When a heavy-duty storm approaches, the general approach is to either seek shelter or head in the opposite direction. But a team of atmospheric researchers at Colorado State University packs up a bunch of specially equipped DJI Matrice drones into vans and drives directly toward nature’s fury.

The team is known by the unusual acronym CSU C3LOUD-Ex, or CSU Convective Cloud Outflows and Updrafts Experiment. Its goal? To collect data from inside a storm while it is gathering. It’s data the group hopes will both further scientific understanding of extreme weather, and also aid in more accurate forecasts. In the case of extremely severe weather, more accurate predictive models with greater lead time could clearly save lives.


You can think of a gigantic storm, particularly the super-cell type storms that can spin off tornadoes, as kind of like an engine. The fuel for this engine is an updraft of warm air that is sucked up the centre of a developing weather system. It’s that upward flow of warm air that creates those towering cumulus clouds with the cauliflower-like tops. When they keep growing in height, and the tops retain that rigid, rippled look – it’s a pretty reliable sign of an impending storm. (BTW, this is one of the key things that storm chasers chase!)

Once it gains a little more structure, the storm’s “exhaust” comes in the form of much colder air (the “cold pool”) that rushes downward, often accompanied by rain or hail, along the storm system’s outer edge. (Having once accompanied storm chasers through tornado alley, we can attest to the awe-inspiring power of a well-structured storm. You can actually feel the warm updraft – while simultaneously observing the rain or hail spitting out at the far end. And when that end hits you, there’s generally an incredible temperature differential – and even windshield-cracking hail.

But what’s going on up there? What’s happening *inside* that updraft and cold pool? And what is the precise relationship between those factors and the eventual outcome? Well, that’s a pretty tricky thing for researchers to figure out. Weather balloons can gather some data, but are at the mercy of the winds once launched.

A DJI Matrice takes off, heading toward an intense weather system. It’s one of six that will record data as an intense storm passes over

Enter drones. Storm-chasers have used them in the past, getting amazing footage of tornadoes while operating from a safe distance. But the use of drones as a data-gathering device for extreme weather is a more recent phenomenon.

Atmospheric scientists have long known that those updrafts and cold pools play a significant role in the intensity and duration of a storm. But in the absence of a significant pool of reliable data, it’s difficult to create new predictive models with a high level of confidence. The program, which involves about 20 students, is led by Professor Susan van den Heever in the Department of Atmospheric Science.

“We don’t do well simulating cold pools in our models,” says van den Heever in a newly published Colorado State University article. “We don’t know their structure, how intense they are, how cold and deep they get. We are making those measurements now to help us evaluate our models.”

And how do drones play into all this? Well, have a look:

All told, the team uses six DJI Matrice drones, which are bristling with various sensors (plus a camera). And while drones are typically restricted in the US by the FAA to flying no higher than 120 meters, the team was able to receive a special waiver allowing flights up to 350 metres in certain areas. In fact, they were able to fly all six drones in formation and then have them hold position as the storm moved over. The result? A wealth of real-time pressure, humidity and temperature data.

“We were able to measure the whole structure of the cold pool,” says van den Heever. “We could see that the cold pools were about 10 degrees cooler than their environment. No one else has ever measured the fine-scale structure of cold pools in this manner, using drones.”

Very cool. And, hopefully the seeds of research that may one day provide more accurate predictive models for potentially destructive storms.