Hydrologic modelling and prediction of extreme streamflow events in the Jock River Watershed, Ontario, Canada

Historical and predicted extreme streamflows were analysed for the Jock River watershed, using Generalized Extreme Values (GEV) and Streamflow Threshold Level (STL) methods. The historical streamflow shows earlier snow melt peaks (~10 days), decreased snow melt peaks, and decreases in consecutive drought days (CDD), whereas the summer-fall season peaks and mean annual streamflow (MAS) have been increasing. The Jock River Watershed Model (JRWM) was developed using the Raven modelling framework and achieved a Nash-Sutcliffe of 0.76 with reduced capacities to characterize low flows. Predicted streamflow changes from climate change scenarios include snow melt peaks decreased by 50% and shifted seven weeks earlier by 2099. GEV analysis showed that extremes are decreasing by ~10% for the snow melt peaks, while increases in MAS are continuing and the summer-fall peaks are increasing by ~30%. STL analysis for CDD shows the magnitude and number of CDD events are increasing by ~300%.