NOAA’s Atlantic Oceanographic and Meteorological Laboratory (AOML) says a small uncrewed aircraft will add a new layer of observations to hurricane forecasting during the 2026 season. The announcement is noteworthy because it reports a measured model benefit, not simply a technology demonstration: retrospective experiments found that including the aircraft’s data reduced maximum-wind forecast errors by about 10% overall.
AOML published the finding on May 19, 2026, as the laboratory prepared to integrate the data into NOAA’s Hurricane Analysis and Forecast System, or HAFS. The result comes from a joint effort involving NOAA scientists, the University of Miami’s Cooperative Institute for Marine and Atmospheric Studies and other research partners. It places near-surface measurements at the centre of a practical question for forecasters: how can models better represent the part of a hurricane where ocean heat and atmospheric motion interact?
What NOAA is adding to HAFS
The aircraft is Black Swift Technologies’ S0, a small uncrewed aircraft system that is released from one of NOAA’s P-3 Hurricane Hunter aircraft. At 2.6 pounds, the S0 is light enough to be carried into a storm and deployed from the crewed research aircraft. Once released, it measures atmospheric pressure, temperature, humidity and wind, along with ocean waves and sea-surface temperature.
Those observations complement the data already collected by NOAA’s Hurricane Hunters. HAFS is NOAA’s next-generation operational hurricane model, designed to provide guidance on both a storm’s track and its intensity. Track describes where a tropical cyclone is expected to move; intensity describes how strong it may become. The new observations are aimed mainly at the second problem, where small changes in the exchange of heat, moisture and momentum can influence rapid strengthening.
Why the lowest part of the storm matters
Crewed aircraft can sample a hurricane’s structure, but there are limits to how close people can safely fly to the ocean surface. AOML describes the marine boundary layer as a difficult-to-access environment where important processes controlling formation and intensification take place. An uncrewed aircraft can enter that layer without putting another person in the measurement platform, extending the information available to the model.
The S0 is not a replacement for NOAA’s P-3 missions. It is launched by those missions and adds another set of measurements to the wider observing system. AOML says that, after testing the platform, evaluating data quality and establishing transmission procedures, forecasters can receive the observations in real time. That connection is important: the value of a measurement depends not only on collecting it, but also on getting reliable data into the forecast process quickly enough to matter.
What the reported experiments found
To estimate the effect on HAFS, scientists ran retrospective experiments with and without S0 data for 10 tropical cyclones from 2022 through 2025. In these comparisons, the data improved intensity forecasts, with an overall improvement of about 10% and an improvement of up to 25% for tropical storms, according to AOML. The laboratory’s accompanying graphic compares maximum sustained 10-metre wind-speed errors in model runs that assimilated the aircraft data with runs that did not.
These figures should be read as an evaluation of model performance across the reported cases, not as a guarantee for every future storm. A retrospective experiment can show how the additional observations changed a forecast system under controlled comparisons; it cannot promise that every storm will be observed in the same way or that every forecast will improve by the same amount. The announcement also does not say that S0 data alone determines an official warning. Forecasters combine many observations, models and expert assessments.
What changes during the 2026 season
AOML said the S0 data would be integrated into NOAA’s hurricane forecast model for the first time during the 2026 hurricane season. The next step is therefore a transition from research evidence to live operational use. That transition will show how the aircraft performs across real missions, how often it can be deployed, and how consistently its measurements arrive in time for forecast cycles.
For the public, the significance is straightforward but limited: better intensity guidance could give emergency managers and communities more useful information about how strongly a storm may develop. It does not make hurricanes predictable in every detail, and it does not remove the uncertainty that accompanies fast-changing storms. Anyone following an active system should continue to use the latest advisories from the National Hurricane Center, rather than treating an individual research result as a standalone forecast.
NOAA’s announcement shows how hurricane forecasting is being improved through a chain of evidence: a platform is tested, its measurements are checked, the data are assimilated into a model, and the result is compared against forecasts made without it. The S0’s first operational season will provide the next test of whether the roughly 10% improvement seen in retrospective experiments can translate into dependable guidance when storms are developing over the ocean.
