Thesis
The practical application of a conceptual framework using social media to understand community-level response to wildfire smoke in the western US
Washington State University
Master of Science (MS), Washington State University
05/2019
DOI:
https://doi.org/10.7273/000003964
Handle:
https://hdl.handle.net/2376/124733
Abstract
Poor air quality from wildfires has a significant effect on public health. In this paper we investigated the community-level social media response in Spokane, Washington to seasonal wildfire smoke events. Surface air quality measurements were assembled from numerous locations and sources around Spokane. Twitter was chosen to investigate the social media response due to its widespread use and data availability. A method was developed for strategic selection of search terms which included extracting terms from literature and public health instruments/tools, and then convening an expert panel with clinical knowledge of respiratory illness. A wide variety of smoke and health related search terms were chosen through this method and placed in five categories: ambient air quality conditions, respiratory symptoms and exacerbations, risk perception and self-efficacy, behavioral responses and self-care management, and quality of life and healthcare utilization. Tweets from the Spokane area that matched search terms were obtained from June through September of 2017 and 2018, which covered the 2017 and 2018 wildfire seasons. This amounted to 160,000 tweets before filtering. It was shown that seasonal smoke is an active topic on Twitter, with the number of tweets about smoke significantly increasing for periods during wildfire smoke events compared to background periods without smoke. For the unfiltered full categories of search terms, there was not good correlation between PM2.5 and the number of tweets (due to the many various uses of the search terms). However, certain search terms were correlated with particulate matter data. The words with the highest correlation were "smoke" (R2=0.77), "smoky" (R2=0.66), and "unhealthy" (R2=0.65). Cumulative frequency distribution plots and word clouds were created that identified the words that individuals use most to talk about the smoke event in each search term category. A list of words was identified that were used by individuals, but that had not been included in the search terms; these can be used to refine the searching method
Metrics
4 File views/ downloads
32 Record Views
Details
- Title
- The practical application of a conceptual framework using social media to understand community-level response to wildfire smoke in the western US
- Creators
- Marissa Anne Grubbs
- Contributors
- Von P Walden (Advisor) - Washington State University, Department of Civil and Environmental Engineering
- Awarding Institution
- Washington State University
- Academic Unit
- Department of Civil and Environmental Engineering
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University
- Identifiers
- 99900890804301842
- Language
- English
- Resource Type
- Thesis