IDENTIFYING FLOW PATHS IN MODELS OF SURFACE-WATER ACIDIFICATION

A clear identification of the movement of water through soils is of central importance to the characterization of surface water acidification. One currently popular view is that the chemical composition of stream water is a function of the mixing of varying volumes of water deriving from different parts of the catchment, all with their own individual chemical “signatures.” The development of models describing this process, and their use for predictive purposes, is a controversial issue. Some would shun the use of models altogether; others would insist on their development based rigorously on the laws of physics; and yet others would accept the aggregation and approximation of simpler conceptual models. As such, this controversy is common to many of the disciplines of geophysics. The paper explores in detail the problem of how a uniquely best set of values for the coefficients of simple conceptual models can be estimated from field observations. In particular, we investigate the role of tracer observations in reducing the uncertainty (and hence ambiguity) associated with the model's constituent hypotheses about hydrological mechanisms. Results from one of the field sites of the (U.K.) Royal Society’s Surface Water Acidification Programme show that even simple conceptual models can lead to quite different interpretations of the dominant flow paths in a catchment. The skill of predicting future behavior will lie therefore in locating this almost inevitable ambiguity and in seeking to make predictive statements that are minimally sensitive to it. Copyright 1990 by the American Geophysical Union.