Development and Application of a Regional Scale Integrated Meteorology-Atmospheric Chemistry Model


RESEARCH TEAM

Dr. Aijun Xiu         aijun@email.unc.edu
Dr. Rohit Mathur   rmathur@email.unc.edu
Dr. Carlie Coats     coats@emc.mcnc.org
Dr. Adel Hanna      ahanna@email.unc.edu
Ms. Uma Shankar   ushankar@email.unc.edu
Mr. John McHenry  mchenry@emc.mcnc.org

 
 
 

Environmental Modeling for Policy Development
Institute for the Environment
University of North Carolina at Chapel Hill

Bank of America Plaza, CB#6116
137 E. Franklin Street
Chapel Hill, NC 27599-6116

BACKGROUND

 

Current Modeling Approaches

  • Diagnostic approach
  • Prognostic approach
  • meteorological data stored for later use as input to drive separate chemistry/transport model
    •
  • One source of uncertainty derives from the fact that the meteorology, emissions and chemistry are modeled separately and connected "off-line"
  • Due to limitations in computer I/O and storage, meteorological data are output less frequently than internal time-steps
  • Temporal interpolation in CTM
  • Potentially important linkages between atmospheric chemistry and meteorology are neglected, such as the radiative feedback of aerosols

    •  

     
     







    Objectives
    Approach

    MODEL COMPONENTS

    MM5
    SMOKE
    MAQSIP
    MCPL


    Table 1. Components of the Integrated Model


    Feature
    Physical/Numerical Representation
    Reference
    Meteorological Dynamical Calculations MM5 Version 2 Grell et al. (1994)
    Meteorology/ Chemistry Coupling MCPL Coats et al. (1998)
    Tracer Transport and Chemistry    
    Gas-Phase Chemistry Modified CBM-IV or RADM2 Mechanism Gery et al. (1989), Kasibhatla et al.(1997), Stockwell et al. (1990)
    Chemistry ODE Solver Modified QSSA Mathur et al. (1998)
    Advection Bott's Scheme Bott (1993)
    Turbulent Mixing K-theory Chang et al. (1987), Alapaty and Mathur (1998)
    Dry Deposition Resistance theory Wesley (1989)
    Aerosol Dynamics Modal approach Binkowski and Shankar (1995)
    Clouds* Kuo or Kain-Fritsch Kuo (1974)

    Kain and Fritsch (1990)
    Emissions** Both anthropogenic and biogenic emissions  

     

    Figure 1. The schematic diagram of the integrated modeling system.







    APPLICATIONS OF THE INTEGRATED MODELING SYSTEM
    Radiative Feedback of Aerosols
    Case Study with the integrated modeling system

    Figure 2. The MM5 domain with the course and fine grids (108 km and 36 km). The integrated model is run within the fine grid domain.
    Sensitivity tests of the refractive index of aerosols
    The real part (Rr) and the imaginary part (Ri) of the refractive index represent the scattering and absorbing components respectively. Rr = 1.5 - 0.27 * WFRAC and Ri = 0 Rr = 1.5 - 0.27 * WFRAC and Ri = 0.01 - 0.01 * WFRAC Figure 3 shows event-average distribution of fine particle mass simulated by the integrated model at the lowest model level. Also shown in Figure 4 are comparisons of these simulations with IMPROVE observations. Taken together these figures suggest that the model can reasonably well predict the spatial distribution of the fine particle mass though the general trend is toward a slight underprediction. These underpredictions could in part ne due to the limited duration of the simulation (10 days). Figure 5 presents qualitative relationships and comparison of simulated distribution of the net radiative effects of aerosol loading with various aerosol loading parameters (mass, size, and number). A comparison of spatial distribution of these parameters indicate that the net reduction in shortwave radiation reaching the surface is not only related to the aerosol mass loading, but is also dependent on particle size and number density. Figure 6 attempts to elucidate these relationships by presenting the same data in a more concise fashion. In construction these figures, we binned the data based on predefined intervals of the various aerosol-loading parameters and computed averages of each parameters and the average reduction in shortwave radiation reaching the surface for the corresponding grid-cells. Also shown in the figures are the upper and lower values encountered within each bin. These plots further illustrates that aerosol size parameters play an important role in dictating the perturbation in radiation and regional feedback patterns.
     
     


    Figure 3. The predicted event-average distribution of fine particle mass at the lowest model level.


     
     

    Figure 4. The scatter plot of predicted fine particle mass at the lowest model level versus observations from the IMPROVE network.



     
     
     

    Firgure 5. The simulated event-averaged shortwave radiation changes with and without direct aerosol radiative feedbacks and three aerosol loading parameters (mass, size, and number).
     
     
     
     


    Figure 6. Simualted correlations between reduction in shortwave radiation reaching the ground and aerosol loading parameters (mass, size, and number).

    SUMMARY

    ONGOING ACTIVITIES