Artificial intelligence (AI) has unimagined potential uses that can prove useful in the fight against climate change – such as the major challenges that wildfires, floods and severe droughts can pose. The irony, however, is that AI itself is not carbon neutral.
With climate change, farmers, to name just these workers, are facing more numerous and complex challenges than ever. The higher temperatures not only lead to more and more diseases, but the harvest calendar is also thrown into disarray. In southern countries, plagues of locusts (pest locusts) are becoming increasingly common, threatening food security.
For reading and listening:
A new tool called Kuzi – in honor of the Swahili name for the European starling, a bird known for eating grasshoppers – is giving farmers a boost by providing real-time data from satellites and ground-based weather observations, as well as surface temperature, soil moisture, vegetation index etc.
Kuzi displays a heat map of high-risk areas along with breeding, swarming and migratory attack forecasts that can alert farmers and ranchers to a possible invasion two to three months before the event. The tool can even send SMS alerts when locusts are likely to attack farms in an area, including livestock farms.
Recognizing diseases in potatoes
The AgriRobot, developed by Charan Preet Singh, a master's student in the University of Prince Edward Island's Department of Sustainable Design Engineering, is a robot that is trained using AI to detect diseases in potato plants.
Open in full screen mode
The University of Prince Edward Island's AgriRobot was trained to detect diseases in potato plants using artificial intelligence.
Photo: University of Prince Edward Island
This small, four-wheeled black device with two outstretched arms moves through a row of green leaves, its huge tires churning up the earth. A map is created with the information necessary to identify and eliminate infected plants, explains Aitazaz Farooque, associate dean of the School of Climate Change and Adaptation at Prince Edward Island University.
Beavers fighting drought
With the information provided by AI, conservationists will also be able to more easily identify new locations for beaver reintroduction.
As droughts, floods and fires increase, beavers are helping to make all three worse, said Emily Fairfax, an assistant professor of geography at the University of Minnesota whose research focuses primarily on the ecological role of beavers.
Open in full screen mode
Emily Fairfax specializes in ecohydrology, the interactions between water and ecosystems.
Photo: Emily Fairfax
In fact, beavers are incredible engineers – building dams, ponds and wetlands that conserve millions of gallons of water.
By storing large amounts of water both above ground in ponds and underground, beavers create these large spongy areas in the landscape from which plants can absorb water in times of drought – and which are simply too wet to burn in dry times. There is one Fire phase, says Emily Fairfax.
These structures also reduce the impact of erosion and flooding, she adds.
Detect forest fires
Germany-based Dryad Networks has developed solar sensors that can detect a fire even before a flame breaks out.
Open in full screen mode
Solar sensors from the German company Dryad Networks are used to detect fire.
Photo: Dryad Networks
Behind [la] The membrane contains a gas sensor that responds to hydrogen, carbon monoxide and volatile organic compounds. […] It's like an electronic nose trained to detect the smell of fire.
The company has already deployed 20,000 of its sensors worldwide, with a pilot in a portion of California's forests.
Forecasting extreme weather conditions
As the risks of climate change increase, engineers and scientists are developing new tools to detect and even predict extreme weather events that have become part of our reality and are likely to occur more frequently in the coming years.
Whether it's helping us better predict solar and wind energy on the grid to better integrate it into grids, or helping us map deforestation and emissions using global satellite imagery, AI is coming to the rescue in a variety of ways used for climate protection measures.
The irony is that while AI helps adapt and mitigate climate change, it has its own problem with carbon dioxide (CO2) emissions. While it may seem invisible to most people, the computers that run it are housed in data centers that require a lot of power. If that electricity comes from a fossil fuel grid, AI in turn contributes to emissions.
Additionally, the servers in these data centers generate a lot of heat and need to be cooled, which often requires more power.
The other side of the coin
However, the computers that run the AI are housed in data centers that produce a lot of heat and require a lot of electricity. Despite the fact that AI has enormous potential, it is imperative that we have a better idea of its contribution to emissions, experts argue.
Open in full screen mode
Experts do not yet know what contribution artificial intelligence makes to emissions from data centers.
Photo: Shutterstock / Oleksiy Mark
We really need to pay attention to AI’s growing emissions footprint. Fundamentally, there is insufficient transparency between the data center providers and the machine learning units that create algorithms to monitor and measure greenhouse gas emissions.
Of course, energy requirements vary depending on usage. A recent study suggests that AI uses enough energy to charge a cell phone every time it generates an image.
Priya Donti therefore emphasizes that we need to re-evaluate what uses are worth the electricity consumed and review our practices. “We should definitely not believe that artificial intelligence has no environmental impact or cost,” she concludes.
With reporting from CBC News' Nicole Mortillaro and Sheena Goodyear