Comparison of the ACM2 vertical mixing scheme with observations taken during the TexAQS II field study

Accurate representation of the diurnal evolution of PBL depth, and turbulent transport within the PBL are critical in both meteorological and chemical processes in the atmosphere. The vertical transport of chemical species is dependent on both small- and large-scale turbulence, which are not properly considered in many vertical mixing options for meteorological and air quality modeling. In these models, local eddy diffusion schemes assume that all of the turbulence is sub-grid, and thus cannot realistically simulate convective conditions. Simple nonlocal closure models (e.g. Blackadar convective model, ACM) represent only large-scale transport driven by convective plumes, neglecting small-scale turbulent mixing. A new vertical mixing scheme, the Asymmetric Convective Model, version 2, (ACM2) has been developed that includes both an eddy diffusion scheme and the nonlocal scheme from the original ACM. Combining both vertical mixing components enables the new ACM2 to better represent the rise and fall of the convective boundary layer.

Preliminary results have shown good agreement between the WRF model, using ACM2, and estimates derived from radar wind profilers. Recently compiled data sets from both the TexAQS II and the Moody Tower TRAMP (TexAQS II Radical Measurement Project) Houston, TX field studies provide an opportunity to further evaluate this scheme in a complex meteorological environment. First, micropulse LIDAR data taken at Moody Tower during the TRAMP experiment was used to determine PBL heights based on aerosol backscatter during September of 2006. Second, mixing heights identified from radiosonde balloon data, courtesy of the University of Houston, are available for August and September of 2006. Lastly, hourly radar wind profiler-based mixing heights created independently by Sonoma Technologies, Inc. are available for August and September of 2006. The above observational data sets will be used to compare the model-derived mixing heights from a 12 km and 4 km horizontal resolution WRF model runs with the ACM2 mixing scheme. These WRF modeling runs were developed by the EPA, and represent the current state of knowledge for the region.

Knowing the accuracy of model-simulated mixing heights is crucial to the continued development of regulatory air quality models. Mischaracterization of the depth and evolution of the PBL and intensity of turbulent transport can lead to severe errors in both meteorological and chemical simulations. PBL processes greatly influence chemical concentrations and reactions within the lowest kilometer of the atmosphere. With the release of a new combined local and nonlocal closure atmospheric boundary layer model, ACM2, and several observational data sets from the TexAQS II field campaign, a comprehensive evaluation is necessary.