Regression-based probabilistic seasonal streamflow forecasts for North American river basins

GeoNetwork record: www.gwfnet.net/geonetwork/srv/eng/catalog.search#/metadata/24e9818b-865f-492f-a0fa-f79718c547be

Canada-wide snow water equivalent (SWE) historical station data (collected by ECCC; http://data.ec.gc.ca/data/climate/systems/canadian-historical-snow-survey-data/?wbdisable=true ) was used as input to a statistical model (linear and logistic regressions) to generate probabilistic seasonal streamflow hindcasts for Canadian river basins.

This project supports the hydrological forecasting theme of the GWF Core Modelling and Forecasting Team. As part of the Global Water Futures project, the computational hydrology group builds tools to simulate and predict hydrologic processes. This work focuses on the last point and aims to: -Set up a North America-wide sub-seasonal to seasonal ensemble streamflow forecasting system -Assess the predictability of streamflow on seasonal timescales across North America

Using information about the amount of snow in winter, we make scientific predictions of the river flow in North American rivers for the following spring to summer months. These predictions can give an early indication of the likelihood of a high or low flow event (e.g. floods or droughts). A partir d’informations sur la quantité de neige en hiver, nous faisons des prévisions scientifiques du niveau d’eau dans les rivières d’Amérique du Nord pour les prochains mois du printemps à l’été. Ces prévisions peuvent donner une indication précoce de la probabilité d’événements de niveau d’eau élevé ou faible (e.g. inondations ou sécheresses)

Case studies: Bow River at Banff, Kootenay River at Fort Steele, Fraser River at Shelley, Yukon River above White River, Stewart River at the Mouth, Grand River near Dundalk, Richelieu (Riviere) aux Rapides Fryers, Gatineau (Riviere) aux Rapides Ceizur. The long-term plan is to produce forecasts for the river basins across North America.