Towards a cold regions hydrometric streamflow uncertainty framework for Canada

Abstract:

The estimation of hydrometric uncertainty is relatively data-intensive, relying on river cross-sectional information, high-frequency measurements, and metadata describing any shifts in river morphology. These data are rarely available for the Canadian hydrometric network. A framework has been developed based on remotely sensed river characteristics and reanalysis gridded climate products to generate a first estimate of hydrometric flow uncertainty for all hydrometric streamflow data in Canada. The framework breaks uncertainty into four “seasons”: open-water, river freeze-up, ice-on, and river break-up using meteorological driven transitions. Results suggest reasonable estimation of freeze-up and break-up timings can be accomplished with meteorologic thresholds and logistical regression. Coastal gauges with near-zero temperatures for extended periods were more difficult to model and the determination of ice jams remains as a challenge for integrating into uncertainty estimation.