Student Research: Angel Ip

Methodology to Assess the Tripod 2006 Wildfire Impacts on Air Quality, Hospitalizations and Public Health in Washington State: A Case Study of the 2006 Tripod Wildfires
MS
Exposure Sciences (ES)
2013
Faculty Advisor: 

Abstract

Fire is an essential ecological process. However, forest fire burn area in the Pacific Northwest is likely to double or even triple by the end of the 2040s based on projected climate change models and ecosystem evaluations. The projected increase in forest fires has negative health implications because of air quality impacts. This study aims to characterize the linkage between air quality degradation due to wildfires and hospital respiratory and cardiovascular admissions, using the 2006 Tripod wildfires as a case study. The 2006 Tripod wildfires included the Spur Peak and Tripod Complex wildfires and were some of the largest wildfires in the past half-century in Washington State.

Air monitoring and air modeling data were used for different research analyses. Air monitoring data from the U.S. Environmental Protection Agency (EPA) and the U.S. Forest Service (USFS) were analyzed to assess any potential air quality degradation and air monitor availability during the Tripod wildfire event. The USFS smoke plume modeling data containing geospatial estimates of exposure for 33 days were used for analysis that linked air quality data to hospitalization and population data. Using a case-crossover epidemiological study, the relationship between air quality and respiratory and cardiovascular hospitalization was assessed. The case-crossover analysis was based on resident zip code exposures, and conducted using USFS air quality modeling data and respiratory and cardiovascular hospitalization records from the Washington comprehensive hospital abstract reporting system. Hospital relative rates addressing cardiovascular and respiratory hospitalizations impacts and exposure impacts in zip codes in Okanogan and surrounding counties were also assessed during the 33-day period, using 2010 Census data because the 2010 data had the most spatial coverage for the study area of interest. Lastly, modeled air quality data were compared against monitoring data from the EPA and USFS to assess the model’s validity, sensitivity, and specificity.

Air monitoring data from the EPA and USFS showed that there were higher than normal air quality degradation during the 2006 Tripod wildfire event. No significant results were found for modeled air quality data and their impacts on cardiovascular and respiratory hospitalizations in several epidemiological analyses by hospital zip codes and residential zip codes. The research found that current air quality model data systems are not accurate predictors of ground-level air monitor data systems.

For future studies, two scientific recommendations were determined. First, the medians and means of the modeled air quality values prior to the hospitalization date were found to be more meaningful, in comparison to mean modeled air quality value on the hospitalization date. Second, age categories for sensitive and non-sensitive populations (three age categories) were found to be more useful than finer divisions of six age categories. Policy recommendations to improve this type of research include increasing governmental agency coordination, improving the air quality monitoring network and air models, and having further research on the impact of wildfire events on air quality and hospitalizations. Additional suggestions were also made for future research into this subject including focusing on a smaller population and using smartphones to measure air quality and other types of data to address health impacts.