Environment Committee on April 9th, 2024

[Monique Pauzé, Bloc, Repentigny, QC]

According to an article published in the journal Nature, artificial intelligence could be used to assess water levels with extreme accuracy, even without gauges in rivers. Do you think the use of artificial intelligence involves safety risks, and has that technology advanced far enough for us to do that?

[Dr. John Pomeroy]

Artificial intelligence certainly has a role. These are essentially data-driven techniques, and where these models have been tested in Canada, they haven't done very well because we don't have very much data. We have parts of Canada that have a stream gauging network that's as sparse as in the developing world—this is when you get into northern Canada or the really remote rural areas—so we don't have the data to train it. Further, with changing climate and new types of floods and things like that, the historical data that would train the AI systems is not what we're going to see in the future.

However, they have a role, perhaps in combination with some of the physics-based models that are out there, where they can provide some benefit. Certainly, on the human side of it, they can provide tremendous benefit.

[The Vice-Chair, Conservative, Dan Mazier]

Thank you very much. That's the end of that two and a half minutes.

We have Ms. Collins for two and a half minutes.

[Laurel Collins, NDP, Victoria, BC]

Thank you, Mr. Chair.

Dr. Pomeroy, you mentioned insurance costs, and in the next panel we're going to be talking to some folks who have been looking at areas that can't get insured because of climate impacts.

I'm curious—this could be to Mr. Scarpaleggia or to Dr. Pomeroy—whether you have any concerns that this kind of forecasting could impact Canadians who are in these areas, and their insurability? What responsibility does government have to ensure that insurance is still available to Canadians?

[Francis Scarpaleggia, Liberal, Lac-Saint-Louis, QC]

That's a really interesting question.

I know the flood maps are being redesigned in Quebec at the moment, and it's creating a lot of controversy because, all of a sudden, if you find your house is in what is now considered a flood zone, you lose a great deal of value on your property. I would think that as long as forecasts are accurate and you're getting an accurate picture of risk, then you're not getting a false picture or a picture that could be called into question.

I suppose it could influence insurance costs, but I think the flood itself would have an impact on your insurance, and that's why the government—and I'm not speaking for the government or representing the government, as I said before—did commit to a low-cost national flood insurance program for those who can no longer get insurance. Some of the areas in my riding have been touched by floods twice now—in 2017 and 2019—and people just can't get insurance anymore.

[Dr. John Pomeroy]

There are two things.

First, the insurance companies are already running their models in secret. There are a number of them across Canada. They're all different, and I don't think any of them are very good. Some people were being charged for floods they shouldn't have been charged for.

Second, if we can predict floods, we can reduce damages, and that should reduce costs across the board. Imagine what the damage to crops would be, due to other events as well, if we had no weather forecasts at all. It's the same thing with flood forecasts. If we can reduce damages this way, we should be able to reduce insurance costs.

[The Vice-Chair, Conservative, Dan Mazier]

That's very good. That's it for two and a half minutes.

We will move on to Mr. Deltell for five minutes.

[Gérard Deltell, Conservative, Louis-Saint-Laurent, QC]

Thank you very much, Mr. Chair.

Mr. Scarpaleggia, I'm glad to see you performing this parliamentary role.

Dr. Pomeroy, thank you so much for your testimony. It's very interesting.

Mr. Pomeroy, earlier you used an image that has stayed with me: You said that every province and territory had built its own wheel. In other words, Canada now has 13 wheels. The aim of this bill introduced by our colleague from Lac-Saint-Louis is precisely to unify. As far as you know, and based on what you know about Canada, do you think we're capable of doing that within a reasonable timeframe, or could it take years for those 13 wheels to turn in the same direction?

[Dr. John Pomeroy]

The wheels are all different right now. Some are pretty sophisticated and others are very simple. This will provide tremendous benefits early on for the jurisdictions that have simple systems or no system. The more advanced jurisdictions will also see it. In our discussions with the Province of Quebec, the province saw a benefit to using both the federal and the provincial models to give them some idea of the uncertainty in the predictions there.

We can see immediate benefits all around, and because these systems all exist, I would think a rapid evolution towards improvement will happen. Eventually, perhaps the concerted generation of the next phase—the next generation of prediction models—can occur on a national basis. That will take longer, but it's certainly possible.

[Gérard Deltell, Conservative, Louis-Saint-Laurent, QC]

You also raised the example of the United States of America. You said earlier that, at some point, it will have the kind of system the member for Lac-Saint-Louis would like to apply.

Can you give us some indication as to how we can get useful inspiration from the Americans' experience?

[Dr. John Pomeroy]

There are a number of things there. One, they do not do it all from Washington. They have regional centres. Even though they're federally led, they're based on river basins: northeast U.S., north-central U.S., Alaska, northwest and so on. They also coordinate heavily with the states involved and with other communities. Finally, they have the central resources for the supercomputer systems and the army of scientists they have in different agencies, such as the U.S. Geological Survey, their weather service and the U.S. Army Corps of Engineers. They have to coordinate all of this as well.

They've done reasonably well over time, but they still feel they need to make further and much more rapid progress, so they set up this new co-operative institute to accelerate those improvements.

[Gérard Deltell, Conservative, Louis-Saint-Laurent, QC]

When did they create that attitude to work together on the same front?

[Dr. John Pomeroy]

This goes back to at least the 1960s, and perhaps earlier. You don't really see the states doing this at all except in providing expertise to the federal system. It's a bit of a different constitutional set-up as well.

