21
Jim Sullivan Ruth Roberson Steve Irwin aq-ppt5-10

Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

  • Upload
    others

  • View
    5

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

Jim Sullivan Ruth Roberson

Steve Irwin aq-ppt5-10

Page 2: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• What we do now? • Why refine what we do? • Scope of existing quantitative practices • Why geostatistic approaches are preferred • Case Study • Driving question – Guidance or Working Practice? • Thoughts and Comments

Page 3: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• Current Guidance • Hierarchy of choices

• Design Value • Representative Monitor • Alternative approaches

• Design Values are low • Representative monitor guidance reflects elements of a

quantitative analysis • Needs clarification to set expectation • Begs the question

• When does clarification look more like a new approach? • Guidance revision? Working practices memorandum? Other?

Page 4: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• Clarity of purpose • More realistic selection of monitor for development of a

background • Based on available data

• Recognize importance of the most relevant factors • Source and Emission Density are primary drivers • Terrain and meteorology are relevant explanatory components

• Make clear that this approach is for the selection of a monitor • Background can be calculated in a variety of acceptable ways

Page 5: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• Qualitative is meaningless in this situation • Thematic data is not indicative of the conditions present in the modeled or

monitored domains • Quantitative data exists for nearly the entire state (some surrounding

states as well)

• Statistical and Geostatistical approaches are available • Statistical is “aspatial” in that it does not account for distances between

points • This is a key dimension that is omitted from traditional statistical

approaches • Most analysts treat data as normal or log-normal

• This is frequently not appropriate

Page 6: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• Geostatistical approaches are preferred • Accounts for spatial distance between points • Quantitatively address

• Characteristics of distributions • Pattern analysis

• MPCA and other State Agencies have digital data sets • Most (if not all) analysis can be completed in ARCGIS

• Most extensions are license-specific • The proposed Guidance refinement reflects standard geostatistical practices

• Considers statistically “representative” analysis between monitoring locations and modeling domain

• Practical differences can be resolved through additional lines of data • Meteorology, surface roughness, terrain, etc.

Page 7: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• Proposed new construction of a large coal-fired brewery in north-central Minnesota (St. Cloud). • Expected emissions of SO2 are 98 Tons Per Year

• Will be higher after planned expansion • No ambient SO2 air quality monitor is present in the area • Design value is too high for the proposers comfort • Project proposer wants to use a more refined approach based on

monitoring data

First Task: Find a “representative” monitor

Page 8: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• What does a “representative” monitor look like?

• What data do I use? • What methods do I need to use to conduct

the analysis? • How do I present the findings?

Page 9: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• What does a “representative” monitor look like? • Statistical and Practical Significance is relevant

• Focus is on point sources • Recognize other sources may be relevant

• Comparison must be made with same size geographic area • 50 kilometer radius from the source under review and selected monitor

• Compositionally, geographic areas must include same/similar type of setting or context • Spatial location of emission sources • Emission inventory data

• Especially “significant” sources in the inventory • Meteorology and terrain

• This is the key feature of the analysis!

Page 10: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• What data do I use? • Existing MPCA Nearby Source Tool has most current data • Spatial and emission

• Will likely augment with an emission density layer to facilitate analysis

• Meteorology • Terrain

Page 11: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• What methods do I need to use to conduct the analysis?• ARCGIS• First step – compare monitoring site(s) with modeling domain.

Visual inspection of an emission density map (tpy/km) shows that there are only a few sites that match with the modeling domain for the brewery:

• Rochester• Metro Region

(Potentially high)

Page 12: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the
Page 13: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the
Page 14: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• Next step – compare distribution and patterns of Rochester & Metro to Brewery

• Distance between center and points • Band Correlation Statistic

• Compares sites through covariance and correlation

• Spatial Autocorrelation • Compares each site to theoretical distribution to evaluate whether a site is

random or clustered

• If it looks favorable – conduct remaining analysis • Meteorology • Terrain

Page 15: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

Rochester Ellenbecker Bros. Brewery

Metro Site

Page 16: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

1 = Rochester; 2 = Metro Monitor; 3 = Ellenbecker Bros. Brewery Site.

Page 17: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• Similar comparison with meteorology and terrain

Page 18: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• How do I present the findings?

•Submit as a memorandum with the protocol •Follow the stepped format • Include summary of analysis with relevant tables and figures

•See case study for example

Page 19: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the

• Feedback is critical • Guidance, Working Practices Memorandum or Work Group

• Each have merit

• Recognize that most firms have GIS talent in-house • May not have all extensions needed

• Analytical tasks in this Case Study are fundamental GIS tasks • Analysis only supports the selection of a monitor – does not

account for a background concentration • Separate analysis

• Current data set is point-source based • Can be augmented to account for other sources • Typically as a modifier

Page 20: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the
Page 21: Jim Sullivan Ruth Roberson Steve Irwin · Minnesota Pollution Control Agency - Jim Sullivan, Ruth Roberson, Steve Irwin \(B. Tegdesch PDF only\) Subject: Presentations given at the