Incorporating Advanced Surface and Subsurface Processes in Mesoscale Climate Models
MetadataShow full item record
Regional anthropogenic climate change poses significant risks to the security of water resources for communities throughout the world. Current climate simulations seek to predict the risks to water resources by employing land surface models (LSMs). While LSMs incorporate biogeophysics, heat, albedo, surface water, and shallow subsurface water, they do not include lateral surface/subsurface flow, groundwater storage, or critical feedbacks between surface and subsurface hydrology. Consequently, the shortfalls of current models severely limit our abilities to predict and understand risks to water resources. Therefore, this study investigates the development of coupling HydroGeoSphere (HGS), an advanced 3D control-volume finite element surface and variably-saturated subsurface model, to two separate atmospheric models to capture the interactions between the deep subsurface, surface, and atmosphere. Initially, HGS was coupled to a simple 0D atmospheric boundary layer (ABL) model, hereafter referred to as the HGS-ABL model. The coupled HGS-ABL model physically resolves boundary layer dynamics, precipitation, evapotranspiration, energy balance, surface water, and groundwater flow. The experimental simulations showed that current LSMs are too shallow for handling deep root-zones and do not provide an adequate representation of subsurface heat storage. Furthermore, the HGS-ABL simulations showed a positive correlation between the soil moisture and the energy feedbacks. To transition from a 0D to a 3D atmosphere, this study then coupled HGS to the Weather Research and Forecasting (WRF) Model, a 3-dimensonal mesoscale nonhydrostatic atmospheric model, hereafter referred to as the HGS-WRF model. HGS replaces the land surface components of WRF by providing the actual evapotranspiration (AET) and soil saturation from the porous media to the atmosphere. In exchange, WRF provides HGS with the potential evapotranspiration (PET) and precipitation fluxes. The flexible coupling technique uniquely accepts independent model meshing and projections and links domains based on their geographic coordinates (i.e., latitude and longitude). The newly coupled HGS-WRF model was then implemented over the entire California Basin. This 3D California Basin Model is 14-layers thick with over 400,000 nodes. The geological model was based on the STATSGO2 soil data, USGS HYDRO1K topographic data, and USGS water use data. Initially, the HGS model was spun-up with historic precipitation and PET data (provided by CMIP5). Once the model reached steady state, groundwater pumping was turned on, and the HGS model was run to present-day conditions. The HGS California Basin Model simulated similar drawdown rates to the Gravity Recovery and Climate Experiment (GRACE), a 21st century remote sensing satellite. Finally, the HGS-WRF model simulated the California Basin for a 200 day period and successfully replicated the Klamath river, Sacramento river, precipitation, and evapotranspiration fluxes.
Cite this work
Jason Davison (2016). Incorporating Advanced Surface and Subsurface Processes in Mesoscale Climate Models. UWSpace. http://hdl.handle.net/10012/10904