Environment Committee on April 9th, 2024

[The Vice-Chair, Bloc, Monique Pauzé]

I now call this meeting to order.

Welcome to meeting number 102 of the Standing Committee on Environment and Sustainable Development. Pursuant to the order of reference of Wednesday, February 14, 2024, the committee is commencing its consideration of Bill C-317, An Act to establish a national strategy respecting flood and drought forecasting.

This is the first time I have chaired this committee. I would like to welcome Mr. Patzer, who is replacing Mr. Mazier for a few minutes as the latter is in the House and will be back soon. I will be happy to turn the chair over to him at that time.

I should remind you that today's meeting is taking place in a hybrid format, pursuant to the order adopted by the House on June 15, 2023. To ensure an orderly meeting, I would like to outline a few rules to follow. I believe that witnesses and members are now familiar with the Zoom application and know how to access the interpretation and that they must raise their hand to request the floor. All the sound tests have been completed, as agreed.

I therefore give the floor to an eminent member of this committee, Francis Scarpaleggia, member for Lac-Saint-Louis and, especially, the sponsor of Bill C-317, which we are considering today.

Mr. Scarpaleggia, the floor is yours for 10 minutes.

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

Thank you, Madam Chair.

First, I would like to thank Dr. John Pomeroy for being here today at the table with me.

Dr. Pomeroy is the Canada research chair in water resources and climate change. He is the inspiration behind this bill. He helped to draft this bill, and he patiently taught me about flood and drought forecasting.

Today, though, I will be concentrating on flood forecasting for simplicity, but I know that Dr. Pomeroy would be happy to take questions about drought conditions in Canada and the art of drought forecasting.

I would like you go back in your mind to when you were in high school, where you drew two-dimensional graphs consisting of the horizontal and the vertical axes, both of which are used in the flood forecasting process.

I will call the x-axis or the x-dimension the top-down process in flood forecasting. When I speak of the top-down process, I'm talking about weather forecasting now. It's top-down. It's not particularly democratic, but it's the technology that dictates that weather forecasting is a top-down process. Images come from satellites down to earth. They form the basis of weather forecasts that are fundamental to flood and drought forecasting.

Technology dictates this top-down process in weather forecasting. Weather forecasting is done by the Meteorological Service of Canada, whose main operations facility, the Canadian Meteorological Centre, is housed in a nondescript building covered with parabolic antennas situated along Autoroute 40 in Dorval, on Montreal Island. Should you ever drive by it, you will definitely see a mass of parabolic antennas on a roof, which you may readily, but incorrectly, think is a sports bar that requires all the antennas to receive signals from various sports events.

The horizontal axis represents the collaborative process that involves the provinces and territories. To forecast floods accurately, you have to do more than forecast the weather. You need data on water levels and flows in water bodies, lakes, rivers and brooks. Data collection necessarily requires the participation of the people closest to those phenomena on the ground, in other words, the people in the provinces and territories.

Water level and flow data are collected by water gauges and stations located in water bodies across the country, water bodies that are in provincial and territorial jurisdiction. As we know, water resources belong to the provinces and territories, and this is done under a collaborative and jointly financed federal-provincial program called the national hydrological service.

Climate change, as we all know, means more frequent and intense flooding and drought events. We therefore need a more accurate flood forecasting system, with greater lead times to allow communities to better prepare for floods and to better buttress against them.

Fortunately, flood forecasting methods and technologies are rapidly evolving. Technology now allows us to build more sophisticated flood forecasting models that cover larger geographic areas, to which we can attach probabilities of flood risk. Because these models cover larger areas, greater co-operation is required among jurisdictions—among provinces and between provinces and the federal government—because, as I said at the very start, the federal government is responsible for weather forecasting and has some in-house capabilities that are important to flood forecasting in terms of gauging water flows and water levels, etc. It's a collaborative process involving provincial, territorial and federal governments.

Because these models cover larger areas, greater computing power is required to run complex, dynamic models. We're talking here about supercomputers. Supercomputers are required for this kind of large-scale, complex probabilistic forecasting.

There are many benefits to this co-operation, which is, in fact, already a reality. There are many benefits when flood forecasters work together. As Dr. Pomeroy has pointed out, while some modellers in Canada may predict major floods every year, others may not be called on to predict even one flood in their entire career, so a national collaborative approach would create opportunities for shared experiences and professional development among flood forecasters in different provinces. There is already a community of forecasters who meet to discuss best practices, but we need a more formal, permanent structure to harness flood forecasting knowledge from jurisdictions across Canada and to better move forward together.

