Sunday, June 19, 2011

Commentary: the tornado warning process, use of technology, and application of social media during severe weather

I've posted in the past on the phenomena known in the weather broadcast community as "wall-to-wall" (or long-form) coverage, in which a local broadcast station (typically TV, though known to occur on radio as well) preempts all routine programming in order to keep the community informed and safe during a severe weather event.  It's not a foreign concept to most in the Mid-South, especially after the spring we have just gone through.

A couple of things prompt me to post again on the warning process and related matters: 1) a blog post by a well-known and respected broadcast meteorologist in Birmingham, James Spann, whom I greatly respect from afar, and 2) a Commercial Appeal article by James Dowd in this morning's (Sunday, June 19) Business section.  I highly encourage you to read each one.

Mr. Spann's main points (not mine) were these:
  • The high false alarm rate in tornado warning situations has created complacency and apathy, and therefore inaction, on the part of the public.
  • Too many people rely on sirens to warn them of a possible tornado and if they don't hear it, they aren't prepared.  Sirens are out-of-date technology that, while at one time served a purpose, no longer do.
  • NOAA Weather Radio will become obsolete unless it is upgraded to the polygon warning system currently in use by the NWS.  If it doesn't get upgraded, the private sector will lead the public into the future of receiving warnings with technology.
  • TV stations must stream their long-form (wall-to-wall) tornado coverage in a way that is accessible to all portable devices, not just some of them.
  • Social media is a lifeline during severe weather and must be embraced by TV meteorologists.
  • When there is a genuine tornado emergency, local broadcasters must go wall-to-wall to warn the public, and have the guts to take the hate mail.
  • TV meteorologists need to get off their off-camera computers, get in front of the chroma-key, and look their audience in the eye. It is a key element of communication.
The key points of Mr. Dowd's article, in which he appeared to get comments from meteorologists at all of the major players in the Memphis market were these (again, not my statements):
  • Memphis TV meteorologists have to maintain a balance between "too little" and "not enough" when it comes to live coverage of severe weather events.
  • Recent events in Joplin, Tuscaloosa, and elsewhere have prompted stations to "have their guard up" and viewer complaints are down, presumably because those events are fresh in their minds.
  • It is preferable to have too much coverage in an event where all turns out OK than to not be on the air when disaster strikes.
  • While re-running preempted programming at a later hour satisfies some viewers (depending on the program that was cut), advertisers are not as happy because they paid big bucks for certain time slots during certain programs.
  • With the expanding use of mobile technology, all agree that it is only a matter of time before these devices will be used for weather updates more than television.
I do not wish to comment again on the use, or lack thereof, of wall-to-wall coverage.  My opinion has changed little and can be found in the posting I mentioned at the top of this page.  However, I would comment on a couple of things that Spann and Dowd's articles touched on.

Tornado Warning False Alarm Rates
I don't wish to go too deeply into this subject or take a hard-line stance on this topic.  Many of my friends in the weather community are the men and women who are the unsung heroes at the National Weather Service that issue these warnings as a matter of "protection of life and property."

Spann differentiates between the "squall line spin-ups" type of tornado (which frequently are brief, weak, and occur on the leading edge of intense lines of thunderstorms) and the supercell twisters which typically cause the greatest damage, most injuries, and highest death tolls.  He takes on the squall-line cases directly, asking the NWS to stop issuing warnings for these brief spin-ups when, in most cases, they are near impossible to detect in advance and last only a few minutes.  By the time the warning is issued and received, the threat is about over and broadcasters are left on the air for up to an hour for a tornado threat that no longer exists, or at least is overshadowed by the greater threat - the damaging straight-line wind in the squall line.  He contends that these warnings desensitize the public.

I won't go so far as to say that the NWS should stop issuing warnings for these types of "lesser" threats.  However, Spann's point is well-taken.  The high false alarm rates (FAR) when it comes to tornado warnings have created a complacent public that is probably more dangerous than an underwarned public. Without having spoken to NWS officials directly on the matter, I am fairly confident in stating that they are likely doing whatever it takes to make sure that they don't miss a tornado, at the expense of 80% or more of their tornado warnings resulting in no tornado (even if the conditions are right and the precursors are present in the specific storm).  The damage control for a missed event has to be much greater than for an overwarned public.

