Flood risk is increasingly shaped by the combined effects of climate change and the vulnerability of built environments, while building-level flood damage severity analysis is often constrained by limited data availability. This study develops a transparent and reproducible framework for analyzing building-level flood damage severity under climate-adjusted hazard conditions in data-constrained environments. The framework integrates administrative post-event damage records, GIS-based terrain information, a terrain-based proxy flood-depth reconstruction procedure, and a standardized Rhine Atlas/ICPR depth–damage relationship. Representative terrain-based proxy flood depths are reconstructed using building locations, terrain elevation, and settlement-level exposure assumptions. Observed damage categories are not used to assign proxy flood depths directly, but serve exclusively as empirical ordinal reference information for ordinal consistency assessment of model-derived damage severity. Climate effects are incorporated through a simplified hazard adjustment based on projected changes in extreme precipitation intensity. The framework is applied to 413 residential buildings affected by flood events in Serbia during the period 2016–2021. Results show a consistent nonlinear relationship between terrain-based proxy flood depth and ICPR-derived structural damage severity, as well as a strong influence of terrain elevation on relative hazard intensity. Climate-adjusted sensitivity scenarios indicate that even moderate increases in extreme precipitation lead to measurable increases in structural damage severity and an upward shift in model-derived damage levels. The proposed framework provides a practical approach for flood damage severity analysis in data-constrained environments, supporting improved decision-making in sustainable flood risk management and climate adaptation planning.

Conventional exposure assessments for toxic release events often rely on spatial hazard representations that provide limited information on how exposure evolves while evacuation is underway. This study proposes a dynamic exposure assessment framework that integrates time-dependent concentration profiles with scenario-specific evacuation completion times, including vulnerability-related evacuation delay. Exposure is quantified as accumulated exposure from release onset until evacuation completion, while the duration above a selected Acute Exposure Guideline Level (AEGL) reference threshold is used as a complementary temporal exposure metric. The framework is demonstrated using an ammonia-release scenario simulated with ALOHA and three illustrative evacuation-delay scenarios evaluated under identical hazard conditions. The results show that progressively delayed evacuation increases both accumulated exposure and the duration above the selected AEGL-2 reference threshold. However, the exposure contribution of additional evacuation time depends strongly on its temporal position within the evolving concentration profile. Additional evacuation time coinciding with the peak concentration plateau produced the largest marginal increase in accumulated exposure, whereas additional time during the later decay phase contributed comparatively less. A structured sensitivity analysis showed that the principal qualitative findings remained stable under $\pm 15\%$ perturbations of the reconstructed concentration profile, although the AEGL-2 exceedance duration was more sensitive to concentration variations. These findings demonstrate that exposure outcomes depend not only on evacuation duration but also on the temporal alignment between evacuation timing and hazard evolution. The proposed framework provides a structured approach for linking time-dependent dispersion outputs with evacuation assumptions and can be adapted to other toxic release scenarios when appropriate concentration–time data and substance-specific reference criteria are available. The approach may provide additional information for temporally informed emergency planning, particularly for populations requiring additional evacuation time or assistance.

Boundary uncertainty arises when a continuous insurance-based severity ratio is converted into ordered damage states using crisp thresholds, causing nearly identical claims to receive different labels solely because they fall on opposite sides of a decision boundary. This study presents a transparent computational framework that explicitly represents this uncertainty through predeclared piecewise-linear transition bands and soft labels while retaining probabilistic damage-state exceedance modelling and ordinal coherence within a transparent and auditable computational workflow. The empirical study uses 802 residential storm-wind insurance claims from Vojvodina, Serbia, recorded during 2013–2018. The response is the Relative Damage Ratio (RDR), calculated as the ratio of Insurance Claim Payout to policy Sum Insured and expressed as a percentage. The framework combines matched hard-label and soft-label cumulative logistic models with leakage-controlled calendar-month-blocked validation, paired blocked-bootstrap comparison and prespecified robustness analyses. Hard-label and soft-label models produced nearly identical proper-score and AUC point estimates; all paired 95% blocked-bootstrap confidence intervals for Brier-score, cross-entropy and calibration-slope differences included zero. Transition-width and local threshold-perturbation analyses likewise showed no consistent empirical advantage of either target representation. A secondary mathematically coupled specification incorporating log(Sum Insured), which also enters the denominator of RDR, yielded higher AUC values of approximately 0.76–0.80 compared with 0.58–0.63 in the primary wind-only models; these results are therefore interpreted as diagnostic rather than as evidence of independent physical predictive information. Rather than claiming predictive superiority for soft labels, the proposed framework makes boundary assumptions explicit while maintaining ordinal coherence by construction and yielding comparable calibration metrics, thereby providing a transparent, auditable and fully specified computational framework for ordinal probabilistic modelling in which continuous outcomes are converted into threshold-based cumulative states, with potential applicability to other ordinal risk-assessment settings.