Research library for resilient housing.

It is estimated that around 1/3rd of the US population lives in unincorporated areas that lie outside of municipal boundaries. Considering the substantial demographic segment and the increasing incidence of disasters, it is important to understand how unincorporated communities plan for, respond to, and recover from disasters; however, limited scholarly attention has addressed this topic with coverage focusing on singular forms of unincorporated communities, such as colonias and AIAN communities, and no coverage of unincorporated communities generally. A more comprehensive understanding of the vulnerability, exposure, risk, and resilience of unincorporated communities to disasters could allow addressing how these populations can better prepare for, respond to, and recover from disasters. This systematic review intends to explore the key problems, solutions, and challenges faced by these communities during different stages of disaster. The paper concludes with recommendations for how unincorporated communities can increase resilience and capacity when faced with disasters.

The Community Development Block Grant–Disaster Recovery (CDBG-DR) program is the primary federal mechanism for financing long-term housing recovery, yet expenditures often lag years behind events. This study examines how administrative processes influenced implementation following the 2017 disasters in Texas, Florida, and California. Drawing on state Action Plans and longitudinal data, the findings reveal that while repeated amendments allowed for adaptation, they contributed to delayed program rollout. The analysis shows that implementation efficiency varied significantly; Texas and Florida achieved closer obligation–expenditure alignment through incremental adjustments, whereas California faced persistent delays driven by prolonged front-end planning. The results suggest that delays in CDBG-DR assistance are systemic, highlighting a need for standardized early guidance and streamlined planning to expedite assistance to disaster-affected households.

The rise in global disasters has highlighted the need for innovative and resilient housing solutions that can withstand and recover from catastrophic events. Conventional construction methods are often time-consuming, labor-intensive, costly, and ineffective in providing adequate resilient housing able to withstand disasters and/or recover from them. 3D printing technology (3DPT) offers a promising solution for disaster-resilient housing. However, comprehensive knowledge of their utilization in this specific area is lacking. This systematic review explores the utilization of 3DPT specifically towards disaster-resilient housing, encompassing both pre-and post-disaster scenarios through retrofitting and recovery applications. In pre-and post-disaster applications, 3DPT offers promising solutions by streamlining construction processes, reducing waste, and enabling rapid customization of housing solutions. The review identifies critical barriers in pre and post-disaster housing and highlights the transformative potential of 3DPT in revolutionizing the construction industry. Through an analysis of the literature, it becomes evident that 3DPT presents opportunities to address significant challenges often faced by conventional construction methods. The review also conducts a t-distributed stochastic neighbor embedding (t-SNE) to help with visualizing the emerged clusters of the gathered reports using Gaussian Mixture Models, topic modeling using latent Dirichlet allocation, and a SWOT analysis, which reveals strengths such as customizability and time efficiency while acknowledging weaknesses like high initial investment and material restrictions. Recommendations for future research include standardization and code development, material innovation, community engagement, long-term performance evaluation, and policy and governance considerations. By addressing these research gaps, 3DPT can maximize their potential to provide sustainable, cost-effective, and resilient housing solutions for communities worldwide.

Reliable models of urban flood losses are essential for climate-resilient planning, yet it remains unclear whether models calibrated for one disaster can be reused for others in different cities or events. This study develops and tests tract-level models of National Flood Insurance Program (NFIP)-insured housing losses for 6,065 census tracts across three major U.S. floods: the 2016 Tax Day Flood and Hurricane Harvey in Texas, and Hurricane Irma in Florida. Using a unified hazard-exposure-vulnerability framework, we integrate peak 6-hour rainfall, floodplain exposure, insurance coverage, population density, social vulnerability, building age, and Community Rating System discounts to predict a normalized loss ratio and the probability of any insured loss. We compare conventional regressions (Ordinary Least Squares, Spatial Lag Regression) and classifications (Logistic Regression) with machine learning models (Random Forest, XGBoost) and evaluate performance using spatial cross-validation that trains and tests models on different geographic areas. Classification models substantially outperform regression, achieving a within event area under the ROC (Receiver Operating Characteristic) curve of 0.76-0.89, whereas regression peaks at 0.32 for Harvey and is near zero for the other events. Transferability is asymmetric: XGBoost models retain high performance when transferred between the two Texas events (Area Under the Curve, AUC 0.91-0.97) but degrade when transferred between Texas and Florida (AUC 0.75-0.88). Population density and rainfall intensity emerge as the most influential predictors across models. Because NFIP data capture only direct insured losses, transferred models require recalibration but can still support rapid post-disaster screening of census tracts likely to experience insured housing losses, informing more equitable inspection, recovery, and adaptation strategies in flood-prone urban regions.

As climate change increases the frequency and severity of disasters, proactive planning for post-disaster housing recovery is essential to mitigate long-term social and economic disruption. Computational models can support this planning by simulating potential recovery trajectories, yet many existing approaches are limited by overwhelming data requirements or narrow applicability to past events. Here, we introduce RAAbIT (Recovery Assessment using Agent-based Tools), a novel agent-based model designed to simulate housing recovery using data available within weeks of a disaster. RAAbIT models individual households, insurers, and contractors as agents governed by empirical behavior rules, and incorporates modifiable system-level constraints, such as contractor availability, to reflect context-specific recovery dynamics. We demonstrate the model’s utility by hindcasting two California wildfires—the 2017 Tubbs Fire in Santa Rosa and the 2018 Camp Fire in Paradise—and capturing their divergent recovery trajectories. Despite similar hazards, the two communities experienced significantly different reconstruction outcomes, with Santa Rosa rebuilding 57% of destroyed homes and Paradise only 9% within five years. RAAbIT can reproduce temporal and spatial patterns of recovery observed in building permit and construction data. By balancing generalizability with data realism, RAAbIT provides a flexible and transferable tool for post-disaster recovery planning, supporting more effective decision-making under uncertainty and enhancing community resilience in the face of escalating climate risks.

This study employed a longitudinal approach to analyze housing recovery following Hurricane Sandy, utilizing survival analysis to assess the time required for property values to return to or exceed their pre-disaster appraised values. Lots’ appraised values before the hurricane and across multiple years, post-disaster, were extracted as proxies for damage severity and recovery progress. The recovery timeline was then linked to household and housing characteristics to determine their significance in long-term recovery. Results indicated that households with higher socioeconomic status and education levels, as well as those residing in older homes, tended to recover more slowly compared to their counterparts. These findings provide critical insights into the factors influencing long-term housing recovery, offering valuable guidance for disaster recovery planning and policymaking at various levels to enhance resilience and equitable recovery outcomes.