
“If you compare land to ocean—if you had the ratio of buoys from the ocean on land—Canada would have two weather stations,” says Sebastiaan Ambtman, chief executive of OceanSync, speaking from Halifax on the eastern seaboard of North America, where his firm is working to turn ships into moving weather stations.
The goal at present is to deliver better data to those at sea, protecting lives and saving fuel in the process. But, Mr Ambtman says, improved understanding and prediction of what’s happening at sea offers huge benefits for those on land, too.
Pointing to the 2024 hurricane season as an example, he talks about both the cost to lives and the economy when a huge storm wreaks havoc. He also talks about the consequences of shutting down, only for a storm to change path or weaken. “We’ve experienced situations where we were under a hurricane warning—schools and businesses were shut down, everybody prepared—but then the storm took a different path. It never hit, but it had a huge impact,” he says. “And then obviously, you get the other side too: where there's an under-prediction and maybe no alert is issued. The more data we can collect out on the ocean, the better we can feed the models that make those predictions.”
Speaking the same language
Marie Drevillon, head of the evaluation of analysis and forecasting systems at Mercator Ocean International, explains that getting the various bodies involved in ocean and analysis and prediction onto the same page is a challenge. “One of the difficulties is to get the various communities—from those doing observations for space agencies, to people in laboratories on the coast or at big, national agencies—to work together to deliver actionable data for the billions people who depend directly or indirectly on the ocean and its resources, from fishing and aquaculture, to shipping, to renewable energies or even for early warning systems for coastal hazards.”
Speaking from the OceanPredict Symposium in Paris in November, Ms Drevillon warns that from observation to the final user is “a very long value chain,” but adds that there is growing awareness of the importance of ocean understanding around the multifaceted threats associated with ocean warming, extreme coastal events and marine pollution. “The rise in extreme events on land and in the ocean does mean people now understand better that the ocean and climate are linked and that these phenomena can be forecasted—and that early warning systems do exist,” she says. “We can use this information better and we can make these systems work efficiently.”
According to Ms Drevillon, what is needed is for the ocean-forecasting community to rely on a co-ordination body and an infrastructure designed to facilitate close co-operation between ocean-forecasting centres and their interaction with ocean-observations systems—as well as weather-forecasting systems. “The ocean-forecasting community is mature and provides the capacity to forecast marine hazards,” she says. “We’ve known for a long time that ocean forecasts improve weather forecasts, but we’ve only recently been able to do this in practice thanks to a close collaboration between disciplines and across operational centres.”
This practical coupling of ocean prediction and weather forecasting is already delivering real results. Ms Drevillon points to improvements in predictions around everything from tropical cyclones to harmful algal blooms and marine heatwaves as examples.
“Marine heatwaves have a tremendous impact on biodiversity,” she explains. “We can already forecast marine heat waves up to 10 days in advance—we’re now working on forecasting them on a seasonal scale.”
Ms Drevillon also talks about an even longer-term focus. “We are working on better understanding the impact of the Intergovernmental Panel on Climate Change (IPCC) scenarios for the next century on the ocean—but at a local level. What’s going to be the sea-level rise locally? What’s going to be the ocean-heat increase locally? It’s about making the link between short-term forecasts and longer-term predictions for the future of marine health.”
A relatively new science
The challenges of ocean prediction are obvious: gathering data on what happens in the high seas is simply more difficult because it is harder to set up weather stations than it is to do so on land. This means it is really a relatively new science. According to the UN, “the development of ocean forecasting began in the early 1980s, with operational services being implemented in the early 2000s.”
But real-time transmission from ships—the OceanSync system, for example—is being added to data gathered by everything from specialised aircraft, satellites, drones and unmanned aerial vehicles, some of which are designed to fly directly into storms.
Even as climate change drives more extreme weather, coastal populations continue to boom. Currently, 2.15bn people live in near-coastal zones and 898m in low-elevation coastal zones around the world, according to researchers who predict these numbers could rise to 2.9bn and 1.2bn, respectively, over the course of the 21st century.
The challenges posed by a more extreme ocean—with greater numbers of people living in potential impact zones—is increasingly being understood, as are the ocean’s connectedness to Sustainable Development Goals (SDGs) that go beyond Life Under Water.
But, writing in the UN’s State of the Ocean Report 2024, Vidar Helgesen, executive secretary of the Intergovernmental Oceanographic Commission (IOC) of UNESCO warns “the ocean crisis is developing faster than our knowledge of it,” calling for a “faster, more dynamic interplay between ocean knowledge, policy and action.” He talks of a “need to transform ocean science and our relationship to it. We need better knowledge as a basis for sustainable ocean planning and management, within and beyond areas of national jurisdiction.”
That report is the latest in the State of the Ocean series, which launched only in 2022, and talks about the need for greater mapping of ocean floors—more than 75% of which remain uncharted. It notes that almost 90% of tsunamis have been generated by large earthquakes, and restates the goal that 100% of at-risk communities should be prepared for and resilient to the danger posed by tsunamis by 2030. It refers to the ocean’s influence on SDGs 11, 13 and 14—respectively, to make cities and human settlements inclusive, safe, resilient and sustainable; to take urgent action to combat climate change and its impacts; and to conserve and sustainably use marine resources. Better ocean weather predictions are needed to improve work on SDG 12 (ensuring sustainable consumption), SDG 6 (clean water and sanitation) and SDG 7 (affordable and clean energy).
An AI leap
As well as improvements in data gathering, the UN is also working on improving data sharing. The Ocean Decade Corporate Data Group works specifically to ease the sharing of vital data, developing frameworks and mechanisms to make privately owned ocean-science data publicly available for the benefit of humanity.
Mr Ambtman also points to developments in AI as holding the potential to expand our understanding of and our ability to predict ocean weather.
“We have really smart scientists, we have really advanced computer models,” he says. “The bottleneck is the data.” Even as AI advances rapidly into ocean and weather forecasting, a 2023 study showed how Google DeepMind took less than a minute to predict future weather worldwide more precisely than other approaches, and using just a desktop computer. However, both Mr Ambtman and Ms Drevillon note the continued need for human oversight and traditional science.
AI models are accurate but not yet mature and Ms Drevillon points to the tracking or pollutants, or search and rescue situations, as examples of where these models aren’t yet accurate enough.
“AI is going to improve many aspects of our existing systems,” she says. “But we still need the science, we still need the traditional approaches to really understand what is going on at sea.”