Flooding and energy instability erode household wealth and local tax bases across the Gulf Coast, disproportionately impacting LMI homeowners. Studies show flood exposure lowers property values by 2–10%, while green stormwater infrastructure can raise prices by up to 8.8%. Evidence on clean-energy and grid-modernization projects remains limited but suggests modest value gains when benefits are visible and equitable. Yet, capitalization effects depend on siting, quality, and neighborhood context—poorly maintained or sited projects may reduce value. Few studies address these interacting factors, limiting insights for equitable and sustainable adaptation planning.

To address this gap, the Community Co-Financed Flood and Energy Resilience (CCOFFER) initiative proposes an integrated “what-if” modeling framework that fuses causal inference with graph neural networks (GNNs) to evaluate how local environmental, infrastructural, and socioeconomic conditions jointly influence property-value responses to resilience and energy-transition investments. The framework generates spatially explicit, context-aware evidence to inform sustainable decision-making across Gulf Coast communities.

In Stage 1 (Causal Identification), modern difference-in-differences and event-study methods isolate localized, time-resolved impacts of flood or energy interventions on property markets, revealing how price responses differ across contexts such as baseline risk, LMI concentration, and infrastructure quality. In Stage 2 (Graph-Based Learning), these causal signals are embedded within a heterogeneous GNN linking parcels, facilities, and risk nodes through spatial, drainage, and energy networks. This design learns nonlinear, place-specific interactions—how flood connectivity, facility accessibility, and social vulnerability jointly shape property values—while retaining causal consistency.

Once trained, the hybrid model functions as a “what-if” simulator, allowing users to test alternative investment or hazard scenarios and assess equity impacts. The initial testbed in Lake Charles, Louisiana will expand to New Orleans, Mobile, and Corpus Christi. Collaborative development with governments, utilities, and community partners ensures transparency and policy relevance. Integrated into the Gulf Coast Resilience Data Center (GCRDC), the system will deliver automated data pipelines and gamified visualization tools for participatory planning.

By linking causal inference, graph learning, and scenario simulation, CCOFFER advances understanding of how infrastructure, risk, and social context interact across the Gulf Coast—supporting equitable, financially sustainable, and community-centered adaptation strategies.