[Gérard Deltell, Conservative, Louis-Saint-Laurent, QC]

Let's just say that, in the next few months, the bill is voted on and passed in the House of Commons, and then in the Senate. Based on what you know about this legislation and on your experience in Canada and the United States, approximately how long might it take for Canada really to turn together as a single wheel with the 13 provinces and territories and for us to be responsible and effective at the national level?

[Dr. John Pomeroy]

I still see it as 13 wheels, but with one engine supplying extra power to those wheels as they need it, because we have very different provincial realities on the ground. I think we can find immediate gains through data flow, through coordination and even through provinces sharing information bilaterally. Alberta, Saskatchewan and Manitoba all predicted the Saskatchewan River system, so that's certainly an easy win right there to make that better.

The final thing would be the development of systems that are suitable for Canada, where we have frozen ground, snowmelt, ice melt and river ice jams, which other countries don't have. We'll have to find those solutions ourselves and make sure they're in our system, so we're not relying on what other countries have developed.

[Gérard Deltell, Conservative, Louis-Saint-Laurent, QC]

This is my last question. Do you think we can do that with no new money?

[Dr. John Pomeroy]

We can take advantage of programs such as the $78-million Global Water Futures program, which is ending now. It's in the rollout period. When we built Global Water Futures, we knew we'd be building these models. I wanted a home for them. I never thought the home would be the United States. I want it to be here. You have that investment that could be leveraged here with just a few personnel, some more meetings and gaining efficiencies.

I don't think it's expensive, compared to the things I've seen in the federal government.

[The Vice-Chair, Conservative, Dan Mazier]

Thank you very much.

Our last questioner is Madame Chatel.

You have five minutes.

[Sophie Chatel, Liberal, Pontiac, QC]

Thank you, Mr. Chair.

Thanks to the witnesses for being here today.

This is a very important bill, especially since Canada is the only G7 country that doesn't have a national flood and drought prevention strategy. What are the consequences of failure to develop a coordinated strategy with the provinces and territories, for example?

[Francis Scarpaleggia, Liberal, Lac-Saint-Louis, QC]

That's what I wanted to discuss at the end of my remarks, and Mr. Pomeroy raised it as well.

In Calgary in 2013, we used a refined, well-devised and sophisticated model, developed in Europe, in a pilot project to analyze the situation in western Canada. That model predicted a devastating flood eight days before it actually occurred, while the Province of Alberta was unable to do it that far in advance. So the benefits of that model are quite obvious.

[Sophie Chatel, Liberal, Pontiac, QC]

Earlier someone said that there were costs associated with the implementation of a new strategy.

However, Mr. Pomeroy, you mentioned that the cost of inaction on climate was enormous. But with forecasts, models and quick action, we could help communities prevent droughts and floods and avoid their devastating effects on the regional economy. Can you confirm that's consistent with your analysis?

[Dr. John Pomeroy]

Accurate flood predictions can reduce the damages. Obviously, they can't stop the rain and they can't stop the snowmelt, but they allow time for reservoirs to be operated differently, for sandbagging and other measures to be put into place, and for individuals to take action, including getting out of town. There are many things that don't happen with the lack of warning. The expenses are dramatically reduced when adequate flood warning is provided.

This has been shown around the world repeatedly. When we look at the deaths and damages from floods in the developing world, the one in Libya last year comes to mind. There was no warning system operating at the time. There were tremendous, horrific casualties and damages in that case.

We already have small damages in Canada in terms of loss of life and things like this because of exceptional emergency services and all of that, but we can't be completely reliant upon those to reduce that. We have to have that previous warning.

We're already starting to see the benefits of the federal system. It helped work with the Province of Quebec on the floods three years ago. It provided an accurate early warning to that region. That was a test of the system. Again, it's not a formalized relationship across the country. It was a demonstration that we could do at least as well as the European systems when it comes to these predictions.

[Sophie Chatel, Liberal, Pontiac, QC]

Climate change is causing a lot of problems for farmers, who are on the front line. How will this strategy be able to help them plan their season more effectively?

[Dr. John Pomeroy]

Our farmers and water managers are seeing events they haven't seen in their lifetime that are outside of their experience. With a lead time of three months, but particularly six months, that's enough time to order different seed. It's enough time to make decisions on forage and others. It's also enough time for the irrigation districts to make their plans and make announcements about the water that will be available to irrigation farmers so that, again, they can make their plans. There are tremendous economic benefits there if we can do this.

Right now, we have a wonderful drought monitoring system through Agriculture Canada, but it's not really a drought prediction system. The same software and computer models can be used for floods and droughts, and many other things, including agricultural productivity and water supply for hydroelectricity. There are many applications of the system, but it has to be in place and it has to be running.

[The Vice-Chair, Conservative, Dan Mazier]

Thank you.

That concludes our line of questioning.

Mr. Scarpaleggia, congratulations, you made it through the hour of questioning unscathed, relatively.

Thank you, Dr. Pomeroy, for your words and for assisting Francis. You certainly bring the discussion of water up to a whole new level, and our understanding, too. I thank you for that, and for what you do back home in terms of all your research. I've dealt with you before in accessing your research. There are some really good points there that we need to discuss further, that's for sure.

With that, I'll conclude, and we'll suspend for a couple of minutes to get ready for the next hour.