Some will say that this structure already exists within Environment and Climate Change Canada and that the department is already involved in flood forecasting. However, to quote Dr. Pomeroy, “ECCC has not established a national hydrological forecasting service. They have established a federal river flow forecasting system over part of the country that is exploratory and top down. But it does not have the participation of the provinces and territories.”

Nowhere in the Canadian Meteorological Service's published core mandate does one find a requirement to engage with provinces and territories on flood and drought forecasting. What is being developed currently is a federal system of stream-flow prediction, not a co-operative flood forecasting system that is national in scope. There is a difference.

At the moment, the Canadian Meteorological Service's stream-flow models help predict the volume of water passing through an area under different weather scenarios. This information is useful in decisions about irrigation and hydroelectricity generation. Stream-flow models are also being used to improve weather forecasts. For example, if the Canadian Meteorological Centre predicts that stream-flow will be high because the soil moisture in a basin is high, resulting in greater runoff because the land is not absorbing rain, this will be an indicator of greater expected evaporation from the soil. Evaporation, in turn, influences atmospheric dynamics. With high levels of soil moisture and evaporation, the energy and water in the atmosphere are greater. This, in turn, influences the weather and the weather forecast.

However, to forecast floods even more accurately, we need a far more refined approach and more local data.

True flood forecasting involves the added complexity of factoring in soil moisture, groundwater saturation, the state of glaciers and snowpack, the topology of the river network, river pinch points, and the state of river bank erosion. Human decisions pertaining to the regulation of dams and reservoirs must also be factored in.

In Canada, flood forecasting is further complicated by the fact that we are a northern country with mountainous regions and lots of ice. River ice jams can raise water levels many metres above the norm. These don't occur only during spring ice cover breakup; they can happen in the fall freeze-up and mid-winter breakup periods as well.

Another factor impacting flood risk that has been recently highlighted by Dr. Pomeroy is wildfires, which of course we've seen a lot of in the last couple of years. Wildfires destroy tree canopy, leading to much denser snowpack as more snow accumulates on the ground. Loss of canopy also translates into less shade and more direct sunlight hitting the snow, causing it to melt faster. The result is earlier and stronger spring runoff and a greater risk of downstream flooding.

The purpose of this bill isn't to reinvent the wheel but rather to urge the federal government to adapt and respond more effectively to the latest scientific developments and flood and drought forecasting methods at a time when floods and droughts have become more frequent and severe as a result of climate change.

If we look to Europe, we see that even independent states can achieve the unity of purpose required for accurate flood forecasting on a continental scale.

According to Dr. Pomeroy, and I'm sure he'll speak more about this, the prototype European flood forecasting—

[The Vice-Chair, Bloc, Monique Pauzé]

Mr. Scarpaleggia, I'm sorry but your time is up.

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

I can't believe it, but all right. I'm going to let Mr. Pomeroy speak later about what's been done in Europe and tell us how relevant it is for Canada.

[The Vice-Chair, Bloc, Monique Pauzé]

Thank you very much, Mr. Scarpaleggia.

Mr. Pomeroy, welcome to the committee for a second time. A reminder to the committee that you are the Canada research chair in water resources and climate change at the University of Saskatchewan. The floor is yours for five minutes.

[Dr. John Pomeroy, Canada Research Chair, Water Resources and Climate Change, University of Saskatchewan, As an Individual]

Thank you very much.

Eleven years ago, my ideas around what would become this bill began to jell when it rained for three and a half days over the mountains west of Calgary, Alberta in late June 2013. Two hundred and fifty millimetres fell on a late-lying snowpack, and the flood started. We had 15 people in the field from the University of Saskatchewan, including several professors who were colleagues. What we found was absolutely incredible. The generation of these floods was in the mountains, and they rushed down towards Canmore, High River and eventually Calgary. What we did not see in time, even after the evacuations were starting in Canmore.... Where was the flood warning for the province of Alberta that a massive flood was on the way? Four people died in that flood. Over $5 billion in damages occurred in the region. It was the most expensive natural disaster in Canadian history at that time.

Almost a year and a month later, in Saskatchewan, which had had only snowmelt flooding since it incorporated as a province in 1905, the rain started. In eastern Saskatchewan, over 200 millimetres of rain caused rainfall-based flooding in basins that only had snowmelt flooding, at a time of year when the creeks are normally dry and farmers are looking after their growing crops and all that. Again, this overwhelmed the provincial capability, which is designed for snowmelt flooding—which you can plan for by watching the snowpack accumulate.

