6.2
Improving regional model predicted ozone distributions over California by using “satellite observation-constrained” boundary conditions from a global model

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Wednesday, 5 February 2014: 10:45 AM
Room C113 (The Georgia World Congress Center )
Min Huang, JPL, Pasadena, CA; and K. Bowman, G. Carmichael, M. Lee, and D. K. Henze

Accurately simulating the near-surface ozone is critical for identifying “non-attainment” areas and for helping develop effective control strategies. Previous studies have shown that near-surface ozone in the western US simulated by regional models can be sensitive to the boundary conditions used. Therefore, improving the quality of boundary conditions may help reduce the uncertainties in the regional model simulated ozone fields. In this study, we simulate California air quality during the NASA ARCTAS field campaign period in Spring and Summer 2008 using a coupled chemical transport modeling system. This modeling system is composed of a global model (GEOS-Chem) on a 2°×2.5° horizontal resolution grid and a regional model (STEM) on a 12 km horizontal resolution grid over California. The STEM simulations used time-varying top and lateral boundary conditions for various gases and aerosol species downscaled from the GEOS-Chem simulations. The GEOS-Chem simulation with no “observational constraints” overall underestimated ozone in the free troposphere in the eastern Pacific and the western US, consistent with previous findings. The three-dimensional variational (3D-Var) approach (provided by GEOS-Chem-M2O2 system, a research task under NASA's ACCESS program whose objective is to provide a mission-generic 3D-Var assimilation capability for the GEOS-Chem adjoint community) was applied to assimilate global ozone observations by the Tropospheric Emission Spectrometer (TES) instrument on board of the Aura satellite. The comparison with available aircraft measurements and ozonesondes during this period showed that the assimilation generally resulted in strongest improvement on the simulated ozone in the free troposphere, where TES has higher sensitivity. We further evaluate the STEM-simulated ozone using the two sets of boundary conditions (with and without assimilation) from GEOS-Chem. Finally, we discuss the helpfulness of using assimilated chemical fields as boundary conditions for determining ozone exceedances in California.