Model Performance Zone Discussion

1
Model Performance Zone Discussion

• Discussion Leader Jeff Underhill
• NHDES
• Inter-RPO Modeling Meeting
• Denver, CO
• May 25-26, 2004

2
Why Model Performance Zones?

• Its a huge task to ask the models to perform
  optimally on a national basis for a period of 365
  days
• Some of the model performance results presented
  so far may be compromised to some degree by
  regions of poor performance
• Some of those areas may not be critical to the
  success of the modeling effort

3
Model Performance Zones

1. Model Performance Zones are designed to ease the
   application of achieving sufficient model
   performance in discrete portions of the modeling
   domain without (yet) taking-on the task of
   optimal performance in all locations at all
   times.
2. Zones should be considered for large-scale
   regional haze, ozone, and PM modeling as well as
   for corresponding meteorological modeling.

4
Options for Zones

1. None require optimal performance everywhere and
   at all times
2. State-by-state
3. By RPO
4. By Class I areas or nonattainment areas
5. By common airsheds

5
Issues

• Option for focusing performance during critical
  periods
• Codes to be developed to automate the application
  of performance zones
• How should targeting performance in one area be
  resolved with performance in other areas?
• Targeted performance for local zones at the
  expense of other zones may lead to result
  disagreement between RPOs

6
Original Zones Used by WRAP
7
Original Proposal for National Zones
8
Original Revised by CENRAP
9
DRI Zones for Haze
10
IMPROVE Monitor Clusters
11
Proposed Merged Concept