How to prevent the next Atlanta ‘Snowpocalypse’

The events in Atlanta this week are an eye opening example of what can go horribly wrong when relatively good weather forecast information is misused or ignored.

It's also a clear example of why every major city in America should be paying for private weather consultants.

With some distance now between the human weather catastrophe in Atlanta this week, it's becoming clear to this seasoned weather forecaster that the events in Atlanta were completely avoidable.

The weather forecasts put out by the Atlanta National Weather Service and others in the days leading up to the storm weren't perfect, but they were clearly good enough for Atlanta and the state of Georgia to make timely and effective operational decisions that would have largely prevented the catastrophe that unfolded.

Grow the Future of Public Media

MPR News is Member supported public media. Show your support today, donate, and ensure access to local news and in-depth conversations for everyone.

Winter Weather Alabama
Joyce Cook, and instructional aide, reads Snowflake Bentley to a group of 5th graders stranded at Oak Mountain Intermediate school due to snow and ice on Tuesday, Jan. 28, 2014, in Indian Springs, Ala. AP Photo/Butch Dill

So what do we do to prevent the next Atlanta?

I don't pretend to have all the answers. But on this one I may be able to offer some unique and unbiased perspective that could be of value to those government officials willing to listen. As chief meteorologist for MPR News, I'm certainly far removed from the emotion and heat of the politics of the situation unfolding in Atlanta.

I also used to work for a private weather consulting firm in Chicago called Weather Command. John Murray and Denny Trettel pioneered private weather consulting  in the late 1940s. They also helped this rookie weather forecaster cut his teeth in the difficult and precise weather business of 'operational forecasting' in the 1980s.

Weather Command still provides highly localized, precise snowfall forecasts and warnings for major cities like Chicago and Milwaukee. They are among the best in the business in my experience. Without the self-promotional hype and attempted media headline grabs of many other private weather providers, they provide incredibly solid forecasts.

Their clients save thousands, even millions over time through increased efficiency and staffing achieved through accurate weather forecasts specialized to their operations. The return on weather investment is many, many times the cost.

I also work in Minnesota where we're pretty good at dealing with snow and keeping traffic moving in conditions far worse than what they experienced in Atlanta this week. Roads and freeways were flowing just fine this afternoon in the Twin Cities following 6.3 inches of snow this morning. No snowpocalyptic traffic jams here.

The Minnesota Department of Transportation is like a finely tuned military operation when it comes to pre-treating and removing snow from highways in Minnesota.


With that backdrop, here is my view of what happened in Atlanta, and how it's an instructive lesson for cities all across America on how to prevent the next avoidable instant weather disaster.

657 20140128_atlabandon1
On foot near abandoned cars in Atlanta Tuesday. Image: Atlanta NWS

What just happened? Some facts

There are lots of accusations and hyperbole flying around these days in Atlanta. Here are some basic facts.

  • 2.6 inches of snow fell in Atlanta Tuesday via Atlanta NWS

  • 11:53 am to 7:53 pm - snowfall was reported at Atlanta Fulton County Airport via NWS hourly observations

657 atl obs
Image: Atlanta NWS

Here's a condensed version of warning timeline from the Atlanta NWS.

NWS Product Timeline

453 AM Mon, Jan 27, 2014

* Winter Storm Watch expanded to include the entire metro Atlanta area (link)

308 PM Mon, Jan 27, 2014

* Tweet issued "Winter precip will make travel risky across GA midday Tues into Weds. Not a bad idea to stay off the roads if you're able!"

Atlanta NWS

936 PM Mon, Jan 27, 2014

* Winter Storm Watch for north Atlanta metro area upgraded to a Winter Weather Advisory. The first paragraph included this text "Please understand that even a slight shift in the moisture could result in significant differences in snow amounts and may require an upgrade to a warning." (link)

* Winter Storm Warning remains in effect for south Atlanta metro area and central GA (link)

339 AM Tue, Jan 28, 2014

* Winter Storm Warning issued for the entire Atlanta Metro area (link)

* Winter Storm Warning remains in effect for central Georgia (link)

Bottom line? A winter storm watch was issued for metro Atlanta more than 24 hours before snowfall began.

winter storm warning was issued about eight hours before snowfall began in Atlanta. There was plenty of lead time to indicate a winter storm was possible with the issuance of the watch. While not ideal, eight hours was enough lead time to be useful for local and state officials to cancel school that day, get all available resources on the roads to treat and plow major freeways, and to alert local media and the public about the impending winter weather event.

Changing forecasts: A survey of forecasts and perspectives

Weather systems are like people. They evolve and change, sometimes quickly. Many National Oceanic and Atmospheric Administration and other global forecast models run every six hours with new, updated data. The track, temperature profile and intensity of storms cranked out by the suite of forecast models can change rapidly with each new computer run. "One to 3 inches"of snow can quickly become "3 to 6 inches- plus." 

Sound familiar? That's why we update weather forecasts so often.

To be fair, some of the model output and forecasts in the days leading up to the Atlanta snow event strongly suggested the bulk of the storm would pass south of Atlanta.

One forecaster in the snow zone who graciously admits he missed the mark this week is James Spann of ABC 33/40 in Birmingham, Ala.


His frank admission is of a forecast gone wrong and his apology to his viewers and readers is impressive, and a good lesson for all of us. It takes a pretty big person to do that, something I will try and remember next time I blow a forecast. And I know it will happen to me again.

