The objective of this proposal is to formulate a two dimensional stochastic estuarine hydrodynamic model, incorporating probabilistic factors that describe the estuarine system's boundary and initial conditions, parameter fields and forcing functions. The research will involve analysis of parameter and uncertainty on the solution of the hydrodynamic equations, use of several solution algorithms for the stochastic prediction problem, Monte Carlo method quasilinearization allied with perturbation and operator analysis and the direct solution of the stochastic differential equations that characterize behavior of the estuarine system. The solution algorithms will be used to generate the probability density functions and/or moments describing water elevations and flow rates in the estuarine system. During the final phase of the project, solution algorithms will be used for the stochastic prediction of water levels in the Redwood Creek estuary in northern California. The results of the investigation are expected to provide baseline information regarding how parameter and forcing function uncertainty affects the output variances of water elevations, and solution algorithms for predicting variations in estuarine water quality that is affected by pollutants from industrial, municipal and agricultural sources.
