This study introduces a novel, building-level flood vulnerability index, which was applied to a case study in Harris County, Texas. This new index is comprehensive and multidimensional, incorporating data beyond standard social metrics to include factors like structural characteristics and traffic congestion. A key finding is that the different domains of vulnerability (such as social, financial, and the built environment) are largely independent. This demonstrates that a "one-size-fits-all" mitigation approach is ineffective. Instead, the research shows that neighborhoods have unique combinations of risk, requiring policy efforts to be tailored to the specific vulnerabilities present in each location.

The analysis also reveals that flood risk is hyper-local, with approximately two-thirds of the variation in vulnerability occurring within single neighborhoods (census tracts) rather than between different ones. This finding is significant because most research is conducted at the tract level, meaning it likely misses substantial internal disparities. Furthermore, the study uncovered nuanced equity implications: while high-poverty and majority Black neighborhoods exhibited relatively low vulnerability, Hispanic neighborhoods displayed the highest levels of flood vulnerability and risk. This highlights the critical need for downscaled analysis to help policymakers equitably allocate resources and target interventions in the most vulnerable communities.