So, while I completely understand their position, the capabilities of the radars these days are allowing the individuals triggering the warnings at the NWS to "see" with even greater detail the precursors.  This in turn has resulted in more warnings that don't verify (precursors don't necessarily mean actual tornadoes, or even funnel clouds).  This should be taken into account during the warning process, as should the societal impacts of an overwarned, "siren-weary" public.

Outdoor Warning Sirens
Spann suggested that tornado sirens were outdated and should be done away with. I partially agree, but take a different stance as to their future use.  Rather than doing away with them, these sirens need to be brought up to date, not only replacing many that are decades old with later models, but to capture the essence of the polygon warning process and only sound sirens that are within the actual area being warned.  ('s page on Outdoor Warning Sirens has more about these devices and allows you to plot their locations in Shelby County.)

The NWS stopped warning entire counties years ago and now issues tornado warnings for areas along the projected path of the storm - using what is called storm-based, or polygon, warnings.  The technology exists for sirens to be programmed such that, if they are not in the polygon area being warned, they do not sound.  I understand that the actual sound from these sirens may bleed over into areas that aren't warned, but I think that is a good thing.  (If you are 1 mile outside the polygon, are you truly "safe" from the storm?)  I also think that there should be a buffer around the polygon from a siren perspective, such that if the siren can be heard within the polygon but is not located inside the polygon, it should sound.  Again, technology will allow you to do this.

However, the time has come for people 20 miles from the warned area to STOP hearing sirens, just because they are in the same county.  The #2 reason for a desensitized public (behind wall-to-wall television coverage) is overuse of tornado sirens.  It's time the money be allocated to upgrade the sirens, add the technology, and activate them from a centralized location.  The money for this should be as sacred as the salaries of police and fire - it's become a matter of public safety.  (And while I am on this soapbox, if a tornado warning is cancelled by the NWS, shut off the sirens for goodness sake.) 

NOAA Weather Radio
Also mentioned by Spann, I agree 100% with his opinion that NOAA Weather Radio must come into the 21st Century.  The ability to only alarm for the county(ies) that are programmed is the first necessary step that was taken years ago.  The next step, along the same lines as the discussion above regarding sirens, is that technology must be used to allow these radios to be alarmed based on their location within a storm-based warning.  Why are weather radios going off in Collierville when the storm is in Millington?  The federal government and radio manufacturers need to invest in the technology as a matter of public safety!

Once again, private industry is leading the way with smartphone apps that will use polygon warnings and GPS to push notifications to users when they are in the path of the storm.  NOAA Weather Radios should have GPS embedded in the device and the warning information that is disseminated should contain the actual area being warned, not just the county.

Use of Social Media
Both articles commented on the rising use of social media in the field of weather, and particularly in the warning dissemination process.  Obviously, if you know anything about how MWN operates, you know that social media is vital to our modus operandi.

During severe weather, Facebook and Twitter are our means of nowcasting weather information to you.  In other words, it's our version of "long-form" coverage.  We were one of the first in the Memphis area to adopt this method two years ago and it has earned us high accolades and many thanks from our growing base of followers.

The great thing about nowcasting on a social media stream is that, if you as a consumer of information do not wish to know about it or it isn't affecting your area, you can turn it off - surf to another website, hide the posts, whatever.  We don't have to "preempt" anything else that you might wish to partake in.  You can come and go as you please.  TV broadcasters don't have this luxury (and I'm not delving into the digital x.1 and x.2 channels, which technically do give them a few options).

Not only does social media give you an option on viewing the information, but it really can be an effective means of spreading the word quickly and efficiently.  It is very easy to pass the information along and grow the number of consumers exponentially when needed.  On broadcast coverage, if you missed the weatherman's statement about the communities in the path of the storm, you may be out of luck.  Not so with social media - the re-tweet and share button allow the information to continue to be spread.  MWN sees social media as part of the future of information dissemination, which is why we adopted it early and use it often where weather is concerned.

