Application%20of%20Multiple%20Models%20to%20Simulation%20Fine%20Particulate%20in%20the%20Southeastern%20US - PowerPoint PPT Presentation

About This Presentation
Title:

Application%20of%20Multiple%20Models%20to%20Simulation%20Fine%20Particulate%20in%20the%20Southeastern%20US

Description:

Denis McNally and T.W. Tesche. Alpine Geophysics, LLC. Gail Tonnesen. UC Riverside ... 36 km and 12 km. January 2002 and July 1999 Episodes. CMAQ and CMAQ-MADRID ... – PowerPoint PPT presentation

Number of Views:54
Avg rating:3.0/5.0
Slides: 15
Provided by: GregYa9
Category:

less

Transcript and Presenter's Notes

Title: Application%20of%20Multiple%20Models%20to%20Simulation%20Fine%20Particulate%20in%20the%20Southeastern%20US


1
Application of Multiple Models to Simulation Fine
Particulate in the Southeastern US
  • Ralph Morris and Bonyoung Koo
  • ENVIRON International Corporation
  • Denis McNally and T.W. Tesche
  • Alpine Geophysics, LLC
  • Gail Tonnesen
  • UC Riverside
  • National RPO Modeling Meeting
  • Denver, Colorado
  • May 24-25, 2004

2
VISTAS Phase I Modeling Multi-Models
  • July 1999 and July 2001 Episodes
  • CMAQ and CAMx
  • 36 km and 12 km
  • January 2002 and July 1999 Episodes
  • CMAQ and CMAQ-MADRID
  • 36 km only (CMAQ-MADRID in progress)
  • Operate Models in a Consistent Fashion
  • Same horizontal and vertical grid
  • CMAQ-to-CAMx emissions, IC and BC processors
  • Different met interface programs (MCIP and
    MM5CAMx)

3
VISTAS Phase I Modeling Multi-Models
  • Evaluation for monitoring sites in VISTAS states
  • IMPROVE, CASTNet, SEARCH, STN (July 2001 only)
  • Soccer Goal plots of summary performance
  • Plot Fractional Bias versus Fractional Error
  • Consider two bias/error soccer goals
  • ?15/35 -- 1-hour ozone performance goal
  • May exceed measurement technology
  • 50/75
  • Not suggested as performance goal, used to frame
    and intercompare performance across models and
    species
  • Always box bias/error by ?100/200
  • If performance statistics exceed these values
  • then set equal to them in plot

4
SO4 CMAQ 36 km Bias/Error
5
SO4 CAMx 36 km Bias/Error
6
NO3 CMAQ 36 km
NO3 CAMx 36 km
7
Carbon (TCM, EC and OC) CMAQ 36 km
8
Carbon (TCM, EC and OC) CAMx 36 km
9
SO4 Hourly Timer Series _at_ SEARCH
Observed (red) CMAQ (blue) CAMx (purple)
? Jefferson St. (Atlanta)
? Yorkville (rural)
10
SO4 and TCM Hourly Timer Series _at_ SEARCH
Centreville, AL Site (rural)
? Sulfate (SO4)
? Total Carbon Mass (TCM)
Observed (red) CMAQ (blue) CAMx (purple)
11
TCM Hourly Time Series _at_ SEARCH CMAQ performance
issue partly due to Kz_min 0.1 m2/s
? Yorkville, GA (rural)
? Oak Grove, MS (rural)
Observed (red) CMAQ (blue) CAMx (purple)
12
TCM Hourly Time Series _at_ SEARCH CMAQ Kz_min
Sensitivity Tests
? Yorkville, GA (rural)
? Oak Grove, MS (rural)
Observed (red) Kz_min 0.1 m2/s Kz_min 1.0 m2/s
13
VISTAS Multi-Model Evaluation Conclusions
  • CMAQ performed better for SO4, whereas CAMx
    performed better for Carbon
  • MM5CAMx currently being updated to better deal
    with clouds and precipitation
  • Both models had NO3, Soil and CM performance
    issues
  • CMAQ lead model coming into Phase I and SO4
    performance was quite encouraging so was selected
    as primary model for Phase II
  • CMAQ-AIM results pending
  • Optional (currently unfunded) VISTAS Phase II
    task to apply alternative models

14
VISTAS Multi-Model Evaluation Conclusions
  • Once CMAQ is set up, 95 of the work for running
    CAMx is also done (and vice versa)
  • CMAQ-to-CAMx emissions, IC/BC Interface Software
    makes these inputs easy to prepare
  • MM5 used by both models
  • Also have CAMx-to-CMAQ emissions processor
  • Advantages of running multiple models
  • Corroborative analysis
  • Diagnostic analysis
  • Additional features available
  • e.g., for CAMx two-way nesting, sectional PM,
    alternative science algorithms, probing tools
    (PSAT/OSAT, DDM and Process Analysis)
Write a Comment
User Comments (0)
About PowerShow.com