The climate is changing, storms are becoming more intense, and more people and property than ever are in harm’s way. Three of the costliest five hurricanes on record, Harvey, Irma, and Maria, struck U.S. shores in 2017 at a price tag of over $200 billion.
The costs of weather disasters and climate change may balloon in coming years if current trends continue. Businesses and government organizations exposed to these threats could well face staggering losses that jeopardize their bottom line and sustainability of operations.
This week, a high-powered, well-funded start-up company has barged onto the scene to help businesses and governments confront their increasing vulnerability to climate change and weather disasters. Using cutting edge technology, it could revolutionize how they receive information about weather and climate threats and make critical planning decisions.
Known as Jupiter, the company was founded in 2017 by Rich Sorkin, a serial entrepreneur who has worked with Steve Jobs, Bill Gates, and Elon Musk. Sorkin was Musk’s first boss as chief executive of Zip2, a company that provided business directories to online newspapers in the late 1990s.
Over the past year, Sorkin and his team have plucked climate and data scientists from top research institutions and corporations while hiring information technology professionals from the likes of Google and General Electric.
The weather and climate risk industry has not seen a company enter the space with this mix of financial backing, technology and intellectual heft in years, if ever.
“The risks related to severe weather and climate change are among the greatest challenges of our generation,” Sorkin said in an interview with The Washington Post. “There are many different pieces of that problem. We’re focused on quantifying those risks in a way that lends decision-makers across-the-board understanding of the human health and safety and economic consequences and [help them] do a better job planning for and mitigating them.”
Headquartered in Silicon Valley and with branch offices in New York and Boulder, Jupiter has 22 full-time employees and has secured $10 million in venture capital. Its advisory team is led by Neal Wolin, who served as deputy secretary of the Treasury Department in the Obama administration; Todd Stern, who was Obama’s chief climate negotiator at the State Department; and Jeff Wecker, chief data officer at Goldman Sachs.
Sorkin said he plans to expand internationally, first in Europe in the next year and in Asia in 2019. The company is working feverishly to roll out products tailored to specific locations, “expanding city by city at an ever accelerating pace.”
Jupiter has developed tools to help customers plan for hazards two hours to 50 years into the future all the way down to the street or even building level. The foundation is its ClimateScore platform, which leverages cloud computing to run and link multiple prediction models that ingest data from millions of ground-based and satellite sensors.
The first two products in Jupiter’s portfolio are called FloodScore and HeatScore. They’re intended to help customers manage the risks associated with intensifying precipitation events and heat waves in a warming world.
FloodScore and HeatScore predict not only precipitation and temperature changes, but also simulate their interactions with the built environment and the surrounding landscape and how they’re altered by climate change. In the assessing flood risk in New Orleans, for example, the analytical framework would take into account the convergence of wetlands and concrete and sea level rise, Sorkin said.
Jupiter will introduce FireScore, that evaluates the risk of wildfires, and other services later this year.
Sorkin believes Jupiter’s use of cloud computing will enable it to expand its products and services at a breakneck pace. “Our ability to chain together multiple models in the cloud in a statistically rigorous way, allows us to go much, much faster — from both a development and deployment perspective,” Sorkin said. “It let’s you go from one customer in one city to hundreds to thousands of customers all over the world fairly quickly.”
Its modeling will also employ machine learning to supplement underlying data and improve the representation of physics-based processes like clouds, whose behavior is very difficult to simulate, and interactions between the natural and built environment, a flood at a chemical plant, for example.
Sorkin pointed out many available climate information tools furnished by government-led assessments are not updated regularly enough to meet the needs of businesses. He said Jupiter’s simulations will be available on demand and based on the latest science and available data.
“The Intergovernmental Panel on Climate Change and regional climate panels do an excellent job but run about every 5 years,” he said. “They tend to be academic, they’re often not statistically rigorous, and there’s no consistent methodology across them. They’re not designed for the private sector. We’re taking the same models and people [used in those assessments] and integrating into a regularly run reusable framework and making them available at scale.”
Sorkin added this same framework will differentiate it from existing approaches that rely on the “fundamentally flawed” assumption that the climate of the past can simply be extrapolated into the future.
Jupiter has already begun working with property developers, critical infrastructure, insurance companies, and the public sector. Sorkin anticipates additional interest from city, state and national government to support solutions for health, safety, and planning as well as intelligence and defense.
The demand for climate services is growing, but a reasonable question is how reliable will information be at a local level for threats years to decades into the future. A 2016 academic paper reviewing the current state of climate services noted many users seek this information but that the “predictability of climate variability on such timescales remains a debated question, and the provision of reliable predictions remains a scientific challenge.”
Sorkin, who says climate change is “real,” at the same time says “its impact in specific locations and at different time scales, as well as the confidence we should have in different predictions, is often poorly understood.”
Jupiter’s products will include information about how well its models are performing so that its users understand their strengths and limitations.
Betsy Weatherhead, an atmospheric scientist from the University of Colorado and one of Sorkin’s early hires, said she is most excited about “translating fundamental science about climate to decisionmaking possibilities.” In an interview, she lauded Jupiter as a “really pure” science-focused culture where providing “the most honest and reliable answers” is a priority.
Weatherhead is joined by Alan Blumberg from Stevens Institute of Technology who built a flood prediction system for New York City and Josh Hacker from the National Center for Atmospheric Research, a pioneer in running weather models in the cloud, among several other climate scientists on Jupiter’s team.
Jupiter is bringing together very smart people to the weather and climate marketplace and new, potentially transformative technology. But David Titley, the director of Penn State’s Center for Weather and Climate Risk and not involved in the company, says it will need to demonstrate its worth.
Titley said competition in the space includes CoreLogic, an information intelligence firm out Irvine, Calif., that conducts property-level risk assessment for the insurance industry and addresses both storm and fire hazards. He also mentioned Coastal Risk Consulting out of Plantation, Fla., which provides hyper-local flood risk assessment.
For Jupiter, “the proof will be in execution, and why a client should believe Jupiter’s projections versus another group’s,” Titley said. “But there is likely a large market for this type of information. If Jupiter can deliver on their sales pitch, they likely can be successful.”