Daewon W. Byun

Review of the Governing Equations, Computational Algorithms, and Other Components of the Models-3 Community Multiscale Air Quality (CMAQ) Modeling System

This article describes the governing equations, computational algorithms, and other components entering into the Community Multiscale Air Quality (CMAQ) modeling system. This system has been designed to approach air quality as a whole by including state-ofthe-science capabilities for modeling multiple air quality issues, including tropospheric ozone, fine particles, acid deposition, and visibility degradation. CMAQ was also designed to have multiscale capabilities so that separate models were not needed for urban and regional scale air quality modeling. By making CMAQ a modeling system that addresses multiple pollutants and different spatial scales, it has a “one-atmosphere” perspective that combines the efforts of the scientific community. To implement multiscale capabilities in CMAQ, several issues (such as scalable atmospheric dynamics and generalized coordinates), which depend on the desired model resolution, are addressed. A set of governing equations for compressible nonhydrostatic atmospheres is available to better resolve atmospheric dynamics at smaller scales. Because CMAQ is designed to handle scale-dependent meteorological formulations and a large amount of flexibility, its governing equations are expressed in a generalized coordinate system. This approach ensures consistency between CMAQ and the meteorological modeling system. The generalized coordinate system determines the necessary grid and coordinate transformations, and it can accommodate various vertical coordinates and map projections. The CMAQ modeling system simulates various chemical and physical processes that are thought to be important for understanding atmospheric trace gas transformations and distributions. The modeling system contains three types of modeling components (Models-3): a meteorological modeling system for the description of atmospheric states and motions, emission models for man-made and natural emissions that are injected into the atmosphere, and a chemistry-transport modeling system for simulation of the chemical transformation and fate. The chemical transport model includes the following process modules: horizontal advection, vertical advection, mass conservation adjustments for advection processes, horizontal diffusion, vertical diffusion, gas-phase chemical reactions and solvers, photolytic rate computation, aqueous-phase reactions and cloud mixing, aerosol dynamics, size distributions and chemistry, plume chemistry effects, and gas and aerosol deposition velocity estimation. This paper describes the Models-3 CMAQ system, its governing equations, important science algorithms, and a few application examples. This review article cites 114 references. DOI: 10.1115/1.2128636

The next generation of integrated air quality modeling: EPA's models-3

Abstract The U.S. Environmental Protection Agency is developing a third-generation modeling system, termed Models-3. This paper provides an overview of the concepts behind this effort. The modeling challenge is large and is addressed at two main user groups, regulatory analysts and scientists. New technology and tools from the federal High Performance Computing and Communications Program present an opportunity to effectively address computational constraints and the modeling challenge, simultaneously. Two goals of the Advanced Air Quality Modeling Project are (1) provide an effective decision support system and (2) provide a framework to support the evolvement of models and modeling systems. The information needed for a decision support system is described and its elements are defined. The need to be able to significantly evolve the air quality models is discussed next, followed by the presentation of a general software approach for avoiding modeling system obsolescence. In the final section, key modeling considerations and target capabilities are outlined to show the directions being undertaken to initiate Models-3.

Design artifacts in eulerian air quality models: Evaluation of the effects of layer thickness and vertical profile correction on surface ozone concentrations

Abstract Previous studies on the regional acid deposition model (RADM) have revealed high bias of surface SO2 and O3 concentrations by the model, especially during nighttime hours. Comparison of the RADM results with surface measurements of hourly ozone concentrations from the National Dry Deposition Network (NDDN) sites showed distinct diurnal variations in the model high bias. Here, we investigate what part of this phenomenon is influenced by the coarse vertical resolution of RADM in representing the deposition layer. For certain deposition species in the model, we apply the planetary boundary layer (PBL) similarity theory to predict the high bias of the model results (volume averages) to the surface observations (time series at a point) for the horizontally homogeneous case. Here, we applied the profile corrections to a secondary species O3, which is one of active reacting species even at night especially with NO. However, we attempted to separate the effect of deposition layer thickness from the effects of other horizontal and vertical resolution such as emissions source distribution during the NOO3 titration process for a clearer presentation of our hypothesis. The study shows that there are situations when a considerable portion of the high bias of model O3 concentrations at night is explained by the coarse vertical resolution in the deposition layer. It is shown that the model needs to resolve, at least, the lower half of the PBL in order to predict surface deposition fluxes correctly. Comparison with several NDDN observations shows that for certain NDDN sites the present hypothesis cannot fully explain the models high bias of daily minimum O3. In a companion paper, the effect of emission source distribution in representing the NOO3 titration process will be studied to investigate causes of the model bias further.

