Optimizing and Streamlining Data Collection and Analysis for Project Level Air Quality Modeling
State DOTs would benefit from guidance that goes beyond the norm on how to conduct modeling for project-level air quality analyses and addresses which modeling inputs are the most cost-effective and important to focus on in terms of improving accuracy and reducing uncertainty in modeling results. Depending on funding levels, the study may also assess the question of proportionality, that is, how accurate and precise the modeling inputs "need" to be in order to meet the specific regulatory requirements and, conversely, not require more accuracy, level of detail and precision in modeling inputs than are really needed to meet the specific regulatory objectives and application. The study would focus on emission and dispersion modeling (for which traffic would be an input) for particulate matter (PM) and mobile source air toxics (MSATs), and may optionally also address carbon monoxide (CO).
Key tasks may include: 1. Document the modeling inputs needed to run EPA emission and dispersion models for project-level air quality analyses for both purposes of transportation conformity and NEPA. Also document the corresponding modeling outputs needed to show compliance for each pollutant with applicable regulatory requirements and guidance, i.e., the units and expected or typical number of significant digits. 2. Review and document common sources of data for each of the modeling inputs, and the typical time and cost for obtaining or generating those inputs. Depending on the budget available for this study, the cost-estimate may be quantitative (preferred) or qualitative (e.g., high, medium or low cost). 3. Conduct a sensitivity study for the modeling chain for each pollutant. For MSATs, the focus would be on emission modeling inputs and outputs. For PM and (if funded) CO analyses, the focus would be on modeling inputs and outputs for the emission-dispersion modeling chain. Rank the inputs according to sensitivity, from the most sensitive to the least sensitive. 4. For each pollutant, use the results from the previous tasks to identify which modeling inputs and sources of data contribute most cost-effectively to reducing the noise (modeling uncertainty stemming from modeling inputs) in the modeling results (emissions for MSATs, and concentrations for PM and, if funded, CO). To the extent feasible, include consideration of default modeling inputs as a baseline. 5. OPTIONAL: Assess proportionality to the extent feasible, i.e., the concept of limiting the level of detail and number of significant digits to just that needed to meet the specified regulatory requirements. For each pollutant, use the results from the previous tasks and additional sensitivity tests as needed, identify limits to how precise the modeling inputs need to be (in terms of number of significant digits) to generate modeling output that would "just" meet the regulatory requirements in terms of number of significant digits for project-level analyses for each pollutant. 6. Draft recommendations for best practices for data collection and analysis on the basis of costs and cost-effectiveness and (if funded) proportionality. 7. Prepare draft and final reports.
State DOTs would benefit from guidance that identifies which emission and dispersion modeling inputs are the most important and cost-effective for modelers to focus on for project-level air quality analyses. State DOTs would also benefit from guidance on the level of detail or degree of accuracy needed in those inputs to just meet the regulatory need (number of significant digits). Overall, the results from this proposed study would allow modelers for state DOTs to best determine how to prioritize the limited resources available for data collection and analysis, and so help optimize the modeling process for project-level air quality analyses.
May 8, 2017
- items posted in the last 7 days (30 days for TERI)