DESCRIBING ANTHROPOGENIC IMPACTS ON STREAM WATER-QUALITY - THE PROBLEM OF INTEGRATING SOIL-WATER CHEMISTRY VARIABILITY

Problems encountered when (a) lumped soil water information is used to assess the relevance of cation exchange equations in the soil and (b) lumped mathematical models are used to describe the hydrochemical behaviour of catchments, are discussed. To highlight the difficulties and to focus the discussion, an example is provided using synthetic data. This data is generated using a mathematical model employing simple cation exchange reactions to describe water chemistry within the microporous regions of the soil. The proportion of the individual cations is assessed by varying the strong anion content within the micropore. These solutions are then mixed and the resultant chemistries are compared with the cation exchange formulations. It is taken that, on transfer from the micropore to the soil water sampler, the mixed solutions do not participate in cation exchange reactions: the chemical speciation programme ALCHEMI is used to characterise the chemical speciation changes occurring after mixing. The model is calibrated on the basis of average soil water data from two acidified spruce forest sites. The results show that by lumping the soil chemistry data, the averaged values for the chemical constituents systematically deviate away from the underlying theoretical cation exchange equations when differing cation charges are involved: no deviations occur when the charges are the same. The findings are extended to suggest that, for the chemically heterogeneous catchments being studied, the use of lumped mathematical models, which smooth out and ignore this variability, nullifies the theoretical equations on which the models are based. An appeal is made for (a) more detailed field studies to determine the degree of catchment inhomogeneity for soil properties and chemistry and (b) an assessment of the importance of catchment heterogeneity within a modelling framework.