颁辞苍迟补肠迟听
听
听Liton听Chakraborty, Partners for Action, Faculty of Environment
听Horatiu Rus, Department of Economics
听
Introduction
This study explores 铿俹辞诲-谤别濒补迟别诲 environmental injustices by deconstructing racial, ethnic, and socio-demographic disparities and spatial heterogeneity in the areal extent of 铿倁惫颈补濒, pluvial, and coastal 铿俹辞诲颈苍驳 across Canada. Environmental justice (EJ) research shows that cultural minorities, persons of lower socioeconomic status, and deprived socio-demographic groups or communities often experience excessive exposure to physical hazards, for example noise and air pollution, hazardous toxic wastes, or industrial pollution through chemical spills. However, EJ studies in the context of 铿俹辞诲颈苍驳 are few, and they typically report ambiguous relationships between the indicators of demographic and socio-economic status and exposure to 铿俹辞诲 hazards. This study is a 铿乺st attempt to assess whether in Canada conventionally recognized socially vulnerable groups, including visible minorities (Black, Asian, Indigenous peoples), females, elderly, and lone-parent households bear a disproportionate burden of 铿倁惫颈补濒, pluvial and coastal 铿俹辞诲 risk.
Methodology
The study utilizes national datasets of 铿俹辞诲 hazards, residential address points, census of population, and census tract (CT)-level cartographic boundaries to determine 铿俹辞诲 vulnerable neighbourhoods and the number of residential properties exposed to 铿倁惫颈补濒, pluvial, and coastal 铿俹辞诲颈苍驳 across 4,458 CTs in Canada. The variables in铿倁encing empirical relationships between 铿俹辞诲 risk, racial, ethnic, demographic, and socioeconomic status of households are grouped from the CT-level population characteristics to represent social vulnerability, socioeconomic and race/ethnicity status associated with a 100-year 铿俹辞诲 hazard exposure.
The dependent variable in this research is 100-year undefended, fluvial, pluvial and coastal 铿俹辞诲-prone land areas at 30-m resolution, based on the international flood modelling 铿乺m JBA Risk Management鈥檚 2018 铿俹辞诲 hazard datasets. These flood maps are national in scope, enabling 铿俹辞诲 hazard assessments at any location in Canada. The independent variables include race/ethnicity, two neighbourhood socioeconomic deprivation indices to capture social vulnerability, including economic insecurity and instability indices, and four variables on socio-demographic status, representing gender, old age, and physical disability. The racial or ethnic variables, including percentages Indigenous peoples, Black, South Asians, and other visible minorities were chosen to test the conventional EJ hypotheses that racial, ethnic or cultural minority groups are inequitably exposed to 铿俹辞诲 hazards and/or disproportionately affected by 铿俹辞诲 hazards.
To deal with spatial-non-stationarity, geographically weighted regression (GWR) on continuous outcome variables like the per cent share of residential properties in a CT-exposed to flood hazards, and geographically weighted regression logistic regression (GWLR) on dichotomous outcome variables like a CT in or out of flood hazard zone.
Outcomes
The statistical analyses suggest that the overall best-铿乼ting model for all three types of 铿俹辞诲 exposure is local GWLR. To visualize the spatial distribution of local parameter estimates, some of the regression estimates are mapped using Geographical Information Systems (GIS) across selected urban areas in Canada鈥檚 three biggest census metropolitan areas (CMA): Toronto, Montreal and Vancouver, where more than one third (36%) of all Canadians live. These maps show how the statistical associations between 铿俹辞诲 risk exposure and residents鈥� racial, ethnic, and socio-demographic status vary spatially by CTs.
Figure听1 portrays the relationships between pluvial 铿俹辞诲 risk and racial or ethnic and other socio-demographic variables at CT level in Toronto CMA. The positive association between a variable and pluvial 铿俹辞诲 exposure is displayed by the dark and green-shaded areas for the respective variable. A few CTs in western Toronto show, for example in Figure 1G a significant positive relationship between neighbourhood instability and pluvial 铿俹辞诲 exposure. In contrast, a few CTs in eastern Toronto show significant positive relationships in Figure 1H between economic insecurity and pluvial 铿俹辞诲 exposure. Estimated GWR parameters as displayed in Figure 1 clearly exhibit spatial-non-stationarity of covariates that explain a variety of relationships between socio-economic status and 铿俹辞诲 risk.
Figure 1: Spatial distribution of local GWR parameter estimates of pluvial flood risk model in Toronto CMA
Conclusions
This study is the 铿乺st attempt to explore the factors associated with the socio-geographic distribution of 铿俹辞诲 risk using GWR in Canada. The GWR technique improves the OLS approach by identifying local differences in socio-economic inequalities and 铿俹辞诲 risk determinants, which help policymakers identify the spatially varying association between 铿俹辞诲 exposure and racial or ethnic and socioeconomic characteristics. The generated GWR estimates, and their geospatial maps are an important starting point for a more detailed investigation of the disproportionate impacts of 铿俹辞诲 risk at Canada鈥檚 local level.
Geospatial mapping of GWR results is a powerful tool for motivating initiatives to target population subgroups for effective 铿俹辞诲 risk communication as it provides a scienti铿乧 basis for the location-speci铿乧 allocation of public resources to reduce socio-economic inequalities. The study finds that racial or ethnic, economic, social, and demographic factors play a signi铿乧ant explanatory role in the distribution of 铿俹辞诲 risk across Canadian neighbourhoods, even after controlling for spatial effects. Speci铿乧ally, exposure to 铿俹辞诲 risk is more signi铿乧ant in Canadian neighbourhoods that predominantly comprise certain vulnerable groups such as females, persons living alone, Indigenous, South Asians, the elderly (age 65 and over), other visible minorities, and economically insecure residents. These 铿乶dings can help promote a socially just 铿俹辞诲 risk management approach emphasizing the need to acknowledge socio-economic heterogeneity within various racial, ethnic, and socio-demographic groups.
Chakraborty, L. Rus, H., Henstra, D., Thistlethwaite, J., Minano, A. and Scott, D. (2022). Exploring spatial heterogeneity and environmental injustices in exposure to 铿俹辞诲 hazards using geographically weighted regression. Environmental Research.听
For more information about听WaterResearch, contact Julie Grant.
听
Photo: 听CC BY-NC-ND 2.0