Sensitivity of ozone to model grid resolution—I. Application of high-resolution regional acid deposition model

Abstract This paper examines the sensitivity of ozone (O3) predictions to grid resolution in Eulerian grid models. A high-resolution version of the regional acid deposition model (HR-RADM) was developed and applied to simulate O3 formation at different grid resolutions. Horizontal grid-cell sizes of 20, 40, and 80 km were selected for this sensitivity study. Individual meteorological and chemical processes that contribute to O3 and its precursors were further separated and analyzed to determine their importance to O3 formation and the effects of grid resolution on these regulating processes. We first examined the model predictions of O3 maxima and minima at different grid resolutions over several major source areas. The results showed that the coarser-grid model tended to underpredict O3 maxima and overpredict O3 minima over the major source areas, because emission strengths were not as well resolved. Process contribution analyses of O3 over these source areas revealed that grid resolution significantly influences the magnitude of O3 formation and loss processes, especially chemistry and vertical transport. We also compared the process contributions between two different grid resolutions over an equal source area with nearly equal emissions to examine the nonlinearities of processes and their interactions with respect to grid resolution. These comparisons showed that for nonreactive species, the average transport applied to a coarse-grid cell is the same as that applied to the same area at higher resolution. For reactive species, however, the average transport is no longer the same between two different grid resolutions because the transport process interacts closely with chemistry, which is nonlinearly related to grid resolution. As a result, over the same source area, the coarser grid tended to predict more O3 but less NO2 from chemistry and to export more O3 and NO but less NO2 by vertical transport than did the finer grid.

Correcting RADM's sulfate underprediction: Discovery and correction of model errors and testing the corrections through comparisons against field data

Abstract A serious underprediction of ambient sulfate (SO42−) by two comprehensive, Eulerian models of acid deposition, the Regional Acid Deposition Model (RADM) and the Acid Deposition and Oxidant Model (ADOM), was found in the National Acid Precipitation Assessment Program phase of the Eulerian Model Evaluation Field Study (EMEFS) model evaluation. Two hypotheses were proposed to explain the cause of the underprediction in RADM: insufficient SO42− production by nonprecipitating convective clouds and insufficient primary SO42− emissions. Modifications of the RADM cloud and scavenging module to simulate nonprecipitating cumulus clouds better are described in detail. Three contrasting pairs of tests using data from the EMEFS were applied to these hypotheses: source vs downwind regions, mid-summer vs late summer seasons and sunny-dry vs cloudy-wet synoptic types. The SO42− emissions hypothesis, tested by artificially boosting SO42− emissions, fared better than expected but was rejected because of its poor performance on the regional and seasonal contrast tests. The RADM nonprecipitating cumulus modification successfully captured the seasonal and the late summer synoptic contrasts but improvement is still needed for the regional and mid-summer synoptic contrasts.

Uncertainty analysis of environmental models within GIS enviroments

Abstract Information on the uncertainties in results from Geographic Information Systems (GIS) is needed for effective decision-making. Current GISs, however, do not provide this information. Uncertainties in application models used within GIS environments are a major cause of uncertainties in GIS results. To analyze model uncertainty and its propagation, model sensitivity analysis is first performed. This paper discusses techniques for model sensitivity analysis, model uncertainty analysis, and analysis of the propagation of model uncertainty, within the context of environmental models used within GIS. A two-dimensional air quality model, based on the first-order Taylor method, is used to demonstrate these techniques.

An automatic differentiation technique for sensitivity analysis of numerical advection schemes in air quality models

