All it takes is a patch of fog to ruin Thomas Swoyer’s day.
Swoyer runs the first and only fully operational hub for researching, testing and developing commercial drones in the United States, from those used for military reconnaissance to ones that may drop packages at your front door someday.
But local weather events like fog, ice, low clouds, or thunderstorms can suddenly pose risks to flights. In fact, about 30 percent of drone flights by Swoyer’s company, Grand Sky, in Grand Forks, N.D., are canceled because of weather.
“If I’m going to deliver a package 20 miles down the road, I have to know what I’m flying into, and weather is my number-one risk,” Swoyer said.
Canceled flights cost money, waste labor hours, slow research and sometimes even delay critical military missions.
Now Grand Sky is looking into how to better predict weather conditions for its drone flights. The solution? Use another drone.
Pending government approval, the “Meteodrone” will launch this fall and be the first drone to record weather data used operationally in the United States. The data collected would then be fed into computer models to improve forecasts for drone flight operations.
The Meteodrone isn’t your average drone. It’s packed with small weather instruments, which can measure temperature, dew point, relative humidity, wind speed and pressure. An onboard camera can take valuable images as storm systems develop and progress.
It’s also designed to withstand a range of intense weather, armed with safety features like heated propellers to ward off icing, and an emergency parachute. A pilot can remotely launch the drone into the atmosphere and send it up to to 20,000 feet, sampling the atmosphere both on its way up and down in a straight line.
“The benefit of the Meteodrone here is being able to get those constant and frequent profiles, particularly when you have interesting or important weather events going on,” said Brad Guay, a meteorologist at Meteomatics, the company that creates the technology.
All of the collected data will feed into a high-resolution computer model for the area around Grand Forks Air Force Base.
Right now, “the data isn’t granular enough. We just don’t have enough data. Our models are not sufficient,” said Don Berchoff, co-founder and CEO of TruWeather Solutions, which will provide decision-support once the high-resolution modeling is completed.
The data will also reveal previously poorly understood weather patterns between the ground and flight level, critical for drone flights across the country.
The Meteodrone will be new to the United States, but it’s already used regularly in Switzerland. Fifteen Meteodrone stations are situated around the European country, collecting data from the atmosphere around the Alps.
“We have done more than 20,000 vertical profiles, thousands of flight hours,” said Martin Fengler, founder of the weather forecast service Meteomatics. In conjunction with the Federal Office of Meteorology and Climatology, Meteomatics has been working on data collection and high-resolution modeling in Switzerland.
Fengler first became interested in the idea of using drones for weather collection about a decade ago in Germany, as he worked to obtain his private pilot license. “I had never experienced a bad weather forecast myself,” he said of his time before becoming a pilot. But many of his drone practice flights were disrupted by poorly forecasted fog or cloud heights. He started thinking about how he could fill the data gap in this key part of the atmosphere, where so much activity takes place.
And there are many gaps to fill.
In large states such as North Dakota, surface weather stations, radar towers, and balloon launch sites are spaced far apart and leave huge weather blind spots.
Other areas are using outdated or broken instruments. The Chatham, Mass., weather balloon launch site, the only site in the six-state region, closed last fall due to coastal erosion, and has yet to be replaced. The National Weather Service says it hopes to have a new site finalized by the end of this summer, but until then, critical data is being missed, potentially impacting the performance of computer modeling and forecasts.
Grand Sky is investing just under $1 million dollars in this new technology, but Swoyer believes it’s well worth it.
With better data collection from the Meteodrone and more accurate high-resolution weather models, he hopes a launch could just be delayed by a few hours, instead of outright canceled.
“If I can get 10 percent more flight hours, it pays for itself in the first year,” said Swoyer. “The real value for me is more flight time, more up time, more mission time.”
For Berchoff, the potential to detect icing with the Meteodrone is particularly exciting.
Icing is incredibly hard to forecast and extremely dangerous. It’s described as one of the biggest risks to drone flight, since ice buildup on the wings can dramatically impact the aircraft’s ability to stay airborne.
“Icing is the key here. We’re going to put that thing up to see if there’s icing,” he says of the company’s plan to regularly sample the lower atmosphere.
Fengler expects drone technology to keep evolving quickly in coming years.
“In about 2 to 3 years from now, we’ll see drones that can fly to 10 kilometers,” he predicts.
Berchoff says drones offer the “best business case ever for detecting micro-weather.”
He’s optimistic that the data from North Dakota is just the start of a wider network for weather drones sampling data entirely missed by forecasters until now.
Michael Page is a Boston-based certified broadcast meteorologist with more than a decade of experience covering meteorology, the environment, and science related news.