Avoiding the Hype Trap
In an interview in Parade magazine, National Weather Service (NWS) Director Dr. Jack Hayes was asked "Is there too much hype in weather forecasting?" His response:
I think meteorologists are sensitive to overhyping the weather and creating a lack of trust with the public. Whenever you're in doubt, check with the National Weather Service. When we issue a weather warning, it means there's a real threat, and it's time to take cover.
The problem here is that weather threats don't always materialize. And when they don't, the perception can be that they were overhyped despite an honest intent not to hype but simply to communicate risk.
Consider last Thursday afternoon, when a winter storm warning was issued for the metro area for the threat of a "dangerous" combination of snow and ice. By most accounts, dangerous conditions never developed and readers here started to complain about the hype.
From a purely meteorological perspective, a case could be made that the NWS should never have issued a warning for that storm and should have simply gone with an advisory (issued for the prospect of less serious conditions than a warning). However, it understandably took a more precautionary approach, especially in the wake of the election night ice storm that paralyzed the area.
Here at the Capital Weather Gang, we were not overwhelmed with the storm's potential to be highly disruptive and actually rated the storm just 2.5 flakes, not a major event. But we felt it would be irresponsible not to report the NWS warning and did not feel it was appropriate to downplay the storm with the election night event fresh in our minds. Forecasters in the broadcast media had a similar mentality. As CWG visitor Howard B and meteorologist from WUSA-9 aptly commented here:
I'd rather be a little more ominous about ice in the forecast and err on the side of caution. At times, we have to forecast based on psychology as well as meteorology.
Attempting to effectively communicate risk responsibly will be a continuing challenge for weather forecasters. We're sensitive about being accused of overhyping storms. We welcome your ideas for presenting information about impending threats effectively -- hype-free.
By Jason Samenow |
February 26, 2008; 11:00 AM ET
Government
,
Recaps
,
Winter Storms
Previous: Forecast: Mild and Wet Then Cold and Dry |
Next: CommuteCast: Rain Showers Moving Through
Posted by: Jake in Reston | February 26, 2008 11:21 AM
It seems like the boy who cried wolf kind of problem. When an event doesn't materialize people are less likely to trust future warnings. Its a balancing act. Frankly, it seems like the nature of the beast and meteorologist have a responsibility to inform people that each forecast is subject to change. The CapWx team does an great job or being honest with their confidence in forecasts and balancing risk vs. hype. My only suggestion would be to have some survey of how accurate different weather forecasting outlets preform.
Posted by: Steve in Silver Spring | February 26, 2008 11:59 AM
I agree about Hype..It can break the trust barrier if things are hyped up early on then the track or forecast changes. Good article.
I am trying to not get over excited about this 3/4-3/5 event. I am noticing some very unusual patterns that havent taken place all winter with this 3/4 system. Why is there such a southen bend in the 850 mb line as the storm approaches..Its alomst like
its system is bending the line around. Can someone explain this? Its almost like the cold pocket follows the storm. This entire winter, storms have pushed out the cold air.
Here is a link I am looking at for the 12z:
180 Frame shows clold air being drawn in from the south: (WE are still warm)
http://raleighwx.easternuswx.com/models/gfs/12zgfs850mbTSLPp06180.gif
192 Frame we are cold, at a point in time before is when the 850 mb line passes through and the changeover from rain to snow I would expect to occur:
http://raleighwx.easternuswx.com/models/gfs/12zgfs850mbTSLPp12192.gif
Cap Team, I know you cannot post early hype on your daily articles with this 3/4 event. Can you answer my question about why cold air is wrapping around from the south as storm approaches. Thanks
Posted by: StormChaser | February 26, 2008 12:14 PM
Thank you for posting this. I hate it when people say that the meteorologists are overhyping everything and call them "liars." I've gotten into what amount to screaming matches with people over this subject, LOL!
Meteorologists are just the messengers. They have to tell you the what COULD happen and that you need to take action based on what COULD happen, so if the worst happens you'll be prepared.
It's the same logic on eating right and exercising and taking cholesterol meds. You want to PREVENT a heart attack by taking action on it. The doctors are giving you a warning that if you don't eat right and exercise, you COULD have a heart attack, so you want to prepare for the worst.
Posted by: weatherdudeVA (Lake Ridge) | February 26, 2008 12:24 PM
"At times, we have to forecast based on psychology as well as meteorology."
