For the past 36 years Frank Marks has accumulated thousands of hours probing the hearts and souls of hurricanes, getting to know them from their birth to their death. Now Hurricane Matthew, the first major cyclone to hit the U.S. mainland in 11 years—is giving him still more quality time with a monster. Marks—meteorologist and director of the Hurricane Research

Division at the National Oceanic and Atmospheric Administration’s Atlantic Oceanographic and Meteorological Laboratory—has been flying into the storm for the past several days and nights, along with some 18 other scientists, pilots, navigators and engineers. Every 12 hours they fly stretches of six to eight hours aboard one of NOAA’s P-3 Orion aircraft, recently refurbished with new wings and engines.

They use the four-engine turboprop airplane as a huge stethoscope to monitor Matthew’s vital signs. In these, as in larger NASA hurricane-hunter aircraft (such as a DC-8 used to study the genesis of the storms) each researcher sits before a console, a computer screen or a laptop mounted on metal racks where there used to be passenger seats. The screens show the flight profile as well as what the storm is doing in real time. Other monitors show rows of changing numbers reflecting the activity of the high-tech instruments connected to the outer fuselage of the aircraft: Doppler radars, microwave sensors, lidars, dropsondes, water vapor and aerosols spectrometers, and a suite of tubes to sample air, salt and dust particles, ice crystals and water drops.

Meanwhile the cockpit is minding its own business. Pilots with good stomachs and steel nerves like Commander Justin Kibbey have to thread their way in and out of the eye, constantly passing through the calm-and-turbulence patterns typical of the galaxy-shaped storms. By now Kibbey and his colleagues are unperturbed by the constant barrage of emergency lights and computer messages—any one of which might make any commercial flight divert to the nearest airport.

“We are out there for situational awareness: knowing what is going on now,” Marks tells Scientific American. “Our job is to collect as many observations as we can to send to the National Hurricane Center in Miami, and to NOAA’s numerical models center in Washington.” And those models must be continually fed with new data as the storm advances.  “Every six hours we put fresh data into each one of those runs, so that the models start the best way possible,” Marks explains.

NOAA's Orion P-3 in its Tampa MacDicc Air Force Base hangar. Credit: J. Cozic, NOAA

This is why, Marks adds, it is so crucial to actually go inside the hurricane. NOAA’s storm behavior models can use satellite data, but that has many limitations because clouds obscure the sea surface. The storm-chasing aircraft’s new scatterometer—a microwave radar sensor that measures the reflection (or scattering effect) produced while scanning the surface of the Earth—can see inside the storm with high resolution, something akin to a magnetic resonance imaging (MRI) scan. Also new this season is an airborne Doppler radar, which offers “three-dimensional snapshots of the winds to show hurricane specialists what the hurricane structure looks like with height, which tells us something about how it’s interacting with the vertical wind shear.”

This weekend’s missions are especially interesting for Marks because he will be deploying a new winged dropsonde for the first time in a strong storm. Called Coyote, it is a small motorized drone that is launched with its wings folded through an opening in the fuselage. It can fly for about one and a half hours, and according to Marks it is part of a new generation of instruments that are controlled by a pilot from a laptop on board the P-3. In 2014 it was deployed into Hurricane Edouard as a demonstration. “But this is the first time we deploy it in a major storm,” Marks says.

Hurricane Matthew is getting a lot of attention. At any given time, while Coyote explores its innards, a Global Hawk drone is flying 20,000 to 30,000 feet over its head. A NOAA Gulfstream IV plane is going around the storm in circles while the little P-3 is tangled in the fabric of its skirts; meanwhile, an Air Force C-130 is doing its own reconnaissance, and to complete the party from space, the instruments of the Aqua and Terra satellites are sending their own blasts of data.

Charles Lynch, an electronics engineer for the National Oceanic and Atmospheric Administration, releases a dropsonde from a hatch in a WP-3D Orion weather research aircraft into the Gulf of Mexico during an observation flight June 25, 2010. The dropsondes measure atmospheric pressure, temperature, humidity and wind speed and direction. Credit: Senior Airman Anna-Marie Wyant, U.S. Air Force

For Marks every storm is different, a little world unto itself.  “Hurricanes carry a lot of interesting stuff inside, from sand grains to dust particles, microvegetation, bacteria and pollution, but the most abundant are salt particles,” he says. “Salt is a very good source of condensation nuclei. It is better than dust and pollution. I remember when we were doing an experiment in the 90s, we were flying 300 feet above the surface collecting data and we had so much salt encrusted on us that the engine shut down.”

All that aside, the big headache and main challenge now for scientists is predicting the intensity of a storm. Recent advances have improved NOAA’s intensity forecasts by 20 percent, but there are so many variables influencing the developing of a hurricane—from the energy they draw from the oceans to their interactions with the surrounding environment and their dynamic inner cores—that storms like Matthew befuddle many experts.

“Rapid intensification of large hurricanes is something that the forecasters have a lot of trouble with,” says Adam Sobel, atmospheric scientist at Columbia University and author of Storm Surge, a book about Hurricane Sandy. “The models don’t predict it very well. That is just the state of current science. Matthew did intensify very rapidly. Before reaching Haiti it went from tropical storm to category 5 in just a little more than a day. It may be the biggest rapid intensification of an Atlantic storm on record.”

Something similar happened last year in the Pacific Ocean with Tropical Cyclone Patricia. To Sobel that is an indication that rapid intensification also plays a role in tropical cyclone climatology, as he mentions in a May 2015 paper in Nature. In other words, that Matthew’s last name might as well be Global Warming.