Modelling transport of dissolved silica in a forested headwater catchment: the effect of hydrological and chemical time scales on hysteresis in the concentration-discharge relationship

The relationship between concentration c and discharge Q in a stream is one of the aspects of hydrochemical catchment response that has been used widely as a diagnostic. In particular, loops in the c–Q curve, commonly referred to as hysteresis loops, are used to infer particular mixing patterns. At the South Fork of Brokenback Run (SFBR) in the Shenandoah National Park, Virginia, we have evidence that stream dynamics reflect a system composed of an ephemeral subsurface stormflow zone perched above a perennial water table. The relationship between dissolved silica and stream discharge exhibits hysteresis in the clockwise (CW) direction. Modelling this relationship in the context of three-component mixing with constant-concentration end members failed to reproduce the observed c–Q pattern. In this paper we examine the possibility that temporal variation in soil-water concentrations of silica can explain how CW hysteresis loops in the c–Q curve can arise. In particular, we examine the role of the ratio of a hydrological time scale to a chemical time scale in determining the nature of hysteresis loops. The CW loops that we observe in SFBR can be explained by time variability in soil-water concentrations only if the chemical (leaching) time constant is less than (or only slightly greater than) the hydrological time constant. The near equality of these time constants is consistent with reports from hydrological measurements and leaching experiments. Copyright  2001 John Wiley & Sons, Ltd.