The Scale-Dependent Relationship Between Building Spacing and Structural Damage Following Hurricane Ian, Fort Myers Beach, Florida

Abstract

Hurricanes cause extensive structural damage in coastal communities, and identifying the factors that govern this damage can inform resilience planning. Distance from the shoreline is an established predictor of hurricane damage, but the role of how densely buildings are spaced remains unclear. Denser spacing could plausibly shelter interior buildings or, conversely, concentrate wave and debris loading on exposed ones. The association between building density and structural damage was evaluated for Fort Myers Beach, Florida, following Hurricane Ian (2022). Publicly available Virtual Damage Assessment data for 3,408 structures were integrated with Lee County pre-storm building footprints for 3,346 structures. Density was quantified using nearest-neighbor distance, neighbor counts, and local density, and each metric was computed under two definitions of spacing: the distance between building centroids, and the true gap between building outlines (edge-to-edge). Both definitions were significantly associated with damage and reached comparable peak strength (Spearman ρ ≈ 0.23–0.25). The associations differed sharply, however, in how they depended on neighborhood scale. When density was recomputed across search radii from 1 to 100 m, the edge-to-edge measure produced a distinct peak at approximately 23 m (ρ = 0.248, p < 10⁻⁴¹) and declined steeply at larger radii, whereas the centroid measure rose to a broad plateau and peaked only near 43 m (ρ = 0.232, p < 10⁻⁴⁴). This sharply defined optimum indicates a characteristic spacing scale of roughly 20 m over which building configuration influences damage — a scale the centroid measure blurs because it conflates building size with separation. Across damage states, the density signal was concentrated at the most severe end: buildings that were completely destroyed (DS6) were substantially more tightly packed than less-damaged structures, whereas density varied little among the intermediate states. Controlling for shoreline distance sharply reduced the association (partial Spearman ρ = 0.10, p < 10⁻⁶), indicating that most of the apparent density effect reflects proximity to the coast, while a small but statistically robust component remains attributable to local building configuration. These findings show that building density is a scale-dependent descriptor of structural vulnerability, that edge-to-edge spacing provides a more physically interpretable measure of this scale than centroid spacing, and that an optimized, physically grounded density metric could improve hurricane damage prediction.