Abstract
This study assessed the spatial distribution of vulnerability to extreme heat in 1990 and 2000 within metropolitan Phoenix based on an index of seven equally weighted measures of physical exposure and adaptive capacity. These measures were derived from spatially interpolated climate, normalized differential vegetation index, and U.S. Census data. From resulting vulnerability maps, we also analyzed population groups living in areas of high heat vulnerability. Results revealed that landscapes of heat vulnerability changed substantially in response to variations in physical and socioeconomic factors, with significant alterations to spatial distribution of vulnerability especially between eastern and western sectors of Phoenix. These changes worked to the detriment of Phoenix's Hispanic population and the elderly concentrated in urban-fringe retirement communities.
Este estudio evaluó la distribución espacial de la vulnerabilidad al calor extremo en 1990 y el 2000 dentro del área metropolitana de Phoenix, sobre la base de un índice de siete medidas igualmente ponderadas de exposición física y capacidad de adaptación. Estas medidas se derivan del clima interpolado espacialmente, del índice normalizado de vegetación diferencial, y datos censales de EE.UU. A partir de mapas de vulnerabilidad también se analizaron grupos de población que viven en zonas con vulnerabilidad a las altas temperaturas. Los resultados revelaron que los paisajes con vulnerabilidad al calor cambiaron sustancialmente en respuesta a variaciones en factores físicos y socioeconómicos, con modificaciones importantes en la distribución espacial de la vulnerabilidad, especialmente entre los sectores este y oeste de Phoenix. Estos cambios se dieron en detrimento de la población hispana de Phoenix y los ancianos concentrados en comunidades de jubilación urbano-marginales.
Notes
*This article was developed from a poster presented at the 2009 Association of American Geographers Annual Meeting in Las Vegas, Nevada. We thank Anthony Brazel, Darren Ruddell, Nancy Selover, Chona Sister, and Sally Wittlinger (Arizona State University) for their helpful comments. We also appreciated critiques from three anonymous referees and the editor that considerably improved this article. This material is based on work supported by the National Science Foundation under Grant SES-0345945, Decision Center for a Desert City. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.
1 VIF is a measure of the impact of collinearity among variables in a regression model. VIF > 10 indicates definite problems of multicollinearity; VIF > 2.5 indicates potential areas of concern. As VIF magnitudes in are between 1.0 and 1.6, this suggests that the collinearity problem among our independent variables is relatively minor.