Record Flooding Risk and Power Outage Restoration

We need to predict the probability of unprecedented flooding of lands and coastlines due to unexpected storms, overflowing rivers, hurricanes, tidal surges and dam failures.This paper addresses new record floods that exceed all prior“historic” levels and are invariably due to extreme or severe weather and/or unexpected precipitation, defeating barriers and causing extensive power system outages. Given their inherently low occurrence, the probabilities of new (rare) record floods are treated as random outcomes and independent events using classical statistical mechanics and related hypergeometric sampling. This analysis straightforwardly replaces tuning or fitting to “normal” precipitation, regular tides and prior flood data and the traditional use of multi-parameter extreme value distributions (EVDs) used for weather-induced flood forecasting and estimating “return periods”.The approach is not reliant on geographic computer models, meteorological forecasting, published “flood zone” charts, or hydrological techniques and images. We illustrate the universal applicability of this Bayesian-style approach of solely addressing new records for a wide range of specific flooding case studies for rivers, major hurricanes, quasi-periodic coastal tides, and dam failures. The quantitative link is shown between extreme event extent and power outage duration, and the results impact disaster resilience, infrastructure vulnerability and emergency preparedness measures.