Grantee Research Project Results
2012 Progress Report: Applying Data Assimilation and Adjoint Sensitivity to Epidemiological and Policy Studies of Airborne Particulate Matter
EPA Grant Number: R833865Title: Applying Data Assimilation and Adjoint Sensitivity to Epidemiological and Policy Studies of Airborne Particulate Matter
Investigators: Stanier, Charles , Oleson, Jacob J. , Carmichael, Gregory R. , Field, R. William , Krewski, Daniel , Kumar, Naresh
Current Investigators: Stanier, Charles , Krewski, Daniel , Carmichael, Gregory R. , Kumar, Naresh , Field, R. William , Oleson, Jacob J.
Institution: University of Iowa , University of Ottawa
EPA Project Officer: Chung, Serena
Project Period: February 1, 2009 through January 31, 2013 (Extended to January 31, 2014)
Project Period Covered by this Report: February 1, 2012 through January 31,2013
Project Amount: $899,401
RFA: Innovative Approaches to Particulate Matter Health, Composition, and Source Questions (2007) RFA Text | Recipients Lists
Research Category: Particulate Matter , Air
Objective:
Source-resolved fine particulate matter (PM) concentrations are needed at high spatial and temporal resolutions for epidemiological studies aimed at identifying more- and less-harmful types of PM. Building on recent advances in air quality modeling, data assimilation, and satellite remote sensing, we aim to develop improved estimates of daily source-resolved PM2.5 at 36 km spatial resolution for major U.S. cities, and at 4 km for Chicago from 2001-2004. The estimates will be made available for public download by other researchers, and used in pilot-scale epidemiological studies to demonstrate one of the many applications of source-resolved PM2.5 at greater spatial-temporal resolutions.
Project objectives are to:
- Evaluate data assimilation (DA) methods for combining the “ground truth” of observations, especially chemically-speciated filter measurements, with the high spatial-temporal resolution of source-oriented air quality models;
- Demonstrate the utility of source-resolved PM2.5 in epidemiological studies and develop best practices for the use of assimilation in conjunction with health studies; and
- Demonstrate the utility of target-oriented and adjoint sensitivity methods for determining the spatial representativeness of samplers, and for visualization of relationships between user-defined air quality/health targets and spatially-resolved emissions.
Progress Summary:
Progress on the project “Applying Data Assimilation and Adjoint Sensitivity to Epidemiological and Policy Studies of Airborne Particulate Matter” can be divided into satellite data assimilation, large-scale CMAQ model runs, adjoint development, and linking air quality model output to epidemiological analysis. To date, results include:
- By using optimized settings for data assimilation of CMAQ PM2.5 and MODIS aerosol optical depth, a domain-wide average improvement in fractional error from 1.2 to 0.97 at IMPROVE monitoring sites is demonstrated, and from 0.99 to 0.89 at STN monitoring sites. Somewhat larger improvements to fractional bias were observed. However, for 38% of the month-region combinations, MODIS OI degraded the forward model skill, due to biases and outliers in MODIS AOD.
- Satellite AOD is most predictive of PM2.5 when matched to PM2.5 station data within an eight to twelve hour time frame, and correlation of PM2.5 and AOD decreases (using Chicago, IL data) when separation distance between the AOD centroid and the monitoring station is greater than 6 km.
- An optimized WRF setup has been identified for the project from adapting WRF configurations from the Iowa DNR/LADCO, SESARM, OTC, and MANEVU comparative analysis.
- Current large-scale CMAQ model run performance statistics for PM2.5 range from -0.7 to +3.8 μg/m3 (mean bias), 5.6-6.9 (mean error), -41 to +25 (fractional bias), and 53-62 (fractional error), evaluated nationally for each of four seasons. They fall into the excellent (fall), good (spring, winter), and average (summer) categories of Morris, et al. (2005). Underprediction characterizes summer, while overprediction is most prevalent in winter.
- Comparing PM2.5 concentrations in the model versus measured values in the ACS cohort shows that a vast majority of the areas are predicted to within a 2:1 envelope, but there are areas where model overprediction is somewhat severe (mostly in the Midwestern U.S.).
- Replacement of default CMAQ 4.7.1 aqueous chemistry routine with a version based on the Kinetic PreProcessor (KPP) should promote efficient calculation of the model adjoint, but numerical problems have not yet been fully addressed. KPP generates Fortran codes that solve the set of chemical reactions using the Rosenbrock solver. The use of KPP is anticipated to lead to higher accuracy and to the efficient generation of an automated adjoint from KPP.
- Collaboration with University of Ottawa co-investigators (Krewski / Turner), and Michael Jerrett and Bernie Beckermann at UC Berkeley, continues. We now have completed development of translation tables that allow association of the ACS cohort members to individual model grid cells and are commencing the epidemiological analysis.
Future Activities:
The next reporting period is the final reporting period and our forward model results, assimilated model results, and adjoint/target-oriented model proof of concept will be presented and published as results become available.
Journal Articles on this Report : 1 Displayed | Download in RIS Format
Other project views: | All 26 publications | 9 publications in selected types | All 9 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Porter AT, Oleson JJ, Stanier CO. On the spatio-temporal relationship between MODIS AOD and PM2.5 particulate matter measurements. Journal of Data Science 2014;12(2):255-275. |
R833865 (2012) R833865 (Final) |
Exit Exit |
Supplemental Keywords:
air, health effects, human health, modelingProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.