Every year, tornadoes claim about 60 lives in the United States. One of their most frustrating qualities is how difficult they are to predict, with weather monitors receiving an average of about 13 minutes notice to take cover whenever one is detected. This is something that Jamey Jacob, a mechanical engineer from Oklahoma State University, hopes to change, by extending warning times for these storms up to an hour or more.
Jacob and his team are developing rugged drones capable of collecting data from these dangerous storms. While meteorologists can see hurricanes and blizzards coming from miles away, tornadoes form a lot quicker, and on smaller scales. Therefore, predicting these storms requires a detailed understanding of atmospheric processes at resolutions you can’t get with conventional weather radar.
Jacob envisions meteorologists using drone swarms to fly into severe weather, where they’ll measure temperature, pressure, and wind velocity at specific points throughout the tornado. Upon being fed into a computer model, this information could determine how likely a thunderstorm is to turn into a tornado. This would enable storm chasers to collect data from a safe distance, who currently use some sensor packages that can collect data from inside the tornado (which have their limitations).
Weather balloons and TOTO sensors dropped in paths of storms largely rely on chance to attain valuable data. Upon being deployed, TOTO sensors can’t even move, and are subject to the sporadic nature of these storms. Insert unmanned aerial vehicles (UAVs), which are steerable and can give a fuller view of a storm. Having said that, drones can’t improve forecasts if they can’t survive tornado-like conditions.
Jacob’s team is developing ruggedized UAVs capable of withstanding conditions like rain, hail, and winds up to 120 mph. These drones will be riddled with sensors, cameras, and made of strong, lightweight, composite materials similar to Kevlar bullet-proof vests. One system named MARIA (Mesocyclone Analysis Research & Investigation Aircraft) can even deploy deopsondes, which are little parachuted sensors capable of collecting data about a storm’s vertical profile as they descend.
To obtain longer-term tornado predictions, researchers will also need better tornado models since tornado formation is largely a mysterious process. Some temperature and wind flow patterns, for example, can form tornadoes in some situations, but that’s not always the case. A large amount of weather patterns can lead to tornadoes and often, similar patterns don’t even produce severe weather.
UAVs capable of handling tornado-like conditions could offer weather forecasters new insights into developing storms, tracking more variables at finer scales to provide valuable data about why tornadoes manifest in some systems and don’t in others. Improved models, however, can turn drone-collected information into more accurate tornado warnings.
Jacob and his team are still working out some kinks in their drone designs. These drones can’t usually fly into a storm at a moment’s notice, as they take up to four hours to prepare. Jacob’s team does occasionally encounter storms when in the field, and use these opportunities to test their UAVs. Other times, the team is met with inconveniences like crashing computers, miscalibrated sensors, or drone flying restrictions that limit their data collection.
Filed Under: M2M (machine to machine)