Places where wildfire potential and social vulnerability coincide in the coterminous United States

Abstract

The hazards-of-place model posits that vulnerability to environmental hazards depends on both biophysical and social factors. Biophysical factors determine where wildfire potential is elevated, whereas social factors determine where and how people are affected by wildfire. We evaluated place vulnerability to wildfire hazards in the coterminous US. We developed a social vulnerability index using principal component analysis and evaluated it against existing measures of wildfire potential and wildland-urban interface designations. We created maps showing the coincidence of social vulnerability and wildfire potential to identify places according to their vulnerability to wildfire. We found that places with high wildfire potential have, on average, lower social vulnerability than other places, but nearly 10% of all housing in places with high wildfire potential also exhibits high social vulnerability. We summarised our data by states to evaluate trends at a subnational level. Although some regions, such as the South-east, had more housing in places with high wildfire vulnerability, other regions, such as the upper Midwest, exhibited higher rates of vulnerability than expected. Our results can help to inform wildfire prevention, mitigation and recovery planning, as well as reduce wildfire hazards affecting vulnerable places and populations.

A Place-based Assessment of Flash Flood Hazard and Vulnerability in the Contiguous United States

“…The least vulnerable counties are located in New England and the upper Great Lakes. Results from the previous assessments 16,21,52 indicate similar spatial geographical patterns. Cutter and Finch 21 showed that the vulnerability is increasing in the southwest parts toward the U.S.-Mexico border regions of Texas in the last four decades, which can be due to the clustering of high vulnerable counties in that region, based on the SEVI calculated in this study.…”

“…This study aimed to generate a comprehensive SEVI; however, it still has some limitations. Although social vulnerability indices can efficiently describe broad-scale vulnerability, they fall short on integrating localized information related to exposure, sensitivity, and adaptive capacity that is often better collected using qualitative methods such as empirical methods and surveying 52,65 . Moreover, the calculated SEVI is based on only one point in time (the recent past), and the spatial patterns of socioeconomic vulnerability that were identified in this study represent the recent past.…”

“…Due to the high number of variables describing the social vulnerability, statistical procedures are commonly used to reduce data dimensions. The Principal Component Analysis (PCA) 48 has been widely utilized for decreasing the dimension of data in order to create a single and consolidated index of social variables 16,21,[49][50][51][52] . Based on PCA, the components which explain the majority of the variance in the data will be chosen to calculate the SEVI.…”

“…One shortcoming of using PCA-based index is that this procedure falls short in dealing with missing values. To overcome this problem, a few studies have suggested replacing the missing values with a value of zero 16,21,52 . However, a zero value cannot accurately represent the true vulnerability based on the particular variable of interest, and it would underestimate the level of vulnerability for the affected regions.…”

Social vulnerability to large wildfires in the western USA

“…The extensive literature on social science related to wildfire issues (McCaffrey, 2015) has studied risk perception, mitigation decisions and perceived consequences (Champ & Brenkert-Smith, 2016;Champ, Donovan, & Barth, 2013;Dickinson, Brenkert-Smith, Champ, & Flores, 2015;Gordon, Luloff, & Stedman, 2012); community pre-fire mitigation (Cohn, Williams, & Carroll, 2008) and adaptive capacity ; residents' actions and adaptation (Brenkert- Smith, 2006); and community social diversity and vulnerability (Paveglio, Nielsen-Pincus, Abrams, & Moseley, 2017;Paveglio, Prato, Edgeley, & Nalle, 2016). However, work to assess wildfire risk by integrating social and natural systems is relatively new (Cutter, Boruff, & Shirley, 2003;Davies et al, 2018;Gaither, Goodrick, Murphy, & Poudyal, 2015;Gaither et al, 2011;Oliveira, Zêzere, Queirós, & Pereira, 2017;Parisien et al, 2016;Wigtil et al, 2016) or applied at limited geographic scales (Fischer, Kline, Ager, Charnley, & Olsen, 2014;Nielsen-Pincus et al, In review;Olsen, Kline, Ager, Olsen, & Short, 2017;Paveglio, Edgeley, & Stasiewicz, 2018;Paveglio et al, 2016), leaving a gap in our knowledge about large-scale transboundary risk in relation to behavioral response to fire. Our first goal is to understand where transboundary large fire events originate and how they spread through a mosaic of land tenures, management jurisdictions and fuel models, and quantify their impacts on the communities of three socially and biophysically distinct fire-prone regions of the western US.…”