How do antidepressants affect fish? Can biological remediation improve water health (e.g. intentionally planting cattails in dirty waterways)? How can the Internet of Things (IoT), blockchain, and artificial intelligence (AI) innovate freshwater research? These are just a few of the questions researchers at International Institute for Sustainable Development’s (IISD) Experimental Lakes Area (ELA) have been tackling over the last few years.
Founded in 1968, ELA is one of the world’s most influential and oldest freshwater research facilities. Data from Winnipeg-based ELA helps researchers examine human impacts on the environment as well as informing national and international policy. As the global population increases and countries continue to feel the intensifying impacts of climate change, this environmental research on freshwater is urgently needed.
“We now have 50 years of intense research on lakes that are now becoming almost a climate-observatory,” says Geoffrey Gunn, former Policy Advisor for Data and Technology at IISD. “By analyzing lake data, we can start to see the more subtle influences of climate change.”
The following conversation between ICTC’s Research and Policy Analyst Maya Watson and IISD’s Geoffrey Gunn is the latest instalment in ICTC’s Green Tech Series. Their discussion dives into the potential of emerging tech to clarify water data and protect the environment.
Maya: Tell me about what drew you to this intersection of water data and emerging tech.
Geoffrey: My first encounter with big data, before we called it big data, was my undergraduate thesis where I used remote sensing to understand grassfire scars. We received United States Geological Survey (USGS) Landsat satellite images on DVD. Processing each one took hours (the “cloud” had not yet taken shape). From there, I did my master’s degree on Arctic climatology, which means even more big data. In geography, we need to understand what’s happening in the atmosphere, in the oceans, and with people. The challenge of fusing these disparate datasets together and my interest in sustainable policy brought me to IISD Experimental Lakes Area.
Maya: At IISD-ELA, you led a research project that used AI and other new technologies to rethink data collection and manipulation in the context of climate change and diminishing freshwater supply. What types of applications did you explore?
Geoffrey: We started researching fintech and the environment, diving into AI, blockchain, big data, and the IoT. For example, we looked at using radio-frequency identification (RFID) tags that World Wildlife Fund (WWF) Australia and New Zealand use to track fish to demonstrate sustainable methods from the catch to the supermarket using blockchain. Then, we realized the most important piece: we need to start getting insights out of water data quicker. AI can help turn data into information that any of us as citizens could use to make better decisions or for scientists to understand some of those deeper complexities as systems change.
Maya: What’s a good example that highlights AI’s potential to make better decisions for freshwater?
Geoffrey: My favorite examples are ones where people have used AI to predict algal blooms in reservoirs and lakes. With warming waters due to climate change in North America, we are seeing more algal blooms, some of which contain neurotoxins produced by cyanobacteria in addition to green scum that depletes oxygen. Industrial phosphate detergents, fertilizer run-off, and manure from livestock remove limitations to algal growth and cause these blooms. That really throws the ecosystem out of whack. In 2014, the City of Toledo became a famous and expensive example because they had to shut down their entire city’s water system for five days at tremendous cost, due to a harmful bloom. Even in smaller cases where cities only shut down briefly, algal blooms can clog filters and damage equipment.
Algal blooms are a big problem, not just for drinking water but [also] for recreation, fisheries, and whole ecosystems. Having AI predict those blooms would be helpful. But in most cases, we don’t have the data to do that, even though in theory we could.
Maya: Why don’t you have enough data?
Geoffrey: Ecology, historically, is a field where a single researcher collects her data, publishes it, and then moves on to the next thing. In other words, environmental data collection is driven by a peer-reviewed journals. While articles now require more publication of data, people just put it on any archive. There are so many places to put data on the internet that even if it is open and published it gets lost. Data is dark because it’s not findable. It’s buried in an archive somewhere or worse — a research notebook.
“Data is dark because it’s not findable. It’s buried in an archive somewhere or worse — a research notebook.”
Maya: How should we solve for this lack of data?
Geoffrey: That’s our focus at ELA over the next few years. We call it accelerated ecology. We’re asking: how can we help people trust what’s going on in the environment and understand it better? And that means turning ecology from a dark data science to one where the data is out, open, and available to anyone.
But we’ve got to make sure open environmental data has a universal identifier or that it’s in a structured dataset. It also needs to be standardized. That way other researchers spend less time fiddling with data and more time doing science. Making data findable means having an authoritative copy, but also making it available in different places. For example, after publishing a paper, it’s a good opportunity to promote the data on social media. Making data open access allows your data to have new light, a new life beyond its intended use.
Maya: Do you have any examples where open data has made a difference in your research?
Geoffrey: Manitoba’s highway traffic conditions cameras are mostly used by travellers wondering how much snow is on the highway to determine whether they should make the drive or by people who decide to send out a plow. Instead, we used basic object detection to see where people were moving trailers and boats to understand how boats move from lake to lake — potentially taking aquatic invasive species like zebra mussels with them. This use was not the data’s intended purpose. We gave a new value proposition to that data stream. You don’t know what your data is good for unless you put it out online and make sure it’s open access.
Maya: How can the government help?
Geoffrey: The way we collect ecological data is somewhat antiquated. With governments, it’s often collected to ensure that we don’t go over thresholds or to ensure we hit this legal obligation of collecting data. Canada was a leader in environmental monitoring data for quite a long period of time, but after some cuts in the 1990s and 2000s, there are fewer resources allocated to it. We need more water data coming from governments, and leadership from national governments and international organizations to develop data standards for environmental water chemistry. On a more positive note, it’s getting better. I’m impressed with how many governments have opened their data and are sharing on repositories like open.canada.ca or the newly-launched Great Lakes Datastream.
Maya: What’s next for environmental data?
Geoffrey: I think it’s a question of how we — whether that’s as individuals, as governments, as communities — use these technologies in ways that help us chart that better pathway forward. AI, IoT, blockchain, when they’re connected, they allow us to do that.
Think about how many weather stations are now around the world. If you have just one, you will know what it is like outside right now, and maybe whether to expect sun or cloud in the next few hours. But a network of weather stations (and with computational power and human expertise behind the scenes), we can save lives by predicting storms better than ever. What’s next? Why not a weekly forecast for our environment, including freshwater?
We’ve made environmental science mainstream before. Let’s do it again for the environment.
Want to learn more about sustainable tech that supports Canada’s freshwater? ICTC’s inaugural Horizon: 2022 Digital Futures Summit this February includes a panel on Fishing for Sustainability, featuring experts from industry, academia, and government (register here).
Can’t wait until February? Read more research about sustainable technology now in ICTC’s newly released report “Canadian Agri-Food Technology: Sowing the Seeds for Tomorrow” and policy brief “Designing Smart and Sustainable Communities.”
