CAMx Training for CAPIA Using CAMx for Ozone Modeling

1
CAMx Training for CAPIAUsing CAMx for Ozone
Modeling

• Greg Yarwood Ph.D.
• ENVIRON
• Novato, California 94945
• gyarwood_at_environcorp.com
• April 15-19, 2002

2
Typical Problem

• One or more ozone monitors exceeding standard
• 1-hr ozone gt 124 ppb
• 8-hr ozone gt 84 ppb
• Impacts of ozone transport must be considered
• background ozone for problem area
• high regional background
• transport between high emission areas
• impact of problem area on downwind receptors
• Use nested-grid regional model, like CAMx

3
Sequence of Steps

• Conceptual model and episode selection
• Modeling protocol
• Base case and performance evaluation
• Future year control strategies
• Demonstrating attainment

4
Conceptual Model and Episode Selection

• Conceptual model based on ambient data
• conditions that lead to ozone problems
• air quality data
• meteorology
• emissions
• basis for model performance evaluation
• Episode selection
• episodes representing all types of conditions
  identified in conceptual model

5
Modeling Protocol

• Required by US air quality regulations
• Formal document governing modeling process
• models (air quality, meteorology, emissions,)
• data sources
• evaluation procedures
• participants and responsibilities
• Builds consensus among stakeholders
• regulators, industries, communities,
  environmentalists
• the process may keep you out of court!

6
Base Case and Performance Evaluation

• Compare model results to ambient data
• ozone concentrations
• precursor concentrations (VOCs, NOx)
• precursor ratios (VOC/NOx ratios)
• ozone/precursor relationships (ozone/NOz)
• Statistical and Graphical evaluations
• normalized bias, normalized error, etc.
• isopleth plots, time-series
• Diagnostic and Sensitivity tests

7
Statistical Evaluation

• Example from Los Angeles 1997 test case
• All statistical measures meet goals

8
Ozone Isopleth Plot
9
Ozone Time Series Plot
10
Diagnostic and Sensitivity Tests

• Understand what the model system responds,
  re-evaluate performance, compare to conceptual
  model
• Diagnostic tests
• altered initial and boundary conditions (e.g.,
  clean)
• no anthropogenic emissions
• alternate meteorology (e.g., MM5 options)
• Sensitivity tests
• alternate emissions, inventory bias?
• sensitivity to VOC and/or NOx

11
Model Tuning

• Right answer for wrong reasons
• compensating errors in model components
• e.g., VOC emissions and mechanism reactivity
• Inadvertent tuning
• choosing the best outcome from many
  sensitivity/diagnostic tests may be tuning
• Dont just rely on statistical measures
• Do look for qualitative agreement with conceptual
  model

12
Ozone Control Strategies

• Evaluate for future year
• time to implement control measures
• assurance that air quality will be maintained
• Estimate emissions growth
• population and activity (e.g. VMT) growth
• new industries (permitting process?)
• changing landuse (urban growth, agriculture)
• Impact of control measures
• specific measures by source category

13
Demonstrating Attainment

• Develop control measures to attain ozone air
  quality objectives
• How to use the model?
• Absolute
• US EPA 1-hr ozone attainment test
• all modeled grid cells lt 124 ppb
• Relative tool
• US EPA 8-hr ozone attainment
• scale current observed ozone by model estimated
  percent improvement