Impact/Purpose:

Major Research Objectives include:

1) Statistical Combination of Environmental Data (e.g. air monitoring data, numerical model output, and satellite data). Develop statistical models and software to significantly improve the characterization of important air quality (PM2.5, O3, Hg, and air toxics) gradients for daily and weekly time periods. Modeling results from this effort will contribute to a better understanding of long-range pollution transport, improved validation of numerical models, accurate delineation of pollution non-attainment areas, and accurate input for modeled linkages to public health data. Sensitivity analyses will consider model runs over different boundary conditions, emission scenarios, and spatial resolutions. Work will be coordinated through the NERL/EMAD long-term research project.

2) Air quality - public health outcome relationships. Develop hierarchical relationships of pollution and public health outcomes adjusted for meteorology and socio-economic factors at individual U.S. cities and across broad regional areas. Through our collaborative research with CDC and state partners, FY05/06 work will focus on establishing linkages between air quality and hospital data collected in Wisconsin.

3) Estimate Temporal Trends in Air Pollution. Provide reliable trend information to the Clean Air Markets Division and the OAQPS for inclusion in their periodic reports on improvements in air quality and deposition in the US. Currently, we are developing statisticl models to estimate emission-related trends in ozone and nitrogen species. As more mercury deposition/concentration become available, our attention will shift to quantifying trends in these variables from 2000-present.

4) Provide guidance for EPA's National Air Monitoring Strategy. With recent reductions in air monitoring budgets, EPA needs to consider optimal approaches for reducing the size of existing air monitoring networks, while still maintaining the ability to reliably quantify non-attainment areas. This effort will develop new iterative algorithms to optimize network design for large-scale air and deposition monitoring networks to optimize network design.

5) Statistical Techniques for Modeling Spatial to Spatial Relationships between Land-use and Water Quality. Provide an assessment of different approaches for aggregation or analysis of water quality data that can subsequently be used in evaluating the relationship between landscape parameters and water quality in large rivers.

6) Statistical Center activities. Coordinate periodic meetings of NERL statisticians, sponsor conference in Environmental statistics on research in 2, and offer on-site training in statistical topics of common interest to NERL scientists.

Description:

EPA's Center of Excellence (COE) for Environmental Computational Science is intended to integrate cutting-edge science and emerging information technology (IT) solutions for input to the decision-making process. Complementing the research goals of EPA's COE, the NERL has initiated a new initiative in environmental statistics to demonstrate how the use of statistics can provide better estimates of uncertainty for making more informed environmental decisions. The NERL statistical initiative will identify major near-term and long-term statistical needs confronting NERL science programs, and perform research to address these challenges.The initiative will focus on reinforcing and forming new collaborative science partnerships within NERL and with other EPA offices, academia, and national laboratories to create a new synergy of shared ideas and meshing of common statistical interests. The timeliness and importance of this initiative are directly related to the ever-increasing need for a better understanding of environmental risks and the impact of human actions on the environment.

Record Details:

Record Type:PROJECT

Start Date:10/01/2004

Projected Completion Date:10/01/2010

OMB Category:Other

Record ID: 137230

Show Additional Record Data

Project Information:

Progress :

Under the Public Health Air Surveillance Project (PHASE), developed methodology for statistically melding daily CMAQ ozone and fine particulate output in 2001 with monitoring data from EPA's FRM PM2.5 monitoring network and the NAMS/SLAMS national ozone monitoring network. Daily spatial surfaces were developed for the Northeast and Midwest U.S. and have been delivered to the CDC and state partners for use in their efforts to link air quality and public health outcomes.

Modeled the spatial association between fine particulate matter and natural mortality counts for June, 2000 for all counties in the conterminous U.S. This analysis led to estimates of log-relative risk, or the per cent increase in mortality with a corresponding increase in PM2.5 on a county basis. Considerable spatial variation was observed in log relative risk estimates across the U.S. Modeling results suggested an increase in the risk of mortality with PM2.5 compared to risk from PM10. Further, risk was estimated for sulfate and nitrate speciated particulate components. Sulfate was found to have a stronger correlation on mortality in the eastern U.S., and nitrate in the western U.S.

Using a new space-time model for fine particulate matter data (2001) in the midwest U.S., provided spatial maps showing the probabilities of exceeding the proposed fine particulate standard. For 2001, many urban areas were found to have high probabilities (> 0.8) of exceeding the new standard. For the first time, these maps were prepared based on aggregating weekly spatial predictions of fine particulate matter for the entire year, accounting for the temporal variability of PM2.5 across the year in the exceedance probabilities.

A regional trend analysis was conducted to estimate changes in rural CASTNet nitrate concentrations in the Midwest U.S. from 1990-1999. No significant trend was found for this period. Ongoing analyses are evaluating trends in data observed from 1990-2003.

Organized joint spatial prediction team with OAQPS (12/03)



Relevance :The development and application of new and innovative statistical approaches and algoritms are critical in making our understanding of physical processes more robust. These statistical approaches will enable us to combine data sets so that the probability and uncertanty of causal relationships between pollution (or other enviromental factors) with environmental and public health are made more explicit. Improve EPA's ability to understand the geospatial extent and transboundary transport of regulated air pollutants. Novel statistical approaches for solving major science problems described below will enable the timely communication of meaningful environmental information, improve emission management decisions, and allow efficient tracking of progress in air quality and public health goals. In a multi-disciplinary team environment, these modeling approaches will further our understanding of the environmental response to air pollution, particularly in exploring potential linkages between air quality and human health data.

Clients :Center for Disease Control, States of NY,WI, and ME, Clean Air Markets Division (OAR), Office of Air Quality Planning and Standards,

Office of Environmental Information, EPA Regional Offices

Project IDs:

ID Code :20957

Project type :OMIS