Examining the relationship among meteorology patterns, air pollution and health outcomes for use in assessing climate impacts

Abstract

Recent research provides evidence that weather patterns in the northern hemisphere may be altered due to changes in climate induced by anthropogenic emissions of greenhouse gases and other pollutants (e.g., airborne particles). This study investigates the use of weather patterns in New York State for forecasting the impact of climate change on ambient pollution levels and health outcomes. Trajectory analyses were performed for eight meteorological regions across ten summers (1997 – 2006) in New York State. These trajectories were grouped by weather patterns and related to ozone concentrations and respiratory-related hospital admissions. Air masses were also associated with major regional emission sources in the Ohio River Valley, making this study one of the first to link emissions, transported pollution, human exposure and health endpoints. These grouped weather patterns will be used to predict the impact of changing weather induced by climate change on human health outcomes. This paper presents the approach for developing these weather pattern indices and discusses the preliminary results in linking emissions, ozone concentrations and health outcomes.

1. INTRODUCTION

The Clean Air Act requires that the U.S. Environmental Protection Agency (EPA) set National Ambient Air Quality Standards for pollutants considered harmful to public health and the environment. Previous research has shown that high ambient ozone concentrations are harmful to humans (e.g., Bell et al. 2005, Ito et al. 2005). While ozone is not directly emitted, the formation and distribution of ozone is driven by chemical reactions involving nitrogen oxides (NOx) and Volatile Organic Compounds (VOCs), as well as interactions with meteorological factors. NOx and the secondarily formed ozone can be transported downwind, contributing to pollutant levels at locations much farther from the emission sources, potentially impacting human health in downwind areas. Changes in climate patterns, such as longer periods of warmer weather or changes in wind patterns (Nolte et al. 2008), can affect ozone pollution levels and its impact on human health. This paper presents an approach for developing weather pattern indices and discusses preliminary results in linking emissions from the Ohio River Valley (ORV), ozone concentrations and health outcomes in New York State (NYS).

2. APPROACH

Back-trajectories were performed from several sites within the NYS domain across ten summers (1997 – 2006) to identify predominant weather patterns. These weather patterns were grouped as either originating from the ORV or not to assess the transport of pollution from this area of significant power plant emissions. The grouped weather patterns were investigated for associations with ozone concentrations and respiratory-related hospital admissions.

Daily maximum 8-hour ozone concentrations were calculated from hourly measurements for the summers (June 1 through August 31) of 1997 through 2006 obtained from the EPA's Air Quality System database (http://www.epa.gov/oar/data/aqsdb.html) and the Clean Air Status and Trends Network (http://www.epa.gov/castnet/). Daily 8-hour maximum ozone concentrations were interpolated to provide estimates for each county to coincide with the hospital admissions data. Back trajectories from selected sites in eight meteorlogical regions were computed using the HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory) model (Draxler and Hess, 1997) for 48 hours back in time, producing a total of 927 trajectories for each site. Health data for the summers of 1997 through 2006 were obtained from the NYS Statewide Planning & Research Cooperative (SPARCS). These data included daily hospital admissions for respiratory-related diseases, including asthma, chronic bronchitis, chronic obstructive pulmonary disease (COPD), emphysema, and pneumonia and influenza.

3. DISCUSSION AND RESULTS

The transport of ozone into NYS was examined by relating the grouped back trajectories to ozone concentrations and hospital admissions using a variety of statistical techniques including tests of statistical significance and Poisson regression models. Based on these back-trajectories, each day was categorized as having a wind flow pattern originating from the ORV or not originating from the ORV. These days were then matched to the corresponding daily maximum 8-hour ozone concentrations and daily respiratory-related hospital admissions. The results of this analysis indicate that the mean levels for ozone concentrations and respiratory-related hospital admissions were significantly higher for those days when the sites were downwind from the ORV versus those days that the sites were not downwind from the ORV. Because of many confounding factors (e.g., temperature, co-pollutants), the health signal associated with the grouped meteorological trajectories is difficult to relate to the variables of interest (weather patterns and ozone concentration levels). Thus, results from applying a Poisson regression model will also be presented.

Acknowledgments: The authors wish to thank Jim Godowitch, Kristen Foley and Jenise Swall for their contributions to this study. Although this work was reviewed by EPA and approved for publication, it may not necessarily reflect official Agency policy.