Department of Geography and Environmental Management, Faculty of Environment, University of Waterloo, Waterloo, Ontario, N2L 3G1, Canada.
Department of Economics and Political Science, University of Waterloo, Waterloo, Ontario, N2L 3G1, Canada.
Environ Res. 2022 Jul;210:112982. doi: 10.1016/j.envres.2022.112982. Epub 2022 Feb 23.
This study explores flood-related environmental injustices by deconstructing racial, ethnic, and socio-demographic disparities and spatial heterogeneity in the areal extent of fluvial, pluvial, and coastal flooding across Canada. The study integrates JBA Risk Management's 100-year Canada Flood Map with the 2016 national census-based socioeconomic data to investigate whether traditionally recognized vulnerable groups and communities are exposed inequitably to inland (e.g., fluvial and pluvial) and coastal flood hazards. Social vulnerability was represented by neighbourhood-level socioeconomic deprivation, including economic insecurity and instability indices. Statistical analyses include bivariate correlation and a series of non-spatial and spatial regression techniques, including ordinary least squares, binary logistic regression, and simultaneous autoregressive models. The study emphasizes the quest for the most appropriate methodological framework to analyze flood-related socioeconomic inequities in Canada. Strong evidence of spatial effects has motivated the study to test for the spatial heterogeneity of covariates by employing geographically weighted regression (GWR) on continuous outcome variables (e.g., percent of residential properties in a census tract exposed to flood hazards) and geographically weighted logistic regression on dichotomous outcome variables (e.g., a census tract in or out of flood hazard zone). GWR results show that the direction and statistical significance of relationships between inland flood exposure and all explanatory variables under consideration are spatially non-stationary. We find certain vulnerable groups, such as females, lone-parent households, Indigenous peoples, South Asians, the elderly, other visible minorities, and economically insecure residents, are at a higher risk of flooding in Canadian neighbourhoods. Spatial and social disparities in flood exposure have critical policy implications for effective emergency management and disaster risk reduction. The study findings are a foundation for a more detailed investigation of the disproportionate impacts of flood risk in Canada.
本研究通过解构加拿大河流、暴雨和沿海洪水泛滥面积的种族、族裔和社会人口差异以及空间异质性，探讨了与洪水有关的环境不公正问题。该研究将 JBA Risk Management 的 100 年加拿大洪水图与基于 2016 年全国人口普查的社会经济数据相结合，调查传统上认为的弱势群体和社区是否面临内陆（如河流和暴雨）和沿海洪水灾害的不公平暴露。社会脆弱性由邻里层面的社会经济剥夺来表示，包括经济不安全和不稳定指数。统计分析包括双变量相关分析以及一系列非空间和空间回归技术，包括普通最小二乘法、二元逻辑回归和自回归模型。本研究强调寻求最合适的方法框架来分析加拿大与洪水相关的社会经济不平等问题。空间效应的有力证据促使该研究通过地理加权回归（GWR）检验连续因变量（例如，普查区内暴露于洪水危险的住宅物业的百分比）和二分类因变量（例如，普查区内是否处于洪水危险区）的协变量的空间异质性。GWR 结果表明，内陆洪水暴露与所有考虑中的解释变量之间的关系的方向和统计显著性在空间上是非稳定的。我们发现，某些弱势群体，如女性、单亲家庭、土著人民、南亚人、老年人、其他可见少数民族和经济不安全的居民，在加拿大社区面临更高的洪水风险。洪水暴露的空间和社会差异对有效的应急管理和减少灾害风险具有重要的政策意义。本研究的发现为更详细地调查加拿大洪水风险的不成比例影响奠定了基础。