These were incredible lessons. We had to better understand flood forecasting in Canada.

The other lesson, as Mr. Scarpaleggia mentioned, was one from the European Centre for Medium-Range Weather Forecasts. It was running an experimental product in 2013 that gave a reasonable estimate of the magnitude of the Calgary flood 10 days before it happened. They didn't communicate it to Canadians. It was just a test product, but it showed what was possible and gave us an aspiration for what Canadians could do if we brought our technologies together and worked together as a country on this exceedingly difficult problem.

Flood and drought damages have risen dramatically in Canada since then and are expected to rise further due to extreme weather and water events, thanks to climate change coupled with our growing communities and increasing agricultural and industrial production. Flood plains are growing, droughts are intensifying and many community farms and industries are impacted by this. In 2022, the “Aquanomics” report estimated that, up to 2050, GDP loss in Canada due to droughts, floods and storms will total $174 billion. In the Global Water Futures program, which I direct, we estimate that Canadian damages from these events since the year 2000 have exceeded $40 billion only up to last year. Things are getting worse.

How do we deal with this in Canada? Prediction in Canada follows a piecemeal governance approach. We have provincial and territorial systems developed bottom-up, which work to meet local needs, and a federal system developed top-down from the weather forecast system, as mentioned. Neither is interoperable and neither meets the full suite of current needs we have in the country. This fragmented approach has led to slow adoption of new technology and methods by the provinces and limited uptake of the more sophisticated federal system. There is a desire and need for common modelling frameworks, common approaches and coordinated forecast systems. This is what countries like the United States do. This is what Europe does. This is what other major countries do.

At Global Water Futures, we established, with the help of Environment Canada, a pilot forecast demonstration project for the Yukon territory. We developed a state-of-the-art prediction system for the Yukon River basin and transferred this to the Yukon government for its operational forecasts. Technical challenges in running such a complex hydrological computer system meant that we have been running the system for the Yukon government since 2018. Remember, the Yukon territory has 40,000 people. It doesn't have the technical expertise, so far, to run a system like this without assistance. A federal-provincial-territorial co-operative system could far better ensure that resources and technologies are available to support operational forecasting and prediction from these co-developed systems.

I have a few recommendations for how we might have a more coherent flood and drought forecasting and prediction framework in Canada.

One, the framework should be developed to coordinate local, regional, federal and international efforts—remember, we have shared river basins with the United States—and enable the authorization of state-of-the-art scientific and technological advances in forecasting and prediction.

The national framework should be co-developed with both a top-down and bottom-up approach to be mindful of local realities and to build credibility and trust between academics, users, and government policy and practice—

[The Vice-Chair, Bloc, Monique Pauzé]

Mr. Pomeroy, I unfortunately have to interrupt you, but you will have an opportunity to finish with your recommendations as you answer the members' questions.

I now yield the chair to Mr. Mazier.

[The Vice-Chair, Conservative, Dan Mazier]

Thank you.

I guess we start off with questioning.

Mr. Leslie, you have six minutes.

[Branden Leslie, Conservative, Portage—Lisgar, MB]

Thank you, Mr. Chair.

Thank you for bringing this bill forward. Speaking as a representative of Manitoba, we have been prone to floods for centuries, but they're getting worse of late, thanks to some of our neighbours and some land use changes in particular. I recently had the chance to speak with Manitoba's hydrological forecast centre to get a bit of a better understanding of how we could benefit from a national forecasting service, because obviously we have a desire to control our water flows and to retain water and mitigate floods. I'm not saying there were any concerns from that conversation, but I'm hoping for some points of clarity.

I'll start with the sponsor, but perhaps Mr. Pomeroy can aid in some of the responses.

One of my concerns is that if we have a federal service, there may be confusion with what comes out of the province. Particularly in Manitoba, as a leader in this space, we rely on those flood forecasts. Would there be two streams of information, or would it be a national, federal organization sharing with the province for them to distribute the flood forecasting? The concern would be that if you have two different sets of forecast information, it might not align, which would lead to some of the challenges you've alluded to already.

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

Is that for me?

[Branden Leslie, Conservative, Portage—Lisgar, MB]

It's for whoever is best placed.

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

Well, I will pass it to Dr. Pomeroy, but I would expect that the coordination would prevent those kinds of overlapping situations from occurring. I think part of the objective here is to rationalize, enhance and improve the system of flood forecasting and drought forecasting.

My understanding is that this new structure would actually prevent that from happening. Is that correct?

[Dr. John Pomeroy]

Yes.

