SWMMCanada: An Open-Source Service for Generating Ready-to-Run Urban Drainage Models Across Canada

Zhonghao Zhang 

Urban stormwater modelling is often slowed more by data preparation than by the simulation itself, because rainfall, terrain, land cover, soil, and pipe network data usually come from different agencies, formats, projections, and data structures. This software paper presents SWMMCanada, an open source and standardized model building service that makes Canadian urban hydrological modelling easier in the AI era. Users draw an area on a map, and SWMMCanada automatically gathers open data, prepares the required geospatial layers, and generates a complete EPA SWMM model that is ready to run. It builds the drainage network in one of two ways. In eight Canadian cities that publish their storm network, it uses the real municipal pipes with their measured inverts and diameters. Anywhere else in Canada, it synthesizes a network from the street map. Beyond the current EPA SWMM, DHI MIKE+, and InfoWorks ICM exports, additional modelling platforms can be connected by reading from the same model-ready datastore. SWMMCanada main contribution is to provide a reliable, repeatable, and transparent starting model that can feed downstream agentic modelling tools such as Agentic SWMM and Agentic MIKE Plus for simulation, calibration, uncertainty analysis, and decision support.