Citation
  • Stadnyk, T., van der Eerden, J. (2024). Peak Flow Assessment under a Changing Climate: Recommendations for the Canadian Highway Bridge Design Code. Canadian Standards Association, Toronto, ON.

Executive Summary

A leading cause of bridge failures is due to scour and erosion around bridge piers and abutments. Suitable protection against these hazards, as well as determining the required hydraulic opening size of bridges that span watercourses is largely dictated by the estimated peak design flowrate. This flowrate allows practitioners to manage flow velocities and acceleration through the bridge site, to ensure adequate bridge soffit clearance is provided, and to assess in-stream and overbank flows. By providing an adequate bridge span and hydraulic opening, scour, erosion, and over-topping can be managed. In the past, design flow estimates were heavily dependent on historical records and typically relied on completing frequency analysis of historical observed streamflow data or a proxy thereof. These historical streamflow-based estimates are becoming increasingly less valid under a changing climate, particularly for bridge infrastructure that is intended to last for decades. For non-rainfall-dominated watersheds (i.e., watersheds in which peak bridge-related flooding cannot be linked directly to individual rain events), hydrological modelling of the contributing watershed is required to estimate future condition peak flow. This approach can overcome the assumption of climate stationarity, which is a key assumption of traditional hydrologic assessments. However, hydrologic modelling of watersheds can be a significant undertaking and the level of effort to build and analyze a climate-driven watershed hydrological model may not be commensurate with the risk imposed at the subject bridge site.

The research conducted for this report investigated the range of simplified and standardized approaches that would allow practitioners to estimate future condition peak flow at bridge sites across Canada. This report is the first step toward an initiative with the aspirational goal of developing upper bound factors that can be applied to frequency analysis of historical records to reflect the impacts of climate change. Although the underlying flood-generating mechanisms may be changing, developing “bounding curves” and resulting “delta-change factors” can be justified by comparing peak flows from modelled historical and future climate windows. The delta-change factors could then be applied to estimate conservative future condition peak flows at other locations within the same hydrological region. Thus, a practitioner could use traditional peak flow estimating procedures, such as single station frequency analysis or regional approaches, and apply a delta-change factor that represents the maximum change in peak flow expected to occur based on a specific emission scenario and time horizon (collectively referenced as degrees of global warming). Delta-change factors would provide the basis for resilient design (i.e., low probability of exceedance) in the absence of detailed, site-specific hydrological assessments driven by future climatic parameters. In turn, this would provide a level of consistency for completing peak flow assessments for bridge design in a changing climate, which would benefit owners and practitioners alike by reducing engineering costs while still incorporating hydrological expertise into the assessments.

While the ultimate objective is to develop a geographical dataset that allows practitioners to apply a delta-change factor to historical flood frequency analysis of naturalized open water conditions to account for climate change, this report is limited to investigating the feasibility of such an endeavour and providing recommendations for an approach and the expected limitations. The investigation included a review of relevant studies and considered whether they could be leveraged or emulated at a national scale. Due to the rapid evolution of hydrologic studies that span a variety of spatial scales, landscapes, and hydrologic conditions, this report identified the most current, robust, and relevant approaches, and included interviews with a variety of agencies and key individuals responsible for peak flow estimation.

This report concludes that a national dataset of bounding curves can be derived from a Canada-wide modelling effort that performs a climate change impact assessment for peak flow estimation. From these bounding curves, delta-change factors can be developed for various hydrologic regions or basins across Canada and used to scale up historical, at-site practitioner-derived peak flow estimates for bridge design to account for climate change impacts. The recommended approach achieves a balance between an adequate amount of rigour for climate adaptation measures, as recommended by the Intergovernmental Panel on Climate Change (IPCC) sixth assessment report (AR6), while recognizing the practical limits of practitioner resources, time, and budget constraints.