As companies around the world grow concerned about the risks of climate change, they have started looking for clarity on how warming might disrupt their operations in the future.
In Charleston, South Carolina, where the ports have been expanding to accommodate larger ships sailing through the newly widened Panama Canal, a real-estate developer named Xebec Realty recently went looking for land to build new warehouses and logistics centers.
But first, Xebec had a question: What were the odds that the sites it was considering might be underwater in 10 or 20 years?
After all, Charleston has repeatedly suffered major floods that can paralyze cargo operations. And scientists warn that flooding will worsen as sea levels rise and storms strengthen with climate change.
Yet detailed information about the city’s climate risks proved surprisingly hard to find. Federal flood maps are based on historical data, and won’t tell you how sea-level rise could exacerbate flooding in the years ahead. Scientific reports on global warming, such as the National Climate Assessment, can tell you that heavy rainstorms are expected to increase in the Southeast, but they won’t tell you whether specific roads leading to a given warehouse might be unusable during those storms.
Most Read Nation & World Stories
So Xebec turned to a Silicon Valley startup called Jupiter, which offered to analyze local weather and hydrological data and combine it with climate model projections to assess the potential climate risks Xebec might face in Charleston over the next few decades from things like heavier rainfall, sea level rise or increased storm surge.
Although Jupiter’s forecasting skill remains unproven, Xebec was eager to participate in a pilot project.
“If we could have reliable predictive analytics in this area, that’s a huge impact for our business,” said Scott Hodgkins, an executive vice president at Xebec.
As companies around the world grow concerned about the risks of climate change, they have started looking for clarity on how warming might disrupt their operations in the future. But governments in the United States and Europe have been slow to translate academic research on global warming into practical, timely advice for businesses or local city planners.
Now some private companies, like Jupiter, are trying to fill the gap.
This remains a young and untested field, and it’s unclear whether Jupiter or others can succeed as profitable enterprises. Scientists caution that predicting short-term climate effects in specific locations remains rife with uncertainty. Jupiter will have to persuade potential customers that its forecasts are reliable enough to give companies a competitive edge.
“In economics, information has value if you would make a different decision based on that information,” said Matthew E. Kahn, an economist who studies climate adaptation at the University of Southern California. “Is that the case here?”
Some insurance companies, such as FM Global, already study climate risks and consult with clients on how to make their buildings more resilient to hurricanes that may get stronger in the future. In 2014, a startup called Coastal Risk Consulting opened in South Florida to offer flood assessments to homeowners nervous about rising seas.
Jupiter, founded in 2017 by Rich Sorkin, a longtime tech entrepreneur, wants to go a step further. The startup has received $10 million in venture capital so far and has been hiring climate scientists, weather modelers and data experts from places like the National Oceanic and Atmospheric Administration. Its co-founders include Todd D. Stern, the lead climate envoy in the Obama administration, and Jeff Wecker, the chief data officer for Goldman Sachs.
The company is developing a variety of predictive tools, some of which look much like Google Maps, that it hopes will allow paying customers to zoom down to the city block level to get a better sense of the potential risks they face from storms, heat waves, wildfires or other climate-change effects in the coming decades.
“We know the planet’s getting warmer and sea levels are rising, but on a hyperlocal basis, the quality of those predictions can be much better than it is,” Sorkin said.
To create its flood maps, for instance, Jupiter looks not just at public data like satellite-based observations of rainfall and ocean currents, but also how changes in the urban landscape affect how water flows through cities. It then aims to harness recent advances in cloud-based supercomputing to combine that data with the latest climate model projections. The company’s scientists plan to continually test their forecasts against observations — to see, for instance, how well they predict flooding from major storms — and publish their research in scientific journals.
Jupiter’s scientists will have to grapple with a number of technical challenges. While current climate models can provide broad statistical projections of how average temperatures and rainfall patterns are likely to shift across large regions over the coming century, it remains difficult to predict such shifts precisely over shorter time scales — which is what companies are often most concerned about.
“Forecasting at 10-20 year time periods is perhaps the most difficult period to forecast,” Simon Mason, a climate scientist at Columbia University’s International Research Institute for Climate and Society who is not involved with Jupiter, wrote in an email. “If that is not enough, trying to predict severe weather events rather than long-term averages is even harder still!”
Jupiter, which acknowledges the uncertainties in climate forecasting, will have to prove that a market exists. But at least one firm in the insurance industry sees potential value in the company’s approach, particularly after flooding and other disasters caused $306 billion in damages last year in the United States and record losses by insurers.
“That certainly raised the stakes in terms of trying to get the best possible science on your side when you’re pricing risk,” said John Drzik, president of global risk at Marsh, one of the world’s largest commercial insurance brokers, which is currently in talks with Jupiter to explore what types of data and risk analyses might be most useful to its clients.
Drzik noted that many of the traditional catastrophic risk models used by the insurance industry are rooted largely in historical data and don’t always grapple fully with how climate change could shift those risks in the future. While his company is still evaluating whether Jupiter’s climate-oriented models are useful enough to be worth paying for, “it grabbed us as something that had a lot of promise.”