Evaluating Aerosol Process Modules within the Framework of the Aerosol Modeling Testbed

Factors that influence predictions of aerosol direct and indirect forcing, such as aerosol mass, composition, size distribution, hygroscopicity, and optical properties, still contain large uncertainties in both regional and global models. New aerosol treatments are usually implemented into a 3-D atmospheric model and evaluated using a limited number of measurements from a specific case study. Under this modeling paradigm, the performance and computational efficiency of several treatments for a specific aerosol process cannot be adequately quantified because many other processes among various modeling studies (e.g. grid configuration, meteorology, emission rates) are different as well. The scientific community needs to know the advantages and disadvantages of specific aerosol treatments when the meteorology, chemistry, and other aerosol processes are identical in order to reduce the uncertainties associated with aerosols predictions. To address these issues, an Aerosol Modeling Testbed (AMT) has been developed that systematically and objectively evaluates new aerosol treatments for use in regional and global models. The AMT consists of the modular Weather Research and Forecasting (WRF) model, a series testbed cases for which extensive in situ and remote sensing measurements of meteorological, trace gas, and aerosol properties are available, and a suite of tools to evaluate the performance of meteorological, chemical, aerosol process modules. WRF contains various parameterizations of meteorological, chemical, and aerosol processes and includes interactive aerosol-cloud-radiation treatments similar to those employed by climate models. In addition, the physics suite from the Community Atmosphere Model version 5 (CAM5) have also been ported to WRF so that they can be tested at various spatial scales and compared directly with field campaign data and other parameterizations commonly used by the mesoscale modeling community. Data from several campaigns, including the 2006 MILAGRO, 2008 ISDAC, 2008 VOCALS, 2010 CARES, and 2010 CalNex campaigns, have been incorporated into the AMT as testbed cases. Data from operational networks (e.g. air quality, meteorology, satellite) are also included in the testbed cases to supplement the field campaign data. The CARES and CalNex testbed cases are used to demonstrate how the AMT can be used to assess the strengths and weaknesses of simple and complex representations of aerosol processes in relation to computational cost. Anticipated enhancements to the AMT and how this type of testbed can be used by the scientific community to foster collaborations and coordinate aerosol modeling research will also be discussed.