A Serious Note: The WeatherTiger Newsletter for March 2025

Taking stock of what firing NOAA employees and dismantling NWS infrastructure means for the upcoming hurricane season and beyond.

On November 8, 2024, from Brownsville to the Florida Keys the people of the Gulf Coast went about their daily lives as usual. Schools and businesses were open, after-work social sports teams practiced, complimentary peanuts were consumed at the Texas Roadhouse Grill.

All this is remarkable only because about 350 miles from Tampa and New Orleans, Category 3 Hurricane Rafael lurked in the central Gulf of Mexico. Thanks to an accurate forecast from the National Hurricane Center that Rafael would harmlessly dissipate over water, no U.S. coastal watches or warnings were issued. No one on the Gulf Coast was asked to evacuate, or even to prepare for a hurricane. Life continued on, as unperturbed as if Rafael had never existed.

The story of modern meteorology is the story of this kind of quiet miracle. Most days, the public doesn’t think too hard about the weather other than a quick glance at an app or the TV news, and certainly doesn’t ponder how that forecast arrives, fully formed, on your device of choice. While no predictive science will ever be free from uncertainty, generally accurate and reliable forecasts are part of the unremarked background of existence, like the air itself.

Of course, the public availability of free, accurate weather forecasts is not effortless. Wrangling domesticated predictability from the dynamic chaos of the ocean and atmosphere is the end result of generations of effort from forecasters and researchers, plus the continual maintenance and incremental improvement of complex observational and computational systems.

Perhaps nowhere has this progress been more dramatic than in my own chosen sub-domain of expertise, hurricane forecasting. A couple of weeks ago, the National Hurricane Center released verification statistics for their forecasts issued during the 2024 hurricane season, which tallied $200 billion in U.S. damages.

Continuing a decades-long trend, the NHC beat their own accuracy records last year. In 2024, a forecast 5 days out was as accurate as one with just 2 days of lead time in 2000. That means fewer false alarms, narrower evacuation orders, reduced storm anxiety, and billions of dollars in efficiencies for the economy as a whole. For those savings, you can thank the dedicated forecasters who work at the NHC, decades of NOAA investments in observation and modeling technology, and researchers like my friend Dr. Andy Hazelton, who evaluates the next-generation HAFS hurricane model that enabled those skillful forecasts in 2024.

Andy was fired last week, swept up despite exemplary performance reviews in a purge of approximately 700 “probationary” (read: at their current job for less than a year) NOAA and NWS employees.

I met Andy at Florida State, where we shared a lab and research advisor. He’s one of those guys for whom weather is not just a paycheck, but a lifelong passion and calling. Unfortunately, these firings had the consequence of not only terminating Andy’s work improving hurricane models, but also excluding buoy data from ocean analyses, cancelling weather balloon launches, and straining understaffed NWS local weather forecast offices to the breaking point.

Nor is that likely to be the end of it. There are indications of further disruptions to come, including the possible liquidation of NOAA-affiliated contractors, dismissal of thousands more NWS forecasters, and termination of leases on operations centers housing critical radar, satellite, historical data, and forecast infrastructure. If these changes go forward, you’ll notice. Given the robust and growing slate of private weather companies in the United States, it’s understandable to assume that your app, the Weather Channel, or people like me can step in to seamlessly fill any gaps in demand that hobbling the NWS might create.

We can't.

Despite accounting for one-thousandth of the federal budget, NOAA and the NWS produce routine forecasts, created by trained human experts, for the entire country, plus lifesaving severe weather warnings. NOAA also operates the computer modeling, radar, satellite, windsonde, station, and buoy networks that inform those forecasts.

By law, all of this modeling and data must be provided for free to the public. That gives everyone from massive companies like IBM to small entrepreneurs like me the freedom to use that trove of information to develop new applications (like WeatherTiger’s real-time seasonal hurricane forecast and agricultural yield models), supporting the specific needs of our clients and deriving yet more value from the data.

If NOAA operations are crippled, almost all the modeling and data that underpins private weather forecast products goes away as well, leaving us blind. Your phone’s weather app is very likely an AI repackaging of forecast output from the NWS’ suite of weather models, and your radar app draws directly from NOAA’s national network of 160+ radars. While other weather models, like the European, do exist, those models are only as good as the data fed into them. Slashing NOAA’s observational, radar, and satellite networks will greatly degrade the skill of these forecasts as well. Almost everything that I do at WeatherTiger hinges on NOAA data, which is true of weather-dependent enterprises as a whole. The capacity to replicate NOAA/NWS public safety functions or data sources simply does not exist in the private sector.

