Dataset of 18O and 2H in streamflow across Canada: A national resource for tracing water sources, water balance and predictive modelling.

Oxygen-18 and deuterium were measured in streamflow samples collected from 331 gauging stations across Canada during 2013 to 2019. This dataset includes 9206 isotopic analyses made on 4603 individual water samples, and an additional 1259 analysis repeats for quality assurance/quality control. We also include arithmetic and flow-weighted averages, and other basic statistics for stations where adequate data were available. Station data are provided including station code, name, province, latitude, longitude and drainage area. Flow data were extracted from the historical database of the Water Survey of Canada. Details on the preliminary application of these data are provided in "18O and 2H in streamflow across Canada" [1]. Overall, these data are expected to be useful when combined with precipitation datasets and analytical or numerical models for water resource management and planning, including tracing streamflow source, water balance, evapotranspiration partitioning, residence time analysis, and early detection of climate and land use changes in Canada.

A groundwater separation study in boreal wetland terrain: the WATFLOOD hydrological model compared with stable isotope tracers.

Monitoring of stable water isotopes (18O and 2H) in precipitation and surface waters in the Mackenzie River basin of northern Canada has created new opportunities for researchers to study the complex hydrology and hydroclimatology of this remote region. A number of prior studies have used stable isotope data to investigate aspects of the hydrological regime of the wetland-dominated terrain near Fort Simpson, Northwest Territories, Canada. The present paper compares estimates of groundwater contributions to streamflow derived using the WATFLOOD distributed hydrological model, equipped with a new water isotope tracer module, with the results of conventional isotope hydrograph separation for five wetland-dominated catchments along the lower Liard River. The comparison reveals highly promising agreement, verifying that the hydrological model is simulating groundwater flow contributions to total streamflow with reasonable fidelity, especially during the crucial snowmelt period. Sensitivity analysis of the WATFLOOD simulations also reveals intriguing features about runoff generation from channelized fens, which may contribute less to streamflow than previously thought.