A local flood warning system typically serves a small community situat
ed in a headwater area and exposed to flash floods or rapid riverine f
loods. Forecasts of such flood events are characterized by short lead
times and large uncertainties, A Bayesian theory is formulated for a l
ocal flood warning system built of three functional components: monito
r, forecaster, and decider. The theory offers a modeling framework and
mathematical concepts necessary for (1) developing optimal decision r
ules for issuing warnings based on imperfect forecasts, (2) evaluating
system performance statistically, and (3) computing the ex ante econo
mic benefits from a system.