Inside Hurricanes: NASA and NOAA Open Science, AI, and Hurricane Hunters

NASA's Curious Universe follows hurricane hunter pilots from NOAA as they fly into storms to collect data satellites cannot. The episode highlights the use of P-3 Orion aircraft, dropsondes, radar, and real-time data sharing between NASA and NOAA to improve track and intensity forecasts, along with open science initiatives and the future role of AI in weather prediction.

Overview: The Curious Universe of Hurricane Forecasting

The episode centers on the collaboration between NASA and NOAA to improve hurricane forecasting through direct data collection from aircraft, advanced instrumentation, and open science. It contrasts satellite limitations with in-situ measurements taken inside hurricanes, and illustrates how data from multiple platforms is fused to improve predictions and inform disaster response.

Section 1: Meet the Hurricane Hunters

The program introduces Dean Legeagus, a hurricane hunter pilot for NOAA, who explains that once a storm nears the United States, the only way to obtain the needed data is to fly through it. Dean notes the two typical reactions from people: awe or disbelief at the danger. He describes the two NOAA aircraft at Lakeland, Florida,—the P-3 Orion reconnaissance planes Kermit and Miss Piggy—that serve as high-tech flying laboratories loaded with meteorological instruments. Rebecca Waddington, another hurricane hunter pilot, appears later in the episode, offering a complementary perspective from the Gulfstream 4 aircraft, Gonzo, which provides a broader environmental view rather than penetrating the storm itself.

Key moments include the phrase “you really can't fight it, you have to go with the ups and the downs,” capturing the improvisational challenge of storm navigation and data collection. The pilots describe the physics of the flight in turbulent eyewalls, where radar reflectivity can become dangerously intense and water in the eyewall can freeze and damage aircraft if encountered at altitude. The crew frequently mentions the difficulty and discipline required to keep wings level amid extreme turbulence, lightning, and downdrafts.

Section 2: Into the Eye and Through the Eyewall

The crew’s flight profile includes penetrating the eyewall, then moving into the calm eye where “hurricane hunting” data collection continues. They describe using dropsondes, which are weather instruments released from the aircraft to gather vertical profiles of the atmosphere as they fall toward the sea, providing crucial information on storm intensity and environmental conditions. They also discuss mesovortices inside the eyewall—mini tornadoes—that crews actively avoid during ascent or transit to maintain safety. As the aircraft pass through the eyewall and into the eye, the atmosphere’s changes are dramatic, including rapid vertical drops and the loss of outside visibility.

Section 3: The Teams and the Tools

Dean flies aboard Kermit with a crew of meteorologists and data scientists who calibrate sensors and collect measurements in real time. Rebecca Waddington flies aboard Gonzo, a Gulfstream 4 jet, which operates high above the storm and captures the surrounding environmental conditions that affect the storm’s track, such as moisture content and wind shear at various altitudes. The G4’s Doppler radar and its tail radar provide a full vertical picture of the storm, while the P-3 aircraft focus on the storm’s interior dynamics. The two aircraft complement each other, offering both micro (inside the storm) and macro (surrounding environment) insights.

Quotes from pilots emphasize the mission’s personal stakes, particularly for those living in storm-prone regions like Florida. The program underscores that the work is not only technically demanding but also emotionally consequential for communities in the hurricane path.

Section 4: Data, Forecasts, and Spaghetti Models

The podcast explains how the central point of a hurricane—the storm’s eye—serves as the initialization point for forecast models. Knowing the exact center of the storm is critical because a few miles’ difference can alter a five-day forecast. The crew describes how the aircraft repeatedly crosses the storm during a mission, dropping weather probes, and feeding data back to the National Hurricane Center. Satellite data alone cannot resolve all questions; in-situ measurements are essential for updating models and making more accurate track predictions.

Forecast models, often depicted as “spaghetti plots” showing many possible paths, improve with each flight as the ensemble converges on a more probable track. The pilots celebrate moments when new data reduce uncertainty in model forecasts, a tangible demonstration of the mission’s impact on public safety.

Section 5: Open Science, AI, and the Prithvi Model

NASA emphasizes its Open Science framework, releasing data, software, and publications without paywalls to enable broad participation in science. Kevin Murphy explains the four pillars: immediate open access to data, open software, no paywalls for publications, and public participation in scientific processes, including meetings. The show then dives into the Prithvi Weather and Climate AI model, a collaboration with IBM that uses a foundation-model approach applied to Earth data. The model is trained on large NASA datasets by masking data patches in a way analogous to masking words in language models. The intent is to enable pattern recognition, gap filling, and rapid testing on hurricane data benchmarks, with Hurricane tracks and intensities serving as strong validation data due to the extensive historical record.

Rahul Ramachandran explains how the AI model might supplement physics-based forecasts, enabling faster run times on consumer hardware and offering new insights by learning from data-driven patterns that physics alone might miss. The consensus is that AI should augment, not replace, traditional physics-based models, with potential future integrations where AI informs physics and vice versa.

Section 6: Disaster Response and Community Impact

Disasters response is a central theme, with the DRCS activating after events like Hurricane Beryl and Milton. Nighttime imagery helps map power outages and guide aid, while satellite and uncrewed aircraft provide situational awareness for responders. The program demonstrates how open data translates into practical relief efforts, including targeted aid to the most vulnerable populations by overlaying satellite-derived outage maps with census data.

The disasters program acknowledges rising risk from stronger and earlier hurricanes as global sea surface temperatures heat the oceans. Milton’s October 2024 landfall, damaging homes and triggering tornado warnings, underscores the critical need for rapid data-sharing and decision support to mitigate harm and save lives.

Section 7: Open Science as a Public Good

The episode concludes with Kevin Murphy detailing TOPS and the principles of open science, emphasizing that broad participation accelerates discovery and enables more trustworthy science communication. The conversations highlight the ongoing questions in hurricane formation and suppression, the role of Saharan dust layers, and climate change's impact on hurricane frequency and intensity. The overarching message is that NASA and its partners are building a global, open, data-rich ecosystem that empowers scientists, responders, and the public to understand and respond to extreme weather more effectively.