In discussions with the provincial forecasters and the federal government, we think the best proposal is to have the high-end, exceedingly complex computer models run federally and tied into the weather forecast, but then tailored to the needs of the provinces so that they can issue the forecast as the one authority of source in there.

It also takes into account multiple model outputs. Remember, there's nothing preventing Google or others from doing this as well. We can get lots of different forecasts, and that adds to the confusion. The idea is to coordinate so that there's something authoritative and this is what we go with.

[Branden Leslie, Conservative, Portage—Lisgar, MB]

As it stands right now, are the provinces and territories harmonized in the statistical techniques they use for the hydrological concepts? Are we currently aligned, or would there be a necessity through this body to have the provinces work with the same processes from a statistical standpoint?

[Dr. John Pomeroy]

In 2019, with the help of Environment Canada, we convened a meeting of the 13 territorial and provincial forecasting groups. They had never met before. We started an informal community of practice to share things. They were all building the wheel separately, by themselves, and everyone said they didn't have enough resources to do it.

There's a lot to be gained by working together. That's where perhaps a federal initiative can help grease those wheels to help it happen. I think a lot of the activity will be provincial, but it varies tremendously, depending on jurisdiction. We have big, rich provinces that have fairly sophisticated systems, and we have smaller provinces and territories that are using Excel spreadsheets. It's pretty different, that capability right now.

[Branden Leslie, Conservative, Portage—Lisgar, MB]

One of the things I learned in talking with our province's forecasters is that the systems are pretty sophisticated, but in the areas of high population. Northern and remote locations don't have that same sort of technology. Would this be a way to have those gaps filled so that more northern and remote indigenous communities might have access to the same level of forecasting information?

[Dr. John Pomeroy]

Yes. I think the objective would be continental-scale forecasting and prediction so that even the really remote parts of the north would have this.

The other thing I would say is that the system can predict many things, including soil moisture, forest fire likelihood due to duff layer moisture and aspects of drought. A single system can have multiple purposes, including drought and wildfire risk, which could be really beneficial.

[Branden Leslie, Conservative, Portage—Lisgar, MB]

How would that information be shared? I think that's a good point. If we could actually better understand and mitigate wildfires, how would that information be shared through this national forecasting service, the Canadian forest service or whatever appropriate body that would be?

[Dr. John Pomeroy]

Right now, the federal system is shared with just a few select users. It's not a public system. A co-operative system could, with the imprint of the provinces and the okay of the provinces, be shared more widely in that sense. Again, the crucial thing is to avoid confusion over what the risks are, but to maximize the outflow of that information in the information sharing.

[Branden Leslie, Conservative, Portage—Lisgar, MB]

Would we anticipate that the national forecasting centre would be using LiDAR flood mapping for the high-risk areas, or would you predict that this would be outside of the scope?

[Dr. John Pomeroy]

The hydraulic routing that is usually done using LiDAR flood mapping has been proven to be best done locally. That's in particular where the communities and the provincial governments have tremendous expertise, and that's where their roles, as one example, would come very strongly into this.

[Branden Leslie, Conservative, Portage—Lisgar, MB]

It's one of the challenges, obviously, with predicting climate change and the impacts. I think what's perhaps more easily accessible are the land use changes that have happened over a number of years, but that's also information that's held and known more locally. How would you see the provinces and municipalities working with the national centre to try to share some of that information so that you have the best assumptions in your modelling?

[Dr. John Pomeroy]

With a good community of practice and a shared system that's truly federal-provincial, you will have that information going up into the national model.

For instance, we've been working on wetland drainage impacts on changing hydrology on the prairies. We've developed algorithms that are now in a federal model called MESH, which can look at that over time and say, “As these wetlands are drained, there will be this change in flood likelihood over time.”

[The Vice-Chair, Conservative, Dan Mazier]

Okay.

Did you need anything tabled there, Branden? No. Okay.

For the next six minutes, we have Mr. van Koeverden.

[Adam van Koeverden, Liberal, Milton, ON]

Thank you, Mr. Chair.

Thank you very much for joining us today, Dr. Pomeroy and MP Scarpaleggia. Thank you for bringing forward this very important bill.

My question is for you, Dr. Pomeroy. It sure seems to me that floods, droughts and things of that sort related to extreme weather events happen more frequently these days. It seems as though a lot of that impression might be driven by a 24-hour news cycle and a heightened awareness of and attention to those matters, but is it accurate to suggest that property damage and natural disasters on that scale with respect to floods, droughts and extreme weather on the coast are happening more and more frequently?