Lake eutrophication: Identification of tributary nutrient loading and sediment resuspension dynamics

In a shallow lake system two problems are of particular significance. It is important to know the distribution of (polluting) nutrient loads among various agricultural and municipal sources within the lake's catchment. It is also important to know what effects the nutrients already deposited in the lake's sediments will have on the biological and chemical features of the lake's ecological system. The paper presents results for the analysis of time-series observations of sediment resuspension characteristics in Lake Balaton, Hungary, and of nutrient loadings in the principal tributary of this lake, the River Zala. Two recursive methods of estimation, an instrumental-variable (IV) algorithm and the extended Kalman filter (EKF), are used for analysis of the time series. The paper takes the view that models are most useful in terms of their failure to replicate observed behavior, in that this is the primary stimulus to insight and progress in understanding. Brief comments on some limitations of method are given. Two potentially fruitful modifications of existing identification methods are outlined, one being broadly consistent with the conceptual analog of engineering structural analysis, while the second makes reference to the development of algorithms based on expert systems.