Neglecting uncertainties surrounding model parameters can drastically
underestimate flood risks
Abstract
Floods drive dynamic and deeply uncertain risks for people and
infrastructures. Uncertainty characterization is a crucial step in
improving the predictive understanding of multi-sector dynamics and the
design of risk-management strategies. Current approaches to estimate
flood hazards often sample only a relatively small subset of the known
unknowns, for example the uncertainties surrounding the model
parameters. This approach neglects the impacts of key uncertainties on
hazards and system dynamics. Here we mainstream a recently developed
method for Bayesian data-model fusion to calibrate a computationally
expensive distributed hydrologic model. We compare three different
calibration approaches: (1) stepwise line search, (2) precalibration or
screening, and (3) the new Fast Model Calibrations (FaMoS) approach.
FaMoS deploys a particle-based approach that takes advantage of the
massive parallelization afforded by modern high-performance computing
systems. We quantify how neglecting known unknowns can drastically
underestimate extreme flood events and risks. Accounting for parametric
uncertainty improves model performance metrics over the best estimate
parameters. Improving the characterization of model parametric
uncertainty improves hindcasts and projections of flood risks.