Skip to main content

A Quiet Revolution in Weather Forecasting

Small, relatively cheap and quickly upgradable microsatellites could soon make weather prediction far more reliable

Artist's rendering of a shoebox-size weather satellite, with solar panels extended

Credit:

Image courtesy of Spire Global

This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American


Smartphone users have come to expect yearly, if not monthly, updates to their software. With one click, bugs are eliminated, apps operate faster, and new features instantly appear. Now imagine if the next big update to your phone wasn’t going to arrive for the next 20 years. This is essentially where we stand today with one of the key ingredients for weather forecasting in the United States; the last big weather satellite program took nearly two decades to prepare for launch.

Weather satellites aren’t the same as iPhones, but they are similar in how they affect us on a daily, and even hourly, basis. The difference of course is that weather has come to be viewed as an unpredictable, mysterious affair, while phones are seen as things that lie within our control. It’s a universal pastime to speculate on the weather, but let’s not forget: literally trillions of dollars are lost each year, and thousands of human lives, due to scary events like floods, hurricanes, and tornadoes.

After a long career in government satellite programs, I recognized that only a diverse network of public and private sector players, including Spire (where I am now a Product Manager), PlanetIQ, GeoOptics, and others, will be able to turn the current fleet of weather satellites—the equivalent of the old Nokia brick-style phones—into something more like the sleek iPhone 6s.


On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.


To be fair, weather forecasts have gotten better over the past several decades. For example, the National Weather Service recently announced the upgrade of two supercomputers, which will triple computing power and result in higher precision weather forecasting.

But even the best forecasting supercomputers— analyzing earth observations at quadrillions of calculations per second—are limited by the quantity and quality of data provided by today’s weather satellites, which are large, aging, and expensive beasts that require decades of work to be improved. Delays, cost overruns, and funding concerns are the norm in these cases as opposed to the exception.

A technician works on a shoebox-size CubeSat weather satellite. Image ourtesy of Spire Global.

The game may soon change with the emergence of commercial weather satellites, which embody Moore’s Law in being smaller, faster, and cheaper. These shoe-boxed sized satellites are tapping one of the most promising (and least-known) new technologies for weather sensing in the past 20 years: GPS radio occultation (GPSRO).

The technology makes use of radio signals continuously broadcast by GPS satellites orbiting the earth, about 20,000 km above earth’s surface. Much closer to earth’s surface, CubeSats listen for GPS signals as they fall over the horizon. Those GPS signals are “bent” around the earth as they are refracted by the atmosphere. Highly accurate temperature and water vapor information is then calculated by measuring how much the signal is bent.

Like every promising breakthrough technology, the story of GPSRO is filled with its share of underdogs in the face of doubt and uncertainty. Back in 2001, the COSMIC Program, of which I was part, devised by the University Corporation for Atmospheric Research (UCAR), NASA, and the National Science Foundation and provided with significant funding help from US agencies and Taiwan, attempted something never done before — launch 6 microsatellites to test the impact of GPSRO for weather forecasting. COSMIC Radio Occultation satellites, and RO in general, promised huge advantages over the legacy government satellites. Each of the 6 COSMIC satellites weighed approximately 70 kilograms (compared to older satellites that can weigh more than 3,000 kilograms) and were projected to cost around $100 million, including ground support, data processing, and the rocket to get them into space. In the space business, this was a revolution, but only if it could be proven to work.

Initially, there were many more doubters than believers. The weather community didn't believe the data would be of value, nor did some people at UCAR. It took a lot of iteration, and lot of evangelizing, to convince people. Yet, the tide turned quickly when the tiny Minotaur 1 rocket launched from Vandenberg Air Force Base in California with the microsatellites onboard. All six deployed safely, on time, and within budget. The world began to take notice. Within weeks after launch, the COSMIC constellation began to deliver data. Testing of the data in US weather models began immediately by a young scientist, Lidia Cucurull

As the results came in, the worldwide weather community was astounded.

GPS-RO demonstrated the largest positive forecast impact of any space based sensor in the past two decades. Early on, people started to recognize the potential for GPS-RO satellites to be built and launched by commercial companies. Many believed that radio occultation would be the perfect “pathfinder” for a new business model in satellite atmospheric sensing, using commercial-level funding and speed of delivery for satellite data. RO could leverage the efficiencies of commercial development like few other technologies in space, which could potentially drive down the future costs of US and worldwide weather forecasting significantly.

NOAA and the US Air Force are currently building COSMIC-2, which will push RO technology to the next level. COSMIC-2 will provide 8,000 to 10,000 high-quality soundings per day – well above the 2,000 soundings per day provided by COSMIC-1. However, recent scientific studies show that we still need many many RO soundings; it is fairly accepted that 100,000 soundings per day will allow for continued improvement in weather forecasting. That’s why we need to push for a global constellation that combines both ultra-high-quality soundings from COSMIC-2 and large amounts of high-quality soundings from commercial satellites to produce RO soundings at levels well above what each can do on its own.

The commercialization of atmospheric satellite data is moving quickly, thanks to advances in the technology and work by Congress to ease this transition but there are roadblocks that may slow progress. Weather has historically been the domain of worldwide governments and while governments are not strangers to public-private partnerships, it’s an arrangement that requires flexibility (especially on the part of the commercial enterprise) and a well-thought-out transition.

How will the advent of commercial weather satellites affect everyone? Expect weather forecasting to improve quickly over the next few years, with accurate forecasts out as far as 10 days. Expect weather and wind states over the oceans to improve dramatically, with a major positive impact on shipping and air travel. Expect the detection of the genesis of hurricanes much earlier with tracking and intensity forecasting improving by more than 50 percent.

The wake-up call that led to the recent NOAA supercomputer upgrade was when the European weather model correctly predicted the storm track of Hurricane Sandy, better than the US model. Hurricane Sandy was the deadliest hurricane in the northeastern United States in 40 years and the second-costliest in  the nation's history.

The ability to get “real-time’ information about the state of the atmosphere and specific, directed information will be huge, and the benefits to daily life are hard to fathom. But let’s agree: it’s clearly better than speculating on the weather, or waiting for the floods to arrive at our doorsteps.

Dave Ector spent the last several decades working on government satellites for the US Air Force and the NOAA, and was the lead program manager for COSMIC-1, a joint satellite project between Taiwan and the University Corporation for Atmospheric Research. He recently joined Spire Global as Product Manager for Spire Stratos, the company's weather data product.

More by Dave Ector