Application of a hydrochemical model and a multivariate soil-solution mixing model to alpine watersheds in the Sierra Nevada, California

Abstract

A computer model that was developed to simulate the hydrology and geochemical reactions at Emerald Lake in California can be applied to three of the four alpine lake watersheds in the Sierra Nevada, California studied in this investigation. The four watersheds were less intensively characterized than the Emerald Lake watershed for which the model, called the Alpine Lake Forecaster (ALF), was originally designed. Reliable parameter estimation for this model requires that samples be collected over the broadest possible range of discharge from the watershed, such as from baseflow to peak snowmelt discharge. A sensitivity analysis of the geochemical formulation of ALF indicates that the streamwater chemical response to acidic snowmelt is highly dependent upon the proportion of more readily weatherable minerals that are contained in the bedrock. The slope of the regression line relating the sum-of-basecations concentration to silica concentration may serve to quantify the relative sensitivity of the watershed to acidification. To further elucidate hydrologic flowpaths in alpine basins, a multivariate mixing model was developed that explains variation in streamwater chemistry as mixtures of soil solutions. Because soil-solution data were available only from Emerald Lake, the mixing model was developed for that site. The results indicate that two soil environments the bench meadow and the inlet meadow or ridge site may explain most of the observed variation in stream chemistry, and indicate that there is a source of calcium to the surface water that has not been identified.