivity to initial and land surface conditions, yielding improvements in comparisons between atmospheric boundary layer simulations and observations of wind speed, direction, specific humidity, and potential temperature.

In this study, a coupled land-surface groundwater model is used to provide high-resolution initial surface conditions for simulations of flow over Owens Valley. The coupled model consists of a variably-saturated groundwater model coupled to a land-surface model driven by meteorological forcing from NARR during a multi-year spin-up procedure. Use of such a physically-based coupled model allows for realistic representation of surface and deep soil moisture distributions that reflect topographically-influenced variations in the valley region, without requiring calibration to the particular catchment or relying on observational data. Results from the coupled model are compared to surface observations from T-REX and are used to initialize atmospheric simulations.