Less paternalism and less projection of one/s own fears is what/s needed.
Make the forecast and leave it go at that.
Posted by: TQ | February 26, 2008 12:31 PM
Good article and I completely understand and agree with the position you guys take in the forecast!
Posted by: HEELS | February 26, 2008 12:32 PM
Augusta Jim: You have been quiet lately. Any thoughts on 3/4.
Thanks
Posted by: VA | February 26, 2008 12:51 PM
Storm Chaser-- The cold air being drawn south is simply a result of the very strong counterclockwise circulation around the deep center of low pressure. If a deep low is moving from south to north, cold air can make the type of push you see in that model. It's the beginning of a cold occlusion which you can read about here.
Posted by: Capital Weather Gang | February 26, 2008 1:50 PM
In addition (according to the GFS forecast), it looks like the cold pool is being enhanced / advected further south with the development of a strong upstream (Rocky) anticyclone, which sets up a very strong northerly flow through the Great Plains and Central US. In addition, the Rockies themselves can enhance further by acting as a western barrier for the colder, more dense air.
It looks like the initial plunge of cold air is being initiated the co-location of the Rocky Mountain anticyclone and an Upper Great Lakes/Southern Canadian clipper system.
Posted by: D | February 26, 2008 2:05 PM
Couldn't the dangerous conditions have existed last week if none of the roads had been treated and every normal Friday morning car had been on the road? I did see ice coating everything in the morning, but the roads (in N. VA) were ok.
Also, pointed out before here, there are the complications of forecasting for a region that can have such diverse conditions (rain, snow, and ice across the same weather map). People tend to make overall weather judgements based on their own personal experience.
Posted by: DM | February 26, 2008 2:26 PM
CWG:
With all the talk of computer models that pops up from time to time on these comments, it would be helpful if you could provide a primer on the GFS, UKMET and other models, explaining what sort of data goes into them, how the models are used and who is responsible for actually developing the models. The NWS seems to rely on the GFS a lot as its "go to" model. One question I have is when was the last time the GFS model was tweaked, or is it it constantly worked on? My apologies if this info is already available on the blog site, but I couldn't find it in the FAQ section.
Posted by: jmbethesda | February 26, 2008 2:58 PM
Just a quick comment here...I hope this does not sound too much like self promotion, but I have just completed a book, that will be out by the end of March, called "How to Find a Good Weather Forecast"...it covers all of the stuff you have been discussing here: How forecasts are made...Who's really doing them...and of course,how you as a weather forecast consumer can find the best products.
My question to everyone out in Capital Weatherland is this: Would you attend a weather forecasting seminar? If we held a confrence that was open to the public on how weather forecasts are made by the pros...would you go? Just curious...keep up the great work guys!
Tony Pann
WUSA Weather
Posted by: Tony Pann | February 26, 2008 3:46 PM
Tony,
Re: the forecasting seminar.
That's something I'd certainly be interested in. Hope you can make it happen.
For those asking about 3/4, the current NWS forecast for that date is a chance of showers and a high of 51. Though not an impossibility, things would certainly have to change a great deal for snow.
Posted by: Ivan | February 26, 2008 3:57 PM
jmbethesda,
The GFS is in general the go to model for the NWS, since it has been developed, supported, and is run by employees of NWS. Additionally, it is in general available to forecasters much sooner than some of the global forecasts produced by other operational centers.
The model is constantly being worked on, tested, improved, etc. However, these changes are implemented into the operational suite on average about once a year. The last big change to the GFS occurred on 01 May 2007, including a new vertical coordinate, as well as a major upgrade to the Data Assimilation (i.e., the code responsible for creating the initial condition for each new numerical forecast).
There is another implementation scheduled for this spring, but we are waiting on central operations for the exact date.
There is a nice chart somewhere that presents information about how the operational global models differ. I'll see if I can find and post a link.
Posted by: D | February 26, 2008 4:22 PM
Tony,
I'd be interested in the seminar. I've been a fan since your Dayton days. I used to intern over at WKEF back in my college days.
Jamie
Posted by: Anonymous | February 26, 2008 5:03 PM
Tony-- Look forward to seeing your book. Keep us posted on the seminar.
Posted by: Jason, Capital Weather Gang | February 26, 2008 9:49 PM
The comments to this entry are closed.











Is there a weather-equivalent of an ombudsman? Should there be?