Investigation into approaches to reduce excessive vertical transport over complex terrain in a regional photochemical grid model

Past photochemical modeling of the western United States (US) using the Comprehensive Air quality Model with extensions (CAMx) and the Community Multiscale Air Quality (CMAQ) model has resulted in large springtime ozone over predictions in the complex high-elevation terrain of the western United States (US). Comparisons against rural measurement data have shown that both models over predicted ozone levels by 20 ppb or more. A systematic investigation using CAMx revealed that excessive vertical transport in mountainous terrain draws down upper tropospheric ozone introduced by the lateral boundary conditions (developed by a global chemistry model), which can routinely exceed 1 ppm near the tropopause. This is not an unreasonable concentration at such altitudes during the spring, and there is observational evidence that stratospheric ozone intrusions result in occasional large ground-level concentrations in the western US, but not at the frequency and intensity simulated by the CAMx and CMAQ models. Past versions of CAMx and CMAQ possess similar algorithms to diagnose vertical velocity, and similar first-order accurate vertical advection algorithms. These similarities, in conjunction with poor vertical resolution of the upper troposphere, have resulted in similar ozone performance issues. Numerous approaches were explored with CAMx in an attempt to externally reduce the rates of vertical transport over complex terrain. Ultimately, we formulated and tested a new vertical advection methodology that included improvements to how vertical velocities are determined and introduced a second-order accurate advection solver technique. Together these improvements proved to yield the most successful results in reducing upper tropospheric ozone transport to the surface. CAMx was then run to simulate ozone throughout the western US for a full year to evaluate the effects of the new vertical transport algorithm on model performance. Ozone performance improvements exceeded those achieved through the application of arbitrary reductions in the upper tropospheric/stratospheric lateral boundary conditions.

► CAMx and CMAQ have over predicted ozone in the inter-mountain western US.
► Primarily caused by treating vertical circulations with simple advection schemes.
► This is important for modeling background ozone and lower ozone standard.
► Several approaches were tested in CAMx to resolve this problem.
► Higher resolution and a higher-order vertical advection scheme were most effective.