Sensitivity analysis, which characterizes the change in model output due to variations in model input parameters, is of critical importance in simulation models. Sensitivity coefficients, defined as the partial derivatives of the model output with respect to the input parameters, are useful in assessing the reliability of the output from a complex model with many uncertainty parameters. Most existing sensitivity methods, however, have one or more of the following limitations: inaccuracy in the results, high cost in human effort, and difficulty in mathematical formulation and computer program implementation. To overcome these limitations, we are exploring ADIFOR, an automatic differentiation technique for systematically studying sensitivities. One can apply ADIFOR without having an intimate knowledge of the algorithms implemented in a model, so manual preparation of sensitivity code is avoided. In this paper, ADIFORs accuracy and computational efficiency are demonstrated by calculating the sensitivity of concentration to a global perturbation of wind velocity in advection models and comparing this with results from the brute-force method of sensitivity analysis. ADIFOR-generated code can produce exact sensitivity information up to the machine epsilon, and can reduce computer CPU time requirements by up to 57% compared with the brute-force method for a single sensitivity calculation (and the savings increases with the number of parameters). Furthermore, we demonstrate the applicability of ADIFOR to models with a large number of uncertainty parameters by calculating the sensitivity of model output to initial conditions in a two-dimensional advection model.

Contributions of inter- and intra-state emissions to ozone over Dallas-Fort Worth, Texas

Simulation of CMAQ with the high-order direct decoupled method (HDDM) for two 2005 episodes was used to assess the impacts of local emissions and regional transport on ozone concentrations in the Dallas-Fort Worth (DFW) region of Texas. The episodes featured east-northeasterly winds conducive to interstate transport of air pollutants. The study revealed that local, intrastate, and neighbouring state emissions of nitrogen oxides (NO x ) all contributed significantly to daytime ozone in DFW. Local NO x emissions exerted the strongest impact on local ozone, though the impact was highly variable temporally and spatially within the region. NO x emissions from Texas areas outside DFW contributed on average about 10 ppb to daytime DFW ozone. Neighbouring states (Oklahoma, Arkansas, Louisiana, and Mississippi) in total also contributed about 10 ppb to DFW ozone. Anthropogenic VOC emissions from outside the DFW region yielded negligible impact on DFW ozone. DFW ozone is shown to respond more nonlinearly to local NO x than to other NO x emission reductions. The CMAQ-HDDM results indicate that for these episodes, a 4 ppb reduction in average DFW 8 h ozone could be achieved by either a 40% reduction in DFW NO x , a 70% reduction in intrastate NO x , or a 50% reduction in NO x from the four neighbouring states.

Development and Implementation of the EPA’s Models-3 Initial Operating Version: Community Multi-Scale Air Quality (CMAQ) Model

For the last fifteen years, the Office of Research and Development (ORD) of the U. S. Environmental Protection Agency (EPA) has been developing three-dimensional Eulerian based air quality models (AQMs) to study air quality problems, such as urban and regional tropospheric ozone and regional acid deposition. These AQMs simulate comprehensively atmospheric processes such as chemical transformations, transports, and removal of pollutants and their precursors. Model application experience with second generation air quality modeling systems has revealed several shortcomings such as slow execution speed, difficulty in implementing improved science algorithms in the model, and complexity in data exchange among system submodels. Byun et al. (1995) listed some of the shortcomings of the present AQM modeling systems in detail.

US National Air Quality Forecast Capability: Expanding Coverage to Include Particulate Matter

The US National Air Quality Forecast Capability (NAQFC), developed by the National Oceanic and Atmospheric Administration (NOAA) in partnership with the Environmental Protection Agency (EPA), currently provides next-day operational predictions for ground level ozone and smoke for 50 US states. Ozone predictions are produced with the Community Multiscale Air Quality (CMAQ) model driven by NOAA’s operational North American Mesoscale weather forecast Model (NAM); routine verification is conducted with monitoring data compiled by the EPA. Smoke prediction relies on satellite detections of smoke sources, US Forest Service emission estimates, with transport and dispersion simulated by the HYbrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model driven by NAM; routine verification is conducted with satellite observations of smoke. Quantitative predictions of fine particulate matter (PM2.5) are in development. Inventory based simulations using CMAQ with aerosol modules show seasonal biases: overestimating in wintertime and underestimating in summertime. Current testing focuses on including intermittent aerosol sources directly emitted by wildfires and dust storms within the forecast domain; longer-range transport of dust is incorporated through lateral boundary conditions. For example, simulations of trans-Atlantic transport of Saharan dust, injected into the prediction domain, contribute enhanced surface PM2.5 concentrations in the southern US, as observed in surface monitoring. Simulated PM2.5 concentrations are being evaluated with speciated observations in order to improve seasonal biases in predictions. Research on assimilating PM2.5 surface observations shows potential to improve predictions. Furthermore, analysis of discrepancies between observations and model predictions that are produced during assimilation can provide insight on impacts of proposed improvements to PM2.5 predictions.