Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out
Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>

Evaluation Tools for CMAQ Model Performance to IMPROVE Data

VIEWS: 20 PAGES: 16

									  Evaluation Tools for CMAQ
Model Performance to IMPROVE
       Observation Data

    Chao-Jung Chien and Bo Wang
             Regional Modeling Center (RMC)
    Center for Environmental Research and Technology
            University of California at Riverside
IMPROVE
• Interagency Monitoring of Protected Visual
  Environments measurement program first initiated in
  1988.
• Objectives:
   – To establish background visibility levels;
   – To identify chemical species and emission sources
     responsible for man-made visibility impairment;
   – To document long-term trends for assessing progress
     toward the national visibility goal; and
   – With the enactment of the Regional Haze Rule, to provide
     regional haze monitoring representing all visibility-protected
     federal CIAs where practical.
• http://vista.cira.colostate.edu/improve/Default.htm

                                University of California-Riverside, Regional Modeling Center (RMC)
 IMPROVE DATA
• 145 monitoring stations, 110 sites in class I
  areas.
• 49 sites data available within WRAP domain.
• Data Resources
  – Aerosol data and Optical data
  – Both raw data (ASCII) and reconstructed data (Excel
    spread sheets) are available for download at:
    ftp://alta_vista.cira.colostate.edu/DATA/IMPROVE/


                          University of California-Riverside, Regional Modeling Center (RMC)
CMAQ Aerosol Species
• Particles are treated in three lognormal
  subdistributions – modes
  – Aitken mode (i-mode)
  – Accumulation mode (j-mode)
  – Coarse mode
• PM2.5  sum of i, j mode species
• PM10 coarse mode species + PM2.5
• Speciation: see handout (Table 1)
                    University of California-Riverside, Regional Modeling Center (RMC)
CMAQ Visibility
• Extinction coefficient
  – Mie theory; “Bext_Mie”
  – Reconstructed extinction; “Bext_Recon”
     • based on aerosol mass concentration

  bext(1/km) = 0.003*fRH*([(NH4)2SO4]+[NH4NO3]) +
              0.004*[OC] + 0.010*[LAC] + [SOIL] +
              0.0006*[CM]



                        University of California-Riverside, Regional Modeling Center (RMC)
IMPROVE Species
• Reconstructed mass species:
   – Sulfates (SO4), as ammonium sulfates;
   – Nitrates (NO3), as ammonium nitrates;
   – Fine soil (SOIL), sum of several inorganic elements;
   – Fine mass (FM), as measured PM25;
   – Reconstructed fine mass (RCFM); sum of
     (SO4+NO3+LAC+1.4*OC+SOIL)
   – Coarse mass (CM); PM10-PM25
   – MT, as PM10
• Visibility
   – Bext=0.003*fRH*([SO4]+[NO3])+0.004*[OMC]+0.01*[LAC]
     +0.001*[SOIL]+0.0006*[CM]


                               University of California-Riverside, Regional Modeling Center (RMC)
CMAQ vs. IMPROVE Species Mapping
• See attached Handout (Table 2)
• Units:
  – CMAQ: ug/m3 for aerosol species
          1/km for extinction coefficient
  – IMPROVE: ng/m3 for aerosol species




                      University of California-Riverside, Regional Modeling Center (RMC)
Software Package for Post Modeling Analysis

 • The programs can automatically …
    – Extract species info from model output files
      (IOAPI/NetCDF) and observation data file (ASCII)
    – Identify monitoring sites within WRAP domain
    – Compute daily average concentrations for each
      species
    – Match model predictions with measured species
      on days for which measurements are reported
    – Generate model vs. observation plots

                          University of California-Riverside, Regional Modeling Center (RMC)
Software Package for Post Modeling Analysis

 • Results:
    – Processed preformatted ASCII files:
     “AmbientData_Query” and “ModelData_Query”
    – Plots: w/ basic statistic analysis
       • AllSite_AllDay
       • AllDay_OneSite    Regression analysis,
                           Mean Normalized Error and Bias
       • AllSite_OneDay
       • Time Series


                          University of California-Riverside, Regional Modeling Center (RMC)
Software Package for Post Modeling Analysis

 • System requirement:
   – Linux
 • Programs needed:
   – g++, perl, pgf90
   – ncdump, gnuplot (included in the package)
 • README file is included.


                      University of California-Riverside, Regional Modeling Center (RMC)
Software Package for Post Modeling Analysis
•    Steps to run:
    1. Get inputs files ready:
       –   Ambient data file; [Station] [Year] [Jdate]
       –   Monitor station list; [StationID] [Longitude] [Latitude]
       –   Relative Humidity file (Optional; default fRH file is provided)
    2. Edit two run script files
       –   Specify evaluation dates (info provided in Run_Eval.scp)
       –   Check model output files location (verify path in
           Run_ModelData.scp)
       –   Make sure two model output files exist:
           e.g. *CONC.1996185 and *AEROVIS.1996185
    3. Verify species mapping
       –   Define species mapping equations for both ambient and
           model species (file: Species_Mapping)
                                     University of California-Riverside, Regional Modeling Center (RMC)
Hands-on Practice
• Analysis period:
  – July, 1996 of CMAQ simulations
    (Julian dates 183-213)
• IMPROVE database:
  – Processed, Sorted ASCII file:
    “AEROSOL.CONC” or “Recon_Mass”
     • Reprocessed from original 100MB ASCII file and Excel
       spread sheets (reconstructed mass concentration)
  – Compare model predictions vs. measurements for
    SO4, NO3, OC, EC, SOIL, RCFM, PM25, PM10 and
    Bext.

                            University of California-Riverside, Regional Modeling Center (RMC)
AllSite_AllDay




      University of California-Riverside, Regional Modeling Center (RMC)
AllSite_OneDay




      University of California-Riverside, Regional Modeling Center (RMC)
AllDay_OneSite




      University of California-Riverside, Regional Modeling Center (RMC)
Time_Series




    University of California-Riverside, Regional Modeling Center (RMC)

								
To top