For what it’s worth, I will continue to write forecasts and discussions this season with whatever information is at hand. Hopefully there will not be interruptions to the reliable availability of U.S. weather data, but I am actively planning for contingencies to mitigate that risk.

This is a studiously apolitical newsletter by design, as weather is one of life’s great equalizers. The rain falls on the just and the unjust alike. And, indeed, over three-quarters of Americans from across the political spectrum support the critical, lifesaving work of the National Weather Service. By 60%+ margins, the clear will of the American people of all stripes is that this valuable work be allowed to continue. Historically, support for the NWS is not controversial: the Weather Research and Forecasting Innovation Act of 2017 that increased NOAA’s funding and responsibilities passed both houses of Congress by acclaim and was signed by none other than President Donald J. Trump.

No institution is perfect. Science is the search for continual improvement, and that should extend to responsibly adapting the structure of our scientific and public service institutions to changing needs. Firing thousands of employees serving in essential public safety roles is not efficient, nor is it an orderly restructuring. It is an indiscriminate, emotional bloodletting that will harm our economy and harm people, the equivalent of trying to lose weight by cutting off your arm. We should not let the search for perfection be the enemy of the very-good-to-excellent institutions of public meteorology we have built in this country, which cost each citizen $4 per year and return hundreds of billions in value in forms as diverse as smoother flights, cheaper food, and faster shipping.

The people like Andy and the infrastructure that make life less nasty, brutish, and short may be out of mind, but are no less real for their invisibility. We live in an era of strange compulsions to touch hot stoves, but I hope the massive unpopularity and economic illogic of destroying NOAA and the NWS can spare these irreplaceable institutions from the most chaotic outcomes.

If you disagree with dismantling the National Weather Service and replacing it with nothing as tornado, flood, fire, and hurricane seasons approach, contact your elected representatives and let them know how you feel.

---

March ENSO forecast

WeatherTiger’s in-house El Nino/La Nina modeling, along with pretty much all other guidance, has been consistent in predicting the demise of the ongoing La Nina in spring 2025, perhaps as soon as… now. Our latest model run suggests that neutral conditions are likely to return no later than April, with cool-neutral conditions persisting through summer. While spring is a time of deep uncertainty in El Nino/La Nina forecasting, the probabilities of El Nino development, which would act as a brake on the 2025 hurricane season, remain minimal.

Beneath the surface, the Equatorial Pacific is mostly cooler than average down to a few hundred meters’ depth, despite some notable surface warmth showing up in the last 2-4 weeks close to South America. With stronger than average trade winds expected over the next few weeks, this patch should cool off a bit, and there is no indication of a rapid transition to El Nino afoot.

---

Florida spring temperature/precip forecast

While La Nina is on its final legs, its influence will continue in Florida for a few more months. Through the end of May, that means there is about a 55% chance of drier than normal conditions, with 25% and 20% chances of near- or above-normal precipitation, respectively, for the state as a whole. North Florida has a better chance of landing closer to normal for spring rainfall. Drier springs tend to be warmer than average in Florida, so look for a 45% chance of average temperatures in March-May running a degree or more above normal, with a 20% or so chance of cooler than normal temps.

Observed rainfall in Florida between 12/1/24-2/28/25 as a % of normal.

This is a fairly high confidence forecast, though there is more uncertainty at in late spring as it’s not clear whether a La Nina overhang to global weather patterns will dissipate heading into early summer. For what it’s worth, WeatherTiger’s expectations for winter were generally met: we called for a most likely outcome of about 60% of normal rainfall in Florida, and most of the state did indeed see 50-80% of normal precip. (Though of course, well above “normal” in sleet and snow.) Winter temps surprised a bit on the cooler side at mostly 0-1F above normal, averaging out a normal December, very cold January, and very mild February. I had put 30% odds on average temps within a degree of normal, and a most likely anomaly of +1.6F above average.

---

Next update

With the hurricane season hype cycle already picking up speed, WeatherTiger’s initial outlook for the season will be released no later than March 28th, and our early and late summer full outlooks will be released around May 23 and July 25, respectively.

And finally, because I think we could all use it, a mood music moment of zen. See you in a few weeks as we start to grapple with the enigma of Hurricane Season 2025.