Search

Vulnerability studies often measure the risk that influences a system's ability to react during the occurrence of a hazardous event. Risk is generally associated with the physical, social, and economic aspects of the system. Climate-related hazards such as floods involve an additional spatial dimension, however. It is challenging combining social and geophysical vulnerabilities to model their interrelations to overall flood vulnerability. This study proposes the partial least squares structural equation model (PLS-SEM) with several latent variables of geographical characteristics of houses and socioeconomic dimensions of households representing flood vulnerability. The former includes GIS/remote sensing data, while the latter comprises household surveys. The results indicate social indicators get contribution mainly from direct effects of spatial neighborhoods and socioeconomic conditions, and indirect effects of land tenure and demographics, with path coefficients equal to 0.47, 0.42, 0.83, and 0.61, respectively, all significant at the p < 0.0001 level. Overall, the goodness of fit equals 0.72 indicates good explanatory power of the model. We conclude that the PLS-SEM successfully incorporated the combined vulnerability from social and geographical factors. The SEM framework allows integrative assessing vulnerability variables from different disciplines, formats, and scales, and thus, offers an opportunity for creating more tailored policies to combat hazards. [ABSTRACT FROM AUTHOR]