Estimation Theory Applications to Design of Water Quality Monitoring Systems

A quantitative methodology, utilizing Kalman filtering techniques, is developed for designing water quality monitoring systems. The basis is established for: (1) improvement of current practices of specification and enforcement of water quality standards; and (2) evaluating the economic trade-off between temporal and spatial frequency of sampling. Monitoring systems are characterized by spatial and temporal frequency of sampling and the variables to be measured. Utilizing a dynamic model of the aquatic environment and estimates of the uncertainty in model error and measurement error, a best sampling program is selected from a set of feasible sampling programs by sequentially minimizing a specified measurement system cost function. An optimal solution is not guaranteed. The power of the technique is based on the unique combination of model and data obtained from filtering techniques. The major shortcomings are: (1) the need for a model of the systems and (2) high computer costs.