NOAA Budget Cuts Threaten Hurricane Forecasting This Storm Season

As the Atlantic hurricane season approaches (June 1–November 30), forecasters are riding high on unprecedented accuracy: the National Hurricane Center’s (NHC) 2024 track forecasts were the most precise on record, with one-day outlooks rivaling the three-day forecasts of just a decade ago. Yet this success story is under threat. Recent staffing reductions and funding proposals at the National Oceanic and Atmospheric Administration (NOAA) jeopardize the advanced instrumentation, data streams and computing resources that underpin modern hurricane forecasting.

Record-Breaking Forecast Accuracy and Satellite Innovations

Thanks to investments in satellite sensors, data assimilation and numerical models, official NHC track errors have shrunk by 75% since 1990. GOES-16 (launched 2016) and GOES-18 (2022) provide visible, infrared and near-infrared imagery at 0.5 km spatial resolution every 60 seconds, while the Suomi NPP and JPSS-1 polar-orbiters deliver microwave soundings critical for probing storm cores through thick cloud decks.

“We’re now resolving convection-scale features that were invisible two decades ago,” says Dr. Michael Brennan, senior hurricane specialist at NOAA’s Hurricane Research Division. “That extra detail feeds into the HWRF and FV3 ensemble systems—running at ≤3 km grid spacing on NOAA’s 15-petaflop HPC cluster—to yield more reliable 5-day forecasts.”

Critical Components Under Threat

• Radiosonde Wind Profiling: Twice-daily launches from ~900 global stations supply temperature, humidity, pressure and wind profiles up to 120,000 ft. Recent suspensions at over a dozen sites (including key Rocky Mountain and Great Plains locations) have created data voids that degrade 4D-Var and ensemble Kalman filter analyses.
• Hurricane Hunter Reconnaissance: Two WP-3D Orion turboprops and an Air Force WC-130J fly directly into storms, deploying 20–30 dropsondes per mission. Each dropsonde records 1 Hz measurements of pressure, temperature, RH and 3D wind down to the ocean surface. The February 2025 layoff of two of eight flight directors threatens mission cadence during multi-storm outbreaks.
• Satellite Observations: NOAA’s earth-observing fleet—including GOES-R series, Jason-3, DSCOVR and the COSMIC-2 GPS radio–occultation constellation—provides the continuous, high-resolution data that calibrates and validates in situ measurements and drives AI-based intensity forecasts.

Advances in AI and Machine Learning for Intensity Prediction

Beyond traditional physics-based models, NOAA researchers and academia are leveraging deep learning to forecast rapid intensification events—when wind speeds jump ≥35 mph in 24 hours. The Global Rapid Intensification Program (GRIP II) uses convolutional neural networks trained on multi-satellite imagery, dropsonde profiles and sea-surface temperature fields. Early trials reduced intensity errors by 10–15% versus HWRF alone.

“AI can detect subtle vortex tilts and convective bursts that human analysts might miss,” explains Dr. Sara Businger of the Cooperative Institute for Meteorological Satellite Studies (CIMSS). However, these algorithms rely on uninterrupted data feeds and high-performance GPUs—both at risk if NOAA’s AI R&D budgets are slashed in the FY2026 passback.

High-Performance Computing and Model Enhancements

State-of-the-art forecasting hinges on rapid turnaround of ensemble simulations. NOAA’s Environmental Modeling Center currently runs 64-member global and hurricane ensembles across IBM Power9 systems with NVIDIA A100 GPUs. Planned transition to an exascale system in 2027 will enable sub-1 km nested grids and real-time Monte Carlo uncertainty quantification.

Budget documents circulated in May 2025 indicate zero funding for NOAA’s Aircraft Recapitalization Plan and cutbacks for the Research and Development High Performance Computing Systems (RDHPCS) program—risking delays to the exascale upgrade that is critical for next-generation model physics.

International Collaboration and Data Exchange

NOAA shares data globally via the World Meteorological Organization’s Global Telecommunication System (GTS) and bilateral agreements with EUMETSAT, JMA and CMA. New real-time direct broadcast from Himawari-8 and Fengyun-4 satellites, along with European Copernicus Sentinel-6 altimetry, have enriched ocean heat content analyses that forecast storm fuel. Continued US leadership in data sharing is jeopardized if NOAA can’t sustain ground station operations and international liaisons.

Budget Outlook and Policy Implications

The Biden administration’s FY2026 budget request sought a 5% increase for NOAA, including funds for six next-generation reconnaissance aircraft and expansion of the Cooperative Institutes network. But the Office of Management and Budget passback proposes flat funding, cuts to Cooperative Institutes, and elimination of satellite lightning mapper investments—essential for convective trend detection.

In early June 2025, the Senate Appropriations Committee advanced a bill that restores $50 million for radiosonde operations and a one-year extension of WP-3D maintenance contracts. The House version, however, remains stalled under a 2% across-the-board cut mandate. “A single under-forecast storm can cost billions in damage and lives,” warns former NWS Director Louis Uccellini in a recent open letter to Congress.

Conclusion: Sustaining Our Forecasting Edge

Hurricane forecasting is a synergy of hardware, software and human expertise. From the high-altitude radiosonde that profiles the jet stream to the dropsonde that measures the eyewall’s eyewall’s eyewall winds, and the petaflop-scale clusters that churn out millions of model ensembles each day—our ability to predict and prepare for storms hinges on sustained investment. As coastal populations swell and climate change intensifies tropical cyclones, maintaining and upgrading NOAA’s data, aircraft and computing infrastructure is not discretionary—it’s a national imperative.