All of the above taken together - a reduced false alarm rate for tornadoes, upgraded siren and radio technology, continued evolution of new techniques in mobile technology from the private sector, and the use of social media - should result in a less siren-weary, more storm-ready public, and eventually a further reduction in injury and death during violent storms.

What do you think?  Feel free to leave your comments below!  I read every one and am interested in your thoughts, particularly with regards to how MWN can play a role in your safety during severe weather.

Erik Proseus
Owner, Cirrus Weather Solutions

For weather information for Memphis and the Mid-South, where and when you need it, visit on the web, on your mobile phone, download our iPhone or Android apps, or visit us on Facebook or Twitter.


Jake said...

Very well-written post. I completely agree and have a good case study of your subject that gives credence to your suggestions: Starkville, MS on April 27.

For those that don't know, Starkville is a college town home to 20,000 permanent residents and around 20,000 college students. On April 27, Starkville was in the bulls eye of the high risk for severe weather. (Starkville is about eighty miles from Tuscaloosa.) As so many other areas in the region, Starkville was placed under a Tornado Warning a few minutes prior to a potentially tornadic cell (not just a brief spin-up, as you described above) would head into the town. However, due to earlier storms, Starkville (and the surrounding region) was without power. During this same time, the two local TV stations had ended wall-to-wall severe weather coverage and returned to normal programming, because they thought that the danger had momentarily subsided.

Based on my prior observations of the town's emergency plans, I immediately started to panic. I knew that, as an aspiring meteorologist, I would be informed because I had a battery-operated weather radio and smartphone to keep me informed. However, the vast majority of people in Starkville – especially the college students – get their weather information from TV/internet sources, warning sires, or from Maroon Alerts, the university's emergency text messaging service.

Though I knew the storm was soon entering Starkville and had warned all my friends, WFO JAN did not issue a Tornado Warning for our area until nine minutes prior to the storm's estimated arrival in Starkville. From previous experience, I knew that the warning sirens did not function properly during power outages. Also, TV/internet sources were not available. Finally, Maroon Alerts typically take ten minutes to arrive on cell phone after a warning's issuance because an emergency manager must manually issue the alert.

Everything was coming together for a catastrophe. Fortunately, the tornadic storm weakened prior to its arrival in Starkville, but had a tornado on the scale of Tuscaloosa or Joplin entered Starkville, there's no doubt in my mind that hundreds, or perhaps even thousands, would have been injured or killed.

So, I say all of this to say that I believe that this issue is incredibly complex, but many things can be changed to help the issue. I certainly agree with everything you said, and I suggest that we explore ways to reduce human intervention from the time a warning is issued to the time it reaches the public. As seen in Starkville, there are so many venues for vital communication to break down in times of crisis.

Anyway, sorry to be rambling, but excellent blog post!

Kevin said...

Very good points from both of you ! It is an extremely complicated problem!

Here are my rambling thoughts.

I would love to see targeted warnings and alerts via NWS radio and the outdoor siren notification net. That one simple thing, I feel, would reduce the "warning fatigue" of the general public.

Broadcasters are slow to change. With the new digital format they now have options for programming that did not exist a couple of years ago. I was glad to see WREG utilize this the other night on 3.2 for the severe weather in the area. A test I am sure, and it did not go unnoticed.

Social media is, as it should be, the next frontier for emergency message notification. I have been in situations when a location specific warning system would have been extremely helpful. A GPS enabled warning system receiver/app to me would be extremely useful.

Anyway, just some random thoughts.....


Meteorologist Erik Proseus said...

Jake - Thanks for your great example! I can see the "need for speed" in the system and you're right, it is very complex, but not something that can't be overcome with a little innovation (oh yeah, and some money).

Kevin - Always appreciate your insight as well. Like I said above, there are way technologically to overcome some of these hurdles. I hope to have MWN riding the leading edge of this wave!