Component sensitivity: A tool for the analysis of complex water resource systems

The value of sensitivity estimates for determining an optimal set of model parameters has been recognized for over a century. However, the inadequacy of the mathematical foundation of sensitivity has prevented the use of systematic gradient optimization techniques for the analysis of complex water resource systems and hydrologie simulation models. The computer time required to increment each parameter and to measure the change in output is considered excessive for optimization of complex systems. The use of component sensitivity provides an alternative to the method of parameter perturbation that greatly reduces the computational effort. The mathematical foundation of component sensitivity is introduced, and the method of computation is demonstrated for a two‐component system. However, the mathematical foundation of component sensitivity is easily extended to more complex systems.