MODELS TO PREDICT SECCHI DEPTH IN SMALL GLACIAL LAKES

This study quantifies and ranks variables of significance to predict m ean values of Secchi depth in small glacial lakes. The work is based o n a new, extensive set of data from 88 Swedish lakes and their catchme nts. Several empirical models based on catchment and lake morphometric parameters are presented. These empirical models can only be used to predict Secchi depth for lakes of the same type, and the models based on ''geological map'' parameters can evidently not be used for time-de pendent and site typical predictions of Secchi depth. However, many of the principles behind the results ought to be valid for lakes in gene ral. Various hypotheses concerning the factors regulating the variabil ity in mean Secchi depth among lakes are formulated and tested. The mo st important variables are: Lake colour (expressing allogenic input of different types of humic materials), total-P and lake temperature (me asures of production of autogenic materials). The most important ''map '' parameters are: The mean depth (linked to resuspension and lake mor phometry) and the ratio between the drainage area and lake area (expre ssing the linkage between catchment and lake). The predictability of s ome of the models cannot be markedly improved by accounting for the di stribution of the characteristics in the drainage area (using the drai nage area zonation technique). The variability in mean Secchi depth fr om other factors, such as precipitation and anthropogenic load, may th en be quantitatively differentiated from the impact of these ''geologi cal'' factors, which can statistically explain 68% of the variability in Secchi depth among these lakes. The model based on map parameters c an also be used to estimate natural, preindustrial reference values of Secchi depth.