Local-scale social vulnerability to multiple hazards

This dissertation explores the processes by which social circumstances interact with physical conditions to produce different levels of suffering and recovery from multiple instances of hazard events. This is called social vulnerability. Understanding and modeling social vulnerability is necessary to communicate lifesaving information, and to inform decision making before, during and after hazard events. In this dissertation, I defined how social vulnerability is measured by different classifications of multihazards, and assessed the implications for understanding social vulnerability. I then identified how disbursement of disaster assistance is influenced by social vulnerability across the conterminous US. Finally, I developed a new way of modeling social vulnerability to multihazards, and through stakeholder assessment, confirmed its usefulness for decision making before, during and after disasters. The results showed distinct classifications of ‘multihazard’ may provide different information and therefore need to be seen as complimentary rather than equivalent in what they say about social vulnerability. Secondly, the study showed that multihazard social vulnerability has different associations with disaster assistance based on where you live and factors of race. Disaster assistance programs therefore need to explicitly consider social vulnerability if they are to become socially equitable, or they risk perpetuating existing inequality. Lastly I developed a new model of social vulnerability to multiple hazards. I show that applying qualitative data on local context of hazard exposure and livelihoods enables a composite index with adjustable indicator selection and weighting. This index outperformed a standard social vulnerability index and was more suitable for local and national disaster risk reduction activities in Guyana.