A temporal recovery index for quantifying post-flood watershed recovery under uncertainty

ABSTRACT

Flood risk assessments generally emphasize pre-disaster physical conditions, whereas post-disaster recovery is often treated as static and homogeneous across watersheds. Yet each watershed has distinct inputs, responses, and recovery pathways: some systems rapidly return to pre-disaster conditions, while others experience prolonged degradation or develop new vulnerabilities. Floods may also alter pre-existing risks, creating uncertainty about how future events will affect changed systems. Quantitatively evaluating watershed condition through time is therefore essential for understanding evolving vulnerability and adaptive capacity. Existing approaches still lack an integrated temporal framework that represents post-disaster dynamics across interconnected components. To address this gap, this study proposes the Temporal Recovery Index (TRI), a time-sensitive composite framework for assessing multi-stage post-flood recovery under uncertainty. The method was applied to two contrasting flood-affected watersheds, Bozkurt–Ezine and Rize–Taşlıdere. Results revealed non-linear, basin-specific recovery trajectories. Under equal weighting, TRI_T3 was 0.954 for Bozkurt–Ezine and 1.187 for Rize–Taşlıdere. Monte Carlo mean TRI_T3 values were 0.957 (95% uncertainty interval: 0.820–1.123) and 1.188 (1.038–1.359), respectively. Comparative ranking was preserved in all eleven OAT configurations and in 98.0% of 10,000 simulations, supporting the internal robustness of TRI under the tested weighting and stochastic uncertainty assumptions and indicating stable differences between the two observed recovery trajectories.