James makes a great point that resonates strongly with me. Some in our profession present weather forecasts with a certainty that's just not supported by the science of meteorology.

Humility in what we do as weather forecasters is critical. What we do does indeed have a high success rate, and saves businesses and the US economy billions each year. But we are still going to have our bad days, and busted forecasts.

Here's a clip from James' humbling post on the storm.

In terms of human impact, yesterday’s forecast “bust” was the most significant for me since Janaury 1982, when we had a timing error of about six hours on the arrival of freezing rain and snow. Instead of starting at 5:00 p.m… it started around 11:00 a.m. People were caught off guard, schools and businesses closed, and the result was traffic gridlock, abandoned cars, separated families, and human suffering. Very much like yesterday, so I have been down this road before. The main difference is that I was a young man of 26 years old in 1982. With the experience and maturity that should come with a long tenure in my position, you would think that kind of error would not happen again, But, it did.

PLACEMENT PROBLEM: As is sometimes the case, the forecast was actually good on the synoptic scale, but the mesoscale placement was bad. By about 100-125 miles. The band of heavier snow that was to have set up over South Alabama (they did get good snow later in the day) initiated over the I-20/59 corridor. Birmingham’s official snow total was 2.0 inches; Anniston’s official total was 1.5″, and Tuscaloosa had 0.3″ (at the Airport… some there had more).

I have said this to both professional meteorology societies in speeches over the last two years. Humility is missing in our science. There are many things we don’t know, and many things we can’t do. Just about the time you think you are infallible, you will be brought to your knees. For the ones in meteorology and climate that say “I could be wrong”, I will listen and respect their opinion. But, for those that claim no error, we all know their time is coming.

So, an apology from me for a botched forecast. Won’t be the last bad forecast I write, but I will keep working to get better and stronger daily. And, no, for this kind of thing doesn’t “get me down”, it energizes me. Thanks for your support and for those that did write an encouraging note over the last 12 hours. Let’s warm up, get back with our families, and enjoy some low 60s this weekend. Scroll down for the morning forecast discussion.

Bravo James! You have my complete respect as a fellow broadcast meteorologist and you are a role model to all in our profession.

Failure to communicate?

One clear lesson from the Atlantapocalypse is that essentially accurate forecast updates were wither not communicated forcefully enough, or were not listened to by decision makers. It's clear that in Atlanta the forecast was good enough to make better decisions, and officials failed to act on good weather information.

I have talked with and listed to presentations by The Weather Channel's Bryan Norcross about his passion for the critical role of effective communication in weather forecasting.

657 norcross

Here's some great perspective from Brian's post on what happened in Atlanta.

Somewhere and somehow somebody has got to take the lead on closing the threat-understanding gap between forecasters, decision-makers, and the public. It’s not simple because of the division of responsibilities between various federal, state, and local agencies in a disaster. But, we’ve seen too many instances where good-enough weather forecasts have lead to bad decisions and poor public communications. The issue is partly science, which we should be able to solve with an organized effort by the National Weather Service, FEMA, and others.

But there’s another big problem, which the Georgia governor articulated very well in his new conference. He was more afraid to be wrong in closing down the city, than he was of people being stranded in their cars. Until we can develop a system that keeps politics out of it and lets science and good judgment drive the decision-making bus, this kind of thing is going to keep happening.

There is more great perspective from Jason Samenow at Capital Weather Gang, from NPR affiliate WBUR, The San Francisco Chronicle and NBC that's worth reading.


Why Atlanta and every other major US city should use private weather forecasters

I know many cities in the U.S. do employ private forecasters for operational decisions like snow and ice removal. For those who don't, Atlanta is a great example of why they should.

In my days working for Weather Command in Chicago, we issued snow and ice warnings to dozens of cities. Cities used those timely warnings to schedule road crews and make timely, critical operational decisions about how to treat and plow roads to keep them as clear as possible.

It's relatively easy to attach a dollar value to these critical services. There are tremendous savings in overtime and scheduling for city crews, efficient use of road salt, and scheduling maintenance of equipment.

The multiplier effect on keeping a city moving, and the economic benefit to business is harder to calculate but of far higher value. How much does it cost to shut down the city of Atlanta for two days?  How much more to pull thousands of abandoned cars off the roadways?

657 20140128_spaghettijunction
Image: Atlanta NWS

A credible experienced operational weather forecast service costs a fraction of what it saves each year for major cities. It's an investment that the taxpayers reap back multiple times. So there's an economic incentive in hiring private weather firms to be in constant contact with operational staff in major, and even mid sized and smaller cities.

There's another benefit. There's plenty of  "free" weather information out there that's either inaccurate, or too many voices with differing forecasts. How is a city road crew supervisor who has little or no meteorological training supposed make sense of six different forecasts and make good operational decisions? A forecast from a single provider gives you one contact point, one forecast.

But this may the biggest reason to use a quality private weather service: Accountability.

Most of the time the forecast will be a good one. But when it goes horribly wrong, there's a direct line of accountability to the forecast provider. If they're wrong often enough, they get fired. If they're good enough, future events like what we saw in Atlanta this week will be fewer. If the forecast is good, and officials fail to act, then responsibility is clear.

Using credible, qualified, experienced private weather forecasters may not be the perfect solution to preventing the next 'Atlanta' but it is one of the best I have seen in my experience.