Great Lakes Runoff Inter-comparison Project, phase 2: Lake Ontario (GRIP-O)

Abstract

The Great Lakes Runoff Inter-comparison Project for Lake Ontario (GRIP-O) aims to compare different hydrologic models, using the same settings, in their ability to estimate runoff for the Lake Ontario watershed. The watershed is challenging because many of its tributaries have a regulated flow regime and a significant part remains ungauged. GRIP-O follows the GRIP-M project which focused on Lake Michigan. It involves a comparison between two different sources of precipitation data (CaPA - Canadian Precipitation Analysis and the GHCND - Global Historical Climatology Network - Daily), and focuses here on two lumped models, GR4J (modele du Genie Rural a 4 parametres Journalier) and LBRM (Large Basin Runoff Model). Results indicate that both models perform very well, with GR4J performing slightly better than LBRM and the GHCND precipitation dataset resulting in better simulations than CaPA, for this area. Performances are, however, always very satisfactory whatever the combination of model/precipitation data used, even for regulated catchments, and do not show any clear correlation to any of the catchments' properties studied here. Results also tend to confirm that the Area-Ratio Method is appropriate for extrapolating flows from the gauged part of a catchment to the whole catchment including its ungauged parts, as demonstrated in GRIP-M.