Isotopic time-series partitioning of streamflow components in wetland-dominated catchments, lower Liard River basin, Northwest Territories, Canada

The distribution of stable water isotopes provides valuable insight into runoff generation processes in subarctic wetland regions of the Mackenzie River basin, a major freshwater contributor to the Arctic Ocean and the focus of intensive hydrological research as part of Canada's contribution to the Global Energy and Water Cycle Experiment (GEWEX). This article describes a streamflow hydrograph separation analysis carried out over three complete annual cycles (1997-1999) for five subarctic catchments ranging in size from 202 to 2050 km(2) situated near the confluence of the Liard and Mackenzie rivers. This heterogeneous landscape, characterized by extensive wetlands (fen and bog), shallow lakes and widespread discontinuous permafrost, is representative of vast flow-contributing areas of the upper Mackenzie Valley, and is suspected to be highly sensitive to climate variability and change. We document seasonal patterns and interannual variability in the isotopic composition of local streamflow, attributable to mixing of three distinctly labelled flow sources, namely groundwater, surface water plus rain, and direct snowmelt, and apply these isotopic signals to partition sources and their temporal variability. Although groundwater input is the dominant and most persistent streamflow source in all five catchments throughout the year, direct snowmelt runoff via surface and shallow subsurface pathways (during spring freshet) and surface waters from lakes and wetlands situated in low-lying areas of the basins (during summer and fall) are also significant seasonal contributors. Catchment-specific differences are also apparent, particularly in the generation of snowmelt runoff, which is more attenuated in fen-dominated than in bog-dominated catchments. The data set additionally reveals notable interannual variability in snow isotope signatures and snow water equivalent, apparently enhanced by the 1998 El Nino event. Copyright (c) 2005 John Wiley & Sons, Ltd.