1 / 33

PBT Monitoring Workshop An Analysis to Refine the PBT Monitoring Strategy Objectives (MDN)

PBT Monitoring Workshop An Analysis to Refine the PBT Monitoring Strategy Objectives (MDN). Sponsored by: EPA, USGS, CDC, NOAA, CEC at Sheraton Capital Center Hotel Raleigh, North Carolina April 22-24, 2002 presented by: Steven M. Bortnick. Acknowledgments.

Download Presentation

PBT Monitoring Workshop An Analysis to Refine the PBT Monitoring Strategy Objectives (MDN)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. PBT Monitoring WorkshopAn Analysis to Refine the PBT Monitoring Strategy Objectives (MDN) Sponsored by: EPA, USGS, CDC, NOAA, CEC at Sheraton Capital Center Hotel Raleigh, North Carolina April 22-24, 2002 presented by: Steven M. Bortnick

  2. Acknowledgments PBT Monitoring Strategy Writing Team Members Russ Bullock, EPA Basil Coutant, Battelle Xiaoling Zhang, Battelle 12-3 A Wkshp.ppt

  3. Goals of the Analysis • Compare mercury deposition network (MDN) to draft PBT monitoring strategy objectives • Refine draft objectives • Aid development of objectives and options • Provide example of model-monitor “integration” • Demonstrate value of “shared data system” 12-3 A Wkshp.ppt

  4. Mercury Deposition Network (MDN) • Weekly concentrations of total mercury in precipitation • Information on spatial and seasonal trends in mercury deposited to sensitive receptors (e.g., surface waters) • Over 50 sites in operation during 2000 (sites anticipated to operate for a minimum of five years) • Managed at the NADP Coordination Office (many collaborating Federal, State, & Local agencies) 12-3 A Wkshp.ppt

  5. Analysis Scope and Limitations • This analysis is NOT a critique or indictment of MDN. In contrast, the choice of MDN for analysis highlights its potential utility. • This analysis is still in draft form and therefore subject to change. It is presented for illustrative purposes. • This analysis relies on model predictions, treating them as some measure of truth. 12-3 A Wkshp.ppt

  6. Draft PBT Monitoring Strategy:Goals, Objectives, and Issues • Goals: (1) Discern trends (2) Evaluate risk management actions • Objectives: • Collect data at geographic-climactic regions (“nested”) • Use representative, probabilistic sampling • Leverage existing programs • Issues: • How to target sub-populations using probability sampling • How to leverage programs not designed probabilistically 12-3 A Wkshp.ppt

  7. Project Data (Hg Wet Deposition Case Study) • Wet deposition model predictions • Regional Lagrangian Model of Air Pollution (RELMAP) • Annual averages (μg/m2, 1989 MET, 1995 Emissions) • Constant degree grid (roughly 40*40 km2) • Provided by Russ Bullock • Locations of active MDN sites • Annual precipitation totals (inches, 2000, NCDC) • Population totals (1990 Census adjusted to 2000) 12-3 A Wkshp.ppt

  8. Stratified Probability-Based Sampling • Why probability-based sampling? • Cheaper than a “census” • yields unbiased, representative estimates • Why stratify? • Target geo-regions • Improve national-level precision • Know region-specific precision • Convenient to administer • How best to stratify? 12-3 A Wkshp.ppt

  9. Stratifying by Combined EPA Regions:monitoring objective impact on site allocation • Consider different population objectives: • Total U.S. land mass (lower 48 states in this case) • Land mass near major water bodies (approximated via precipitation) • Total number of people • Consider optimal site allocation in each case: • Proportional to product of variability & sub-population (stratum) total • Maximizes national-level precision • How does site allocation change depending on “population” objective? 12-3 A Wkshp.ppt

  10. Optimal ‘Regional’ Allocation

  11. In Summary … • Population objective impacts regional allocation, e.g., • Total population objective: ~17% of sites in EPA Region 9 • Total precipitation objective: ~6% of sites in EPA Region 9 • Population objective has little affect on national-level precision • Regional estimates are less precise than national estimates, e.g., ~13% CV nationally versus ~23% - 43% CV regionally under a 50-site network. 12-3 A Wkshp.ppt

  12. Evaluate MDN Siting vs. Stratified Probability-Based Optimal Allocation • Bias • Percent bias [(MDN - Truth)/Truth]*100 • Region-specific, National • Precision • Percent CV (side-by-side) • Region-specific, National • Overall Accuracy (mean square error=variance+bias2) • Percent efficiency [(optimal / MDN)*100] • Region-specific, National 12-3 A Wkshp.ppt

  13. Conclusions of the Analysis • MDN could be leveraged for PBT trends goal, but: • Must caveat any “national” interpretation or extrapolation due to lack of probability-based design. • At ~20% point-in-time CV, smaller or short-term trends would be difficult to discern as statistically significant. • MDN exemplifies value of “shared data system”: • 53 independent, distinct sites • 30 funding agencies and 34 operating agencies • In aggregate, provides national & geo-regional information 12-3 A Wkshp.ppt

  14. Analysis Conclusions (cont’d) • Stratification is recommended for targeting geo-regions and improving national-level precision • Ecoregions and state clusters are only marginally better stratification variables than EPA Regions • Optimal allocation is most efficient for national-level precision, but region-specific concerns should not be ignored (e.g., consider targeted over-sampling) 12-3 A Wkshp.ppt

  15. Analysis Conclusions (cont’d) • As expected, MDN over-samples the eastern U.S. compared to several potential PBT “population” objectives • Not as expected, eastern MDN sites (bulk of network) appear to be spatially distributed in a relatively unbiased manner • MDN remains critical for leveraging, but not ideal; so either … • re-consider PBT objectives (e.g., target high deposition areas); or • provide MDN options (e.g., more western sites) • Multi-media integration can be facilitated through careful consideration of objectives. 12-3 A Wkshp.ppt

More Related