April 20, 2015 by Jason Contant, Online Editor
By focusing on riverine flooding in its newly launched G-CAT Canada Flood Model, Guy Carpenter & Company, LLC felt that “it captures the majority of the hazard,” said company research hydrologist Joseph Becker at the flood model launch in Toronto last week.
“[We focus on] identifying those that are most dangerous or the ones that pose the most risk, because that’s where we can concentrate our resources for a better understanding,” he said.
Riverine flooding – which typically occurs next to rivers where floodwaters or waters spill over the banks, or breach a levee and spill into the adjacent flood bay – is one of the easier drivers to measure, Becker noted during his presentation. “We can go out and put a river gauge and estimate the quantity of flow of a particular location on the river – that’s fairly easy to do,” he said during his presentation on Thursday at the Design Exchange in downtown Toronto.
Pluvial flooding, Becker explained, is “when it’s raining so intense the ground can’t absorb it, but it’s not really going anywhere.” Currently, the G-CAT Canada Flood Model evaluates riverine overland flood risk, together with the off-plain component of such events, but does not include pluvial, coastal flooding, flash flooding, mudslide (slope failure), groundwater, sewer back-up or tsunami risk.
“Understanding pluvial or slope failure or coastal failure is far more difficult because now you are not working on an individual one-dimensional river, you’re working on a two-dimensional surface, or even three-dimensional if you are talking about pluvial,” Becker told attendees. “If we start talking about groundwater and backwater or sewer failure, it becomes exponentially more difficult. It’s a monumental task.”
Becker (pictured right) said that flood modelling is broken up into two major components – hydrology (the quantity of water moving through systems) and hydraulics (what actually happens to the water as it’s on the ground). For the G-CAT Canada Flood Model, Guy Carpenter partnered with JBA Risk Management, who has a “local knowledge of flooding in Canada.” The model uses six different return periods of the hazard – the likelihood of a particular flow in a particular river. For example, one hazard map in London, Ont. features train tracks, which are incorporated into the flood map.
“They use a two-dimensional approach to draw essentially a box around a particular area, they feed what is known as a hydrograph in, which is flow versus time, and they propagate that water within that box,” Becker explained. “They put these boxes end on end, every 400 metres and then stitch the final results together. At the end of that, we end up with a 100-year flood model or a 20-year flood or a 1,500-year flood and we make sure we do this at the variety of return periods to make sure we have a full representation of the hazard.”
Guy Carpenter’s model also factors in continental rainfall and hurricane rainfall, with a hurricane-induced precipitation data set and a non-hurricane-induced precipitation data set. “This transforms that precipitation into the flow,” Becker said. “Once we have the river flows, we are then able to transform that into probabilities and when we have the probabilities, we do a peak over threshold,” which essentially grabs all the major pieces. “Once we have [those] peak over thresholds, we can think of them as dots on a 2D surface in 3D time, we start drawing circles around them and saying, ‘These are the events.’”
Flooding in Canada came to the forefront with the June 2013 floods in Alberta, considered among the worst in the country’s history. Guy Carpenter estimates that the floods resulted in insured losses of Cdn$1.8 billion and economic loss estimated at Cdn$6.2 billion.