From Watts Up With That?

By Anthony Watts

If you listen carefully to the collective wailing coming out of Boulder, Colorado this week, you’d think the Trump administration had just announced plans to shut down the National Weather Service and replace Doppler radar with tea leaves.

In reality, what’s on the table is far more nuanced, and far less apocalyptic, than the headlines suggest.

According to USA TODAY, the administration is moving to dismantle or restructure the National Center for Atmospheric Research (NCAR), long regarded as a flagship institution in atmospheric and climate science . Predictably, this has been framed as an “attack on science,” “public safety at risk,” and — my personal favorite — a blow to America’s “competitive advantage.”

Cue the hysteria.

But let’s pause for a moment and separate weather science, useful computing infrastructure, and policy-driven climate alarmism, because conflating the three is exactly how we got into this mess in the first place.

NCAR was founded in 1960 to advance atmospheric chemistry, physical meteorology, and weather prediction. On that front, it has an impressive résumé. Hurricane dropsondes, numerical weather prediction advances, and severe weather research are real, tangible contributions that have saved lives and property. Even longtime critics of climate exaggeration — including Roger Pielke Jr. — acknowledge NCAR as a scientific “crown jewel” that deserves improvement, not destruction .

The problem is not that NCAR does weather.

The problem is what NCAR increasingly became.

Over the past two decades, NCAR — like much of federally funded climate science — drifted from observational rigor toward model-driven storytelling. Computer models, tuned to assumptions and parameterizations that consistently overstate warming and extreme weather trends, became the dominant product. Those outputs were then laundered through press offices and compliant media outlets as “settled science,” stripped of uncertainty and error bars.

That’s not atmospheric science. That’s advocacy.

When an institution begins hosting “justice-centered” programs, art exhibits about water relationships, and ideological framing exercises under the banner of Earth science, it invites scrutiny. The administration’s criticism of “woke” or policy-driven research may be blunt, but it is not entirely misplaced .

Climate Models are Useful Tools, but Terrible Oracles

Let’s be clear: climate models are not useless. They are tools — exploratory tools — not crystal balls.

Yet NCAR’s climate modeling efforts have been routinely used to justify sweeping policy claims: existential crises, “thresholds passed,” and forecasts of unprecedented catastrophe. Many of these claims fail basic validation against historical data. Satellite records, surface observations, and extreme weather statistics simply do not support the worst projections being sold to the public.

When models diverge from reality, the scientific response should be recalibration and humility — not louder press releases.

This is where skepticism becomes not only reasonable, but necessary.

That said, dismantling NCAR entirely — especially without a scalpel — risks throwing out the baby with the bathwater.

The USA TODAY article notes that NCAR operates a federally owned supercomputing center in Cheyenne, Wyoming. That facility is a national asset, regardless of one’s views on climate policy. High-performance computing is not inherently ideological. It can support:

• Short-term weather forecasting
• Severe storm modeling
• Aviation and military applications
• Hydrology and flood prediction
• Energy load forecasting
• Even non-climate scientific research entirely unrelated to CO₂

Shutting it down simply because some of its cycles were used to run overheated climate scenarios would be an act of bureaucratic vandalism.

A far more rational approach would be repurposing and refocusing.

Strip out the advocacy. Re-center on observational science. Decouple supercomputing from policy-driven climate modeling. Move critical infrastructure where it can serve broader national interests — or at least put it under management that understands the difference between uncertainty and certainty theater.

The media narrative wants a villain-versus-hero storyline: Trump versus science. Reality is messier — and more interesting.

There is legitimate criticism to be made of how climate science has been politicized, how uncertainty has been buried, and how institutions like NCAR have increasingly blurred the line between research and activism. Ignoring that problem would be irresponsible.

But responding with a blunt instrument risks collateral damage to areas of atmospheric science that actually work, and matter.

Reform NCAR? Absolutely.
Audit its programs? Long overdue.
Throttle back the model-driven hysteria? Please do.

But before anyone pulls the plug entirely, it might be wise to remember that weather is real, models are fallible, and supercomputers don’t vote.

If the goal is better science — not louder narratives — then precision, not demolition, should be the order of the day.