Private Companies Step up to Gather Weather Data for NOAA as Staffing Cuts Hobble Agency Forecasting – Inside Climate News

When staffing shortages caused the National Weather Service (NWS) to suspend weather balloon launches at its Kotzebue, Alaska, station earlier this year, a startup deploying next-generation weather balloons, WindBorne Systems, stepped up to fill the void. 

The company began selling its western Alaskan atmospheric data to the NWS in February, plugging what could have been a critical data gap in weather forecasting. 

Weather balloons collect real-time atmospheric temperature, humidity, wind speed and pressure data that meteorologists use to predict the weather and understand longer-term changes to the climate. The Alaska office was one of about a dozen to suspend or scale back balloon launches in response to deep staffing cuts instituted by the Trump administration’s Department of Government Efficiency (DOGE). Critics claim that the cuts have weakened the NWS’ forecasting capacity as hurricane season bears down and extreme weather events, like the floods that ripped through Texas, claim lives and destroy property.

As the beleaguered weather service struggles to maintain its forecasting and other services, it’s leaning on private companies to pick up the slack. For example, WindBorne, which is backed by Khosla Ventures, a venture capital firm focused on investing in companies with innovative business models and technologies, is opening five new balloon launch sites in the U.S. this year as it expands its work with the National Oceanographic and Atmospheric Administration (NOAA), the parent agency of the NWS.

“We’re flying more balloons every day and collecting more observations to help improve forecasts in light of some of these systems going down,” said John Dean, WindBorne’s co-founder and CEO.

Sofar Ocean Intelligence, Tomorrow.io, Black Swift Technologies and Saildrone are among other startups with innovative technologies and AI forecasting models that are increasingly supplying NOAA with critical atmospheric and oceanic data through its Mesonet Program. Such collaboration isn’t new, but former NOAA officials worry that the current administration, with its zeal for privatization, will jettison core federal observing systems and rely instead on private sector data to forecast the weather. While they lauded the companies’ innovations, they said that NOAA must maintain ownership of its “backbone” data assets like weather balloons to ensure public safety and maintain the historical climate record. New technologies, they said, should supplement NOAA’s core data collection efforts rather than wholesale replace them. 

“NOAA has always had a robust relationship with the private sector exactly for the sorts of things that WindBorne does,” innovate and supply data, said Tom Di Liberto, a meteorologist and former NOAA spokesperson who is now media director at Climate Central. Under the current administration, however, “the concern is…what is it going to replace?”

If private services take the place of, rather than supplement, the agency’s core data assets, that could prove problematic, because “less data is bad,” he said. “Are we actually saving money or just giving taxpayer dollars to a private company?”

Data as a Service

In the past NOAA bought sensors and hardware from companies with promising innovations to bring the technology in house. More recently, it’s adopted a model of “data as a service,” in which it buys data from companies that maintain their own hardware and intellectual property rights. 

“While that can be fruitful for everyone, what I worry about is becoming so dependent on some of these innovative solutions,” said Rick Spinrad, who led NOAA during the Biden administration. “What happens when the founder [pivots]?” 

The agency also needs more staffing to effectively manage the growing use of commercial data, he said. “There’s a contradictory nature to what this administration is doing, advocating for private sector delivery of data and then removing a third of the weather service. Who’s going to manage these programs and make sure they’re effective?”

NOAA already lost access to a vital tool developed by Saildrone for improving hurricane forecasting and warning accuracy because it didn’t issue a request for contract proposals far enough in advance of hurricane season.

And there are risks that come with some of the technologies the agency is becoming reliant on when they are proprietary and unique to an individual company. 

Agency dependence on one company for critical services or data is especially worrisome for Brad Colman, a private meteorologist who previously worked at NOAA. “It’s a vulnerable position because you now have data that you have built your forecasting system around,” he said. The company could demand more money, which could limit NOAA’s ability to invest elsewhere, or have the business challenges it faces affect the product it provides the government.

“There’s a contradictory nature to what this administration is doing, advocating for private sector delivery of data and then removing a third of the weather service.”

— Rick Spinrad, NOAA administrator during the Biden administration

Data ownership is another crucial concern. Historically, NOAA has strived to make the commercial data it buys freely available to anyone who wants to use it for forecasting or research, said Mary Glackin, a former high-ranking official at NOAA who also worked at The Weather Company. 

That’s best for public safety, she said.“There is no weather forecast that’s produced in this country that isn’t dependent on NOAA,” she said.

But free and open data licensing agreements can be costly for the government, and companies often want to retain some data to sell to private buyers. In those situations, NOAA may buy data for its own purposes but withhold it from forecasters outside the agency for a set period.

The first Trump administration showed a willingness to choose this latter option. A contract negotiated in 2020 with a company that had what many considered a superior hurricane forecasting model, for instance, constrained NOAA from publicly releasing the forecasts for five years, drawing criticism from hurricane specialists and private forecasters.

