Category‐unit loss functions for flood forecast‐response system evaluation

Abstract

Economic evaluation of the performance of a flood forecast‐response system requires a large amount of information for the estimation of a loss function. This function must encode the cost of protective action undertaken in response to flood warnings and the flood damage sustained. The concept of category‐unit functions is exploited to reduce field survey demands and to circumvent high computation requirements. The loss function consists of a unit damage function, a unit cost function, and a unit reduction function; the degree of response that can be actually implemented is bounded by a decision constraint function. These four functions are estimated from data of various origins for seven categories of structures and three categories of decision makers.