Complexity in Columbia River Systems Modeling

Abstract

This paper reviews the problem of model complexity in evaluating the tradeoffs between fish and power production on the Columbia River. Characterizations of model complexity are presented and appropriate levels of model complexity are discussed. Models currently used to investigate the conflict between fish and hydropower production in the Columbia River Basin are reviewed. It is concluded that the complexity of the models used to generate tradeoffs between conflicting uses of water in this setting hinder their successful application and, because of their reliance on historic streamflow records, do not adequately address the problem in a probabilistic framework. To demonstrate the value of a simplified model, the operation of an equivalent composite reservoir was simulated. After calibration of the simplified model (CRISP), a series of Monte Carlo simulations were run to indicate possible impacts of alternative operational strategies on firm and secondary power and average annual generation.