WindBorne’s Innovation

WindBorne’s AI-guided balloons stay aloft for months and collect vastly more data in the upper atmosphere than traditional weather balloons, which only fly for about two hours before popping and descending back to earth.

Called radiosondes, after the instruments they carry, traditional weather balloons cover just a fraction of the Earth, because it is logistically challenging to launch and receive data from them over the oceans and in remote areas.

WindBorne’s balloons, in contrast, can collect and distribute data from remote regions. That makes them more adaptive, and especially useful for monitoring atmospheric rivers that bring extreme precipitation to coastal regions, said Glackin. “I’d like to see them in the suite of observing systems.”

The company deploys about 100 balloons from six launch sites globally, a fraction of the 92 launch sites operated by NOAA, but it aims to expand to launch up to 10,000 balloons globally over the next five years, Dean said.

Windborne’s data are less costly than radiosonde data “on a per observation or per station basis,” Curtis Marshall, the director of the Commercial Data Program for the NWS, wrote in an email. 

And while its data are now free and open to the public, as the company expands, it wants to hold some back some of the information it gathers for 48 hours so that it can sell it to private buyers, Dean said. That data would no longer be useful to other forecasters.

Radiosondes’ Old School Technology is Difficult to Replace

Radiosondes collect one vertical profile—a line from ground level to the point where the balloon explodes—of data in the atmosphere, which is important for understanding climate change signals. WindBorne’s balloons, in contrast, collect thousands of data points, at different altitudes, across a horizontal expanse. Their path is somewhat ad hoc, determined by where the wind blows them, whereas radiosondes collect data in a line rising from a location that stays the same for each launch. 

While WindBorne’s lack of a consistent path doesn’t matter for short-term weather forecasting, it could matter for understanding longer term changes to the climate, which are currently based on decades of vertical profile data collected at the same spot, Glackin said. WindBorne’s data would not be comparable with that historical record.

 “We have a very cleaned up climate record that allows us to talk about how the climate is changing,” she said. “If all the radiosondes went away tomorrow, it would be hard to figure out what’s changed, and what [to] attribute to technology versus what really happened in the atmosphere.”

There are methods for transitioning to new instrumentation, Colman, the meteorologist who used to work at NOAA, said, but the NWS would need to proactively plan for that changeover to maintain a consistent data record.

The NWS isn’t moving to replace radiosondes—yet—but it is in the “early stages” of planning for a new suite of upper atmospheric observing systems that would provide data “substantially similar to the federal radiosonde network,” Marshall wrote. 

The new observing systems would come from commercially operated balloons, drones and aircraft, and “complement our federal balloon network.”

However, Austin Tindle, a co-founder of Sorcerer, a WindBorne competitor, said that officials within NOAA are increasingly asking him, “what it could look like to be a true replacement to a radiosonde.”

“It’s been a vibe shift recently, coming up in conversation a lot,” he said.

WindBorne’s Dean declined to respond when asked if he’d been having similar conversations.

NOAA’s partnership with WindBorne “could be completely on the up and up, [meaning an add-on rather than a replacement] but folks don’t have a lot of trust in the broader strategy for the NOAA weather enterprise, based on everything that’s happened,” said Di Liberto, citing the agency’s June 25 announcement that it was permanently ending—within just five days—a vital microwave satellite program used for forecasting hurricanes.

Dean at Windborne is none too eager to replace core NOAA functions. “You’re better off augmenting than you are replacing traditional weather balloons, but we want to fill gaps wherever they form,” he said.

He’s not alone. Tindle, whose solar-powered balloons are smaller and travel further than WindBorne’s, said that Sorcerer “was never intended to be a replacement” for radiosondes, but to cover places in the world with no traditional balloon launches.

One reason for private weather monitoring companies’ caution about how much service they provide government lies in the directive that federal agencies have to serve the public, which is sometimes mismatched with business interests.

“The mandate of the government is not ours,” said Tim Janssen, a co-founder of Sofar, which has created a network of buoys that deploy from vessels and aircraft with sensors to monitor ocean conditions. “It would be impossible for us to spend millions of dollars to do something just for societal benefit, [when] there isn’t a direct business case.”

Sofar provides shipping companies with forecasts to help them plan the safest and most fuel-efficient routes and partners with the U.S. Navy, NOAA and the European Centre for Medium-Range Weather Forecasts.

“The biggest concern for us over the last months has been the [lack of] awareness around the importance of those partnerships,” Janssen said, noting an increasing attitude of, “let’s just rip it out and industry will take care of it. That is a nonsensical idea.”

But the NWS may find itself backed into a corner, with limited options to gather, analyze and distribute critical weather information.

“If I were NOAA, I’d be looking for a cost-benefit analysis on [WindBorne], although the politics are outpacing everything else and now your back’s up against the wall,” said Glackin. “But I don’t think they can dance into the secretary’s office and say, here’s the answer to all our problems.”