**
As the United States braces for an intense hurricane season and record-setting summer heat, concerns mount over the impact of the Trump administration’s cuts to crucial climate and weather data programmes. The National Oceanic and Atmospheric Administration (Noaa) has recently rolled out new artificial intelligence (AI)-driven weather forecasting models, but experts caution that without adequate historical data, these innovations may falter precisely when they are most needed.
AI and Weather Forecasting: A Double-Edged Sword
In late 2025, Noaa introduced a series of AI-powered global weather forecast models, touting enhancements in speed, efficiency, and accuracy. According to agency officials, these models are trained on centuries of weather history. Yet, as Monica Medina, former principal deputy undersecretary of commerce for oceans and atmosphere, points out, the effectiveness of AI in forecasting hinges on the robustness of the data it is fed.
“AI can help us analyse data at unprecedented speeds,” Medina stated. “But right now, we’re cutting back on data collection, which is counterproductive.” Under the Trump administration, investment in climate and weather data collection has dwindled, with a proposed overall budget cut of 40% to Noaa, despite a modest increase for the National Weather Service.
Staffing Cuts and Data Limitations
Critically, staffing reductions have led to the scaling back of essential tools such as satellite monitoring and weather balloon launches. Erica Grow Cei, a spokesperson for the National Weather Service, defended current data collection efforts, asserting that a variety of weather data is collected daily through multiple sources, including satellites, buoys, and land sensors. However, widespread reports suggest that these staffing cuts are hindering the operational capacity of Noaa’s data collection systems, leading to gaps in the information needed for accurate forecasting.
Craig McLean, Noaa’s former acting chief scientist, emphasised that reductions in climate research directly affect the quality of weather forecasts. “Cutting climate research impacts the skill of our weather forecast and halts our advancement in forecasting,” he explained. As the nation anticipates more extreme weather events, including the predicted “super El Niño” that could exacerbate heat and potentially increase hurricane activity, the stakes have never been higher.
The Shortcomings of AI Models
Historically, scientists relied on physics-based models to predict weather, utilising complex equations to simulate atmospheric dynamics. However, AI models represent a shift in methodology, identifying patterns in extensive historical data. While these models often require less computing power and can outperform traditional methods in certain areas, they have significant limitations—especially in predicting extreme weather events.
Recent research published in *Science Advances* indicates that AI models tend to underperform when forecasting unprecedented weather scenarios, often failing to account for the increasingly erratic climate conditions. Sebastian Engelke, a co-author of the study, highlighted that traditional models are less constrained by historical data, allowing them to better evaluate new weather phenomena.
In a sobering comparison, forensic meteorologist Chris Gloninger noted that AI models were trained on a climate that has fundamentally changed. “The infrastructure and models we have are built on the assumption of a stable climate, but extremes are on the rise,” he said. This disconnect has already impacted forecasting accuracy, as evidenced by the AI models’ struggles during a notable blizzard in February 2026.
The Political Landscape and Its Implications
Despite the potential pitfalls of relying heavily on AI in forecasting, Noaa has not fully transitioned away from traditional methods. Instead, the agency is integrating AI into its existing models, maintaining a blend of techniques to generate potential outcomes. However, the overarching concern remains: as the government reduces data collection and climate research funding, the integrity of these AI models is jeopardised.
Neil Jacobs, the current administrator of Noaa, is recognised for his expertise in modelling. However, as a Trump appointee, he faces political pressure to align with budget cuts that could undermine the agency’s effectiveness. Critics argue that while Jacobs is committed to improving weather forecasting, his allegiance to the Trump administration may compromise the quality of the data and models used.
Medina warns that the consequences of less accurate weather forecasts are far-reaching. “Weather forecasts are essential for our economy, public safety, and health,” she stated, stressing that unreliable predictions could endanger lives and disrupt vital sectors such as agriculture and energy.
Why it Matters
The ongoing cuts to climate and weather data collection not only threaten the accuracy of US weather forecasts but also place public safety and economic stability at risk. As extreme weather events become more frequent in the face of climate change, the need for reliable forecasting has never been more critical. Without a robust data infrastructure, the move towards AI in weather prediction may prove to be a perilous gamble, one that could leave communities vulnerable in their most desperate moments.
