CO2 Dynamical Downscaling in 2016 over the Contiguous United States Using WRF-VPRM, a Weather-Biosphere-Online-Coupled Model

The elusive impacts of many factors including land cover and land use changes, drought, and synoptic-scale weather (e.g., mid-latitude cyclones) on regional scale CO₂ fluxes and the subsequent CO₂ concentrations can be investigated using a weather-biosphere-online-coupled CO₂ model. As one effort to develop such a model, the Vegetation Photosynthesis and Respiration Model (VPRM, Mahdevan et al [2007]) was coupled into the Weather Research and Forecasting (WRF) model in ~2008 to simulate "weather aware" biospheric CO₂ fluxes and subsequent transport/dispersion. The resultant modeling system WRF-VPRM, however, was not thoroughly evaluated over the contiguous United States (CONUS) partially due to scarcity of appropriate observations. In this study, calibrated VPRM parameters using eddy covariance tower data over North America from Hilton et al [2013] are implemented into WRF-VPRM. The updated WRF-VPRM is then used to simulate CO₂ over CONUS with a resolution of 12 km for year 2016 using an optimal downscaling configuration. The downscaled fields are evaluated using data from the Orbiting Carbon Observatory-2 (OCO-2), Total Carbon Column Observing Network (TCCON), and Atmospheric Carbon and Transport (ACT)-America mission. Evaluation shows that WRF-VPRM reasonably captures monthly variation of column averaged CO₂ concentrations (XCO₂) and episodic variations due to frontal passages. The downscaling also successfully captures the horizontal CO₂ gradients across fronts, as well as vertical CO₂ contrast across the boundary layer top. WRF-VPRM modeling results indicate biogenic fluxes dominate anthropogenic sources to modulate XCO₂ over most area of CONUS except a few metropolitan areas such as Los Angeles in growing season from May to September.