Sensitivity of Catchment Transit Times to Rainfall Variability Under Present and Future Climates

Hydrologists have a relatively good understanding of how rainfall variability shapes the catchment hydrograph, a reflection of the celerity of hydraulic head propagation. Much less is known about the influence of rainfall variability on catchment transit times, a reflection of water velocities that control solute transport. This work uses catchment-scale lumped parameter models to decompose the relationship between rainfall variability and an important metric of transit times, the time-varying fraction of young water (<90 days old) in streams (FYW). A coupled rainfall-runoff model and rank StorAge Selection (rSAS) transit time model were calibrated to extensive hydrometric and environmental tracer data from neighboring headwater catchments in Plynlimon, Wales from 1999 to 2008. At both sites, the mean annual FYW increased more than 13 percentage points from the driest to the wettest year. Yearly mean rainfall explained most between-year variation, but certain signatures of rainfall pattern were also associated with higher FYW including: more clustered storms, more negatively skewed storms, and higher covariance between daily rainfall and discharge. We show that these signatures are symptomatic of an inverse storage effect'' that may be common among watersheds. Looking to the future, changes in rainfall due to projected climate change caused an up to 19 percentage point increase in simulated mean winter FYW and similarly large decreases in the mean summer FYW. Thus, climate change could seasonally alter the ages of water in streams at these sites, with concomitant impacts on water quality. Plain Language Summary Scientists use information about the age of water in streams to help diagnose the source of water and what pollution it may carry. Most stream water originates from rain, and its age'' is the time that has passed since it fell from the sky. Old'' water tends to seep into the stream from deep underground where it has been stored and filtered for years or even decades. Young'' water, on the other hand, comes from recent storms and tends to carry pollution from the land surface directly into streams. This study used stream data and computer simulations to understand how the amount of young water in streams changes in response to wet and dry periods at two small streams in Plynlimon, Wales. We show that the proportion of young water is much higher during wet periods and can be surprisingly sensitive to the rainfall pattern. We also show that changes in rainfall due to climate change could significantly increase the proportion of young stream water in winter and decrease it in summer. We conclude that changes in rainfall under present and future climate conditions can significantly alter the age of water in streams and, in turn, the pollution it carries.