Why continental scale is such a big deal
Innovative solutions to hard problems
Jeremy Chen developed the first-ever single hydrologic model of Canada more than 10 years ago as a PhD candidate. We talked to him about how modelling has changed since his groundbreaking achievement — and why next-gen continental-scale models like C1W are so important.
SEEN HERE FIRST - Chen’s Canadian Continental Basin Model generated a first-of-its-kind simulation of surface water drainage patterns and mean surface water depth distribution across a single continental domain.
Q: What got you into hydrologic modelling?
Jeremy Chen: I took a modelling course led by [Aquanty co-founder] Ed Sudicky at the University of Waterloo when I was doing my Masters. I learned he’d developed a fully integrated model that simulated groundwater and surface water simultaneously, and I thought that was brilliant because it’s how the world works. There’s only one water cycle, but people had imposed an artificial separation between groundwater and surface water. So I chose to do my PhD with Ed and we got funding to build an integrated continental-scale model. That became the Canadian Continental Basin Model (CCBM).
Q: How was your original model different from C1W?
JC: A lot of the fundamentals are the same, but we had slower computers and our resolution was limited. We could only get down to a resolution of about 30 square kilometres, which leaves out important local details such as smaller creeks and the like. The CCBM output is good for large surficial areas but there’s not a lot of vertical layering. Canada1Water has more nodes, higher grid resolution, better surficial geology data and stronger linkages between climate and hydrology. And our model was entirely steady state, while Canada1Water has a seasonal cycle, which adds a lot of insight and makes the development much more complicated.
“Bigger models with better resolution give you more meaningful answers.”
Q: There seems to be a movement underway to develop continental-scale models: Canada1Water, Parflow-CONUS in the U.S., and two separate initiatives in Europe. What’s driving that?
JC: Partly it’s that the technology is finally available. Twenty or 30 years ago all you could do were tiny catchments, nothing large-scale. Now we have the computing power to build much larger models. The numerical algorithms have become more efficient, and there’s more and better data available, including satellite imagery and geophysical data. And there is also a need for answers to big questions. Obviously, there is a lot of interest in understanding climate change. It’s a very hot topic, and that’s driving the requirement for larger-scale models.
Q: Why is scale so important?
JC: Bigger models with better resolution give you more meaningful answers. All models have boundary conditions to constrain them. The bigger you go, the more accurate the boundary conditions become. And for climate — in the real world, climate has no boundary. The feedback loops are global. So if you want to model climate with any accuracy at all, you need to be working on a semi-continental, continental or ideally global scale.
Q: What kinds of questions can a continental-scale model like C1W answer?
JC: If you’re in Alberta and you care about surface water for irrigation, you may want to know how much you can safely draw from a particular river. Seeing just part of the river isn’t helpful. You need to see the whole thing to represent the entire system. Or if you want to know what’s going on with one of the Great Lakes, you can’t simulate just that one because they’re all connected.
Q: How does Canada1Water measure up against some of the other continental modelling initiatives underway today?
JC: As far as I understand, Parflow-CONUS, the European Parflow-CLM model and Denmark’s work to extend its national model continentally all have sophisticated computer code and fully coupled groundwater and surface water. Canada1Water may be the most physics-based to date, and its regional climate downscaling is definitely very sophisticated. What’s most important is the value it will bring for Canada and Canadians. Researchers who don’t have the resources to build their own models can take excerpts from Canada1Water and add local, site-specific data to do their own simulations. It’s going to answer practical questions and at the same time advance our fundamental understanding of hydrology, climatology and geology overall.
Jeremy Chen is a senior engineer at the Nuclear Waste Management Organization and the developer of the Canadian Continental Basin Model — a predecessor to Canada1Water.