Darren Sutton

Organizations such as the Toronto and Region Conservation Authority (TRCA) could benefit from a flood forecasting method that uses GIS tools and spatial data because they have already invested in GIS software. This research project utilized a 30-metre resolution Digital Elevation Model (DEM), shapefiles of land cover and soil type, and a Landsat 7 Enhanced Thematic Mapper (ETM+) satellite image to delineate hydrological response units. Snowpack depths were simulated using data from The Weather Network and modelling using the Guelph All-Weather Storm Event Runoff (GAWSER) equations. The Curve Number Method was used to predict peak discharge and time to peak discharge.

Snowpack depth was overestimated in some cases, which is likely due to error in snowfall simulations and underestimation of the melt factor. A systematic error, which was caused by averaging of rainfall inputs was corrected for by reducing simulated peak discharge by a factor of 10. The model is most adept at simulating peak discharge for a 71.6 mm rainfall event. For smaller events, a multiplier was introduced to improve the model's performance. Snowpacks did not have a significant impact on peak discharge although runoff increased slightly. Underestimation of time to peak for some simulations is due to the effects of tillage and error in estimation of channel flow velocity. Despite some errors, a workable model was obtained that could be useful to organizations like the TRCA.