The Crisis in the National Weather Service

The National Weather Service is falling apart. 

 

That is the only reasonable conclusion to be drawn by informed observation of its cumulative output during the past, highly active, 30 days of severe thunderstorms and tornadoes. 

 

Before proceeding, I want to stipulate that the Trump Administration’s handling of the cost cutting and downsizing of NOAA earlier this year was unhelpful. Yes, there was – and still is – fat in the NWS and NOAA. Cutting the deficit should be a high priority but those cuts need to be well thought out and well executed. 

 

I also want to state there are exhausted people in the NWS who are suffering from vast amounts of overwork due to manpower shortages that existed before President Trump took office -- that were made worse by President Trump’s layoffs and then rehiring, in some cases,  people with little to no experience. The quality of work of burned out meteorologists, reasonably, is not optimal. 

 

The above stipulated, it is a fact that the decline in the NWS began during the Obama Administration and has continued unabated. 

After record, and rising, accuracy scores for tornado warnings and watches from 2005 to 2010, it all seemed to come crashing down with the horrific – and poorly warned – 2011 Joplin Tornado. The NWS published an highly flawed after-storm report (the tornado killed 161) that was reasonably assumed by most readers to be factual. Relying on that report lead to a number of unfortunate decisions. Combined with that misinformation, the Service also seemed to lose confidence. Instead of staying with what had been highly successful, the National Weather Service decided to make its warning system far more complex. They also created levels of products beyond the state-of-the science to do consistently well. When they, predictably, failed – confidence seemed to drop farther. 

 

Underlying the above is the issue of training. When the National Weather Service conducted its last reorganization in the early to middle 1990’s, it opened a radar and storm warning school in Norman, Oklahoma. For a month, NWS meteorologists received state-of-the-art training on radar interpretation and storm warnings. Then because senior management (presumably) viewed their workforce as “trained,” the NWS shut down the school. Since then, the level of training of new hires varies – from adequate to nearly nonexistent -- depending on where they are stationed and other factors. 

 

The above has been reflected in the increasingly poor quality of tornado warnings and, perhaps, tornado watches. I have published the numbers before (brown link, above), so I won’t recount them here. But let me give you just a few of many examples from the past month.

 

Tornado Warnings

This is a bad tornado warning. Issued Monday, there was no actual rotation. What appears to be rotating precipitation (image on right) is in an area where there is no radar echo (no raindrops, etc.) at all (left). This is due to an artifact called a "side lobe" -- a type of false radar echo. In addition, there was no reflectivity signature of a tornado (e.g., hook echo) and the thunderstorm was well into the cool air north of a quasi-front. This tornado warning should not have been issued. 

A tornado warning never should be issued in this case. This one would have been bad enough but the St. Louis office of the National Weather Service issued 27 consecutive tornado warnings Monday without any tornadoes occurring! Conventioneers in St. Louis were forced to interrupt their meetings – twice – for false alarms (the complaints can be viewed on LinkedIn) for the City of St. Louis. Businesses let people leave early. Rush hour traffic was far below the norm. Cumulatively, this huge number of false alarms caused serious economic losses and loss of credibility for the warning system – meaning people might not take shelter when the “real thing” occurs in the future. 

 

Tornado Watches

The category of tornado watch that denotes the most serious threat is called a “particularly dangerous situation” (PDS) watch. They are rare. One was issued Monday afternoon for the Mississippi Valley for 8 1/4 hours. Spotters were deployed, television stations brought in extra meteorologists, little league baseball games postponed. There were no tornadoes until 29 minutes before the watch expired at 11pm. It was a weak tornado near Jonesboro -- instead of the EF-5's and EF-4's one typically associates with PDS tornado watches. 

 

The night before, a tornado watch was issued in southern Kansas and western Oklahoma until 4am, with a "moderate" chance of "strong" tornadoes. The watch forced people in certain jobs to stay up all night monitoring the weather along with others who have storm anxiety. Were there any tornadoes? No. Not even a weak one. 

 

Weather Observations

 

This is Sunday's (April 26) 30,000' (300 mb for meteorologists) map of weather
balloon observations. There are huge holes in the West -- from where the large-scale low pressure system which in turn caused the tornado, giant hail and flash flood-producing thunderstorms was originating. 

Observations: weather balloons, radar, satellite data, surface data (temperature, wind, etc.) and all other types of atmospheric observations are the most basic information needed to make weather forecasts and storm warnings. Without this data, no one can make accurate forecasts. The NWS is failing in its obligation to launch weather balloons at the internationally agreed-upon time of 1200 hours UTC (7am CDT). This jeopardizes forecasts and other activities each day by every meteorological organization; federal, military, academic and private sector. 

 

Amidst all of this, the management of the NWS and NOAA seems lost. Instead of focusing on these, and other, critical issues by (for example) reopening the storm warning school, NWS management had been pursuing the latest “shiny thing” – starting a training program to teach NWS meteorologists – when some offices are already badly understaffed – to offer services to third parties at locations away from the offices where they are employed. The need for accurate forecasts and warnings of the (taxpaying) public almost seems to be an afterthought.  

 

The NWS says it is “reorganizing.” But, as of today’s Grok AI analysis of the reorganization, none of it involves training, bringing back retired meteorologists as mentors, hiring part-time college meteorology students to launch balloons or answer phones, or similar ways to meet its obligations and bring its current workforce up to speed. 

 

Since 2012, I have been working tirelessly in attempt to persuade Congress to create a National Disaster Review Board (NDRB, modeled after the hugely successful NTSB) to assist the NWS, NOAA, FEMA, and other players in improving their products and services. That is an important suggestion and the House is considering two bills to create a Board. An effective NDRB is a necessity. 

That said, I have been making daily weather forecasts and storm warnings since 1971. I have never seen the NWS perform at this level. It puts lives in jeopardy. It causes serious economic losses. 

 

I call on Congress to open an investigation and hold hearings on these issues as soon as possible. To the extent possible in today’s Washington, I also call on them to be non-partisan -- as quality weather information from the federal government is a necessary public service. A weather balloon is neither Republican nor Democrat. 

 

In my book, Warnings: The True Story of How Science Tamed the Weather (2010), I explain how the National Weather Service was the very best value among federal agencies. We must, quickly, restore that level of quality and trust.