Framework for Anticipatory Water Management: Testing for Flood Control in the Rijnland Storage Basin

Abstract

Anticipatory water management uses weather forecasts and water system simulation models to take operational water management actions before an event occurs. In this paper a framework consisting of the different steps and challenges of developing and evaluating an anticipatory water management strategy is discussed. The framework makes use of recent developments in ensemble weather forecasting and modeling of controlled water systems. In particular, flexible water-system control models enable simulation of a wide range of control strategies in multiyear hindcast analyses. With the increasing number of archives for weather forecasts and water-system-state variables, and the reduction in computational time resulting from increasing processor speed and parallel computing, hindcast verification analysis will become the basis for development and optimization of new operational water management strategies. The framework is successfully applied to a flood control case study in the Netherlands. Archived ensemble precipitation forecasts of the European Centre for Medium-Range Weather Forecasts (ECMWF) were fed into a simulation model of the water system to create water-level hindcasts. Decision rules to take anticipatory flood control actions on the basis of these ensemble forecasts have been developed. The optimized decision rules proved successful in reducing total damage costs resulting from deviations from target water levels for an analysis period of seven years.