The highly urbanized Cooksville Creek watershed, located in Mississauga, Ontario, is approximately 34 km2 and is vulnerable to flooding. The objective of this study was to compare the currently established HECRAS flood lines to those generated by a SWMM5 model using a 1D-2D overland mesh.
Two flood damage centers in Cooksville Creek were identified by CVC as areas significantly affected by flooding and were used for the 2D flood line analysis. An existing 1D Cooksville Creek major system model, developed in 2009, was implemented into a real-time flood forecasting tool, and is used by the Credit Valley Conservation Authority’s (CVC) flood duty officers. The model was imported and geo-referenced 52 existing HEC-RAS modelsinto SWMM5. Historical flow and rain gage and flow data from 2007 to 2009 was used to calibrate the 1D model. One of the limitations of 1D modeling is that the flow splits are not separately identified. These splits are a result of a low lying area close to the flow path and can be hidden in a 1D representation.
Using the recent PCSWMM 2D capabilities the flood line for the Cooksville Creek watershedwas independently investigated. 1D-2D connections were represented as SWMM5 bottom orifices with assigned offsets set to be the difference of invert elevation for the 2D and 1D connection nodes. To avoid duplication of the natural channels in the model each transect was cropped to the identified bank stations. Two 1D-2D models were created; each model included the entire 1D Cooksville Creek model and the 2D overland mesh of the identified flood damage center. The first model was located in the lower portion of Cooksville Creek with the second flood damage center located further upstream. The 2D component for the first flood damage center had 58844 links and 20201 nodes in the overland mesh while the second had 52144 links and 18262 nodes. The paper discusses the differences between the original flood line analysis and the results of the 2D modeling. Clearly a primary advantage of combining the 1D-2D SWMM5 models is the ability to better predict the extent and duration of overland flow.