Model Descriptions

Regional, hemispheric, and global models exist that are capable of simulating oxidants and aerosol mass. Several models are currently being used for ACP research; a list is provided in the breakout session report titled "Scientific Issues in Oxidant and Aerosol Modeling." The capabilities of a four of the models are described here.

GChM-O (Global Chemistry Model driven by observation-derived synoptic meteorological data); C. Benkovitz and S. Schwartz, BNL. GChM-O is a three-dimensional (3-D) Eulerian transport and transformation model for sulfate. The model represents emissions of the several sulfur species, chemical conversion of SO₂ to sulfate by H₂O₂ and O₃ in the aqueous-phase of precipitating clouds and by OH in the gas phase, chemical conversion of DMS to SO₂ and MSA by OH, and removal by wet and dry deposition. Reactions in nonprecipitating clouds are currently not represented in the model. The model includes seasonally dependent concentrations of OH, H₂O₂, and O₃. The meteorological data are the six-hour forecast fields obtained from the European Centre for Medium-Range Weather Forecasts (ECMWF); use of these observation-derived meteorological data allows comparison with observations at specific times and locations. Model resolution is 1.125^o in the horizontal and 15 levels from the surface to 100 hPa in the vertical; model time step is one hour with results captured every six hours. GChM-O has been used in subhemispheric simulations; material advected into the model domain was assigned representative background concentrations and is carried as a separate variable. In order to permit interpretation of sources of sulfate, the SO₂ and sulfate variables are distinguished as to origin of material (biogenic or anthropogenic), geographical region of origin, and formation mechanism (primary and secondary by gas-phase or aqueous-phase oxidation). The model is initialized with the concentration of all species set to zero.

MOZART (Model of Ozone and Related Chemical Tracers); G. Brasseur and X. Tie, NCAR. MOZART is a 3-D chemical-transport model (CTM) driven by global winds, temperature, humidity, and cloud fields provided by the NCAR Community Climate Model version 3 (CCM3). The version of the model currently in use computes the time history of 50 chemical species on the global scale from the surface to the upper stratosphere. In its present configuration, the model is run with a spatial horizontal resolution identical to that of the standard CCM3 (triangular truncation at 42 waves) with a corresponding numerical grid of 64 gaussian latitudes and 128 equidistant longitudes (corresponding to a 2.8^o horizontal resolution). In the vertical, the model uses a hybrid coordinate with 25 levels extending from the surface to 5-mbar level. The numerical timestep for both transport and chemistry is 20 min, so that the diurnal evolution of chemical species is explicitly represented. MOZART is run "off line'' from CCM3, with dynamical variables (e. g. wind components, pressure, temperature, water vapor, convective mass fluxes, diffusion parameters, cloudiness) provided every three hours from pre-established history tapes. A version using assimilated winds will also soon become available. The model accounts for surface emissions of chemical compounds, advective transport, convective transport, diffusive exchanges in the boundary layer, in-cloud wet scavenging, and surface dry deposition. The chemical scheme with its 50 species includes approximately 130 chemical and photochemical reactions, as well as wash-out processes for approximately 10 soluble species. The heterogeneous conversion of NO₃ and N₂O₅ into HNO₃ on the surface of sulfate aerosols is crudely parameterized using pre-calculated sulfate concentrations. The NO_x sources represented in this model include the contributions of fossil fuel combustion, biomass burning, soil emissions, lightning, and aircraft emissions.

IMAGES (Intermediate Model of the Annual and Global Evolution of Species); G. Brasseur and X. Tie, NCAR. IMAGES simulates chemical processes, wet and dry deposition, and surface emissions as in MOZART. The resolution of the model is 5^o by 5^o in latitude and longitude; there are 25 levels, with three of them located in the stratosphere below 50 mb. The model has oneday time step, except during the first three days of each month, during which a full diurnal calculation is performed with onehour time step. The wind components used to calculate the transport of long-lived species are monthly mean climatological values derived from the analyses of ECMWF. The effect of temporal wind variability (on time-scales shorter than one month) is therefore not explicitly taken into consideration in the advection term of the transport equation. However, large-scale mixing associated with wind variability is parameterized by a diffusion process, which is a function of the observed wind variance. The additional transport components (subgrid processes) which are included in the model are turbulent mixing in the planetary boundary layer and vertical transport associated with convection process. Convection is represented using a probabilistic parameterization.

Models at the State University of New York at Albany; W.-C. Wang et al. The University of Oslo's 3-D CTM has been used in ACP efforts to study the impact of aircraft emissions in years 1992, 2015, and 2050 on the distribution of gases like NO, NO₂, OH, O₃, and CH₄. Changes in the background distribution due to changes in surface emission of pollutants were also included in the calculations. These calculated changes in O₃ and CH₄ were then used in CCM3 to calculate changes in radiative forcing. The climate parameters needed in the radiative forcing calculations are taken to be the 1992 mean distributions of temperature, moisture, clouds, and surface albedo simulated in the AMIP runs in which the observed monthly SSTs were used.

Some Recommendations

Rigorous evaluation of model results has not been possible due to the lack of observational data. ACP should develop "infrastructure" to compile database/s of observational data suitable for model evaluation. Such an effort would involve defining the needs of models for evaluation data for long-term simulations and simulations for specific times and places.

ACP should develop a means to use global and regional models in "nested" mode to help evaluate model results and to help analyze field experiments.

Aerosol microphysics (size distribution, chemical composition, etc.) must be added to aerosol models to allow proper evaluation of climate forcing by aerosols.