Outdoor warning sirens

After adding my last post to Google Plus, a friend asked my thoughts on tornado sirens. I replied that I thought they provided a rather poor return on investment. This eventually lead to a day-long discussion with a coworker who disagrees with my assessment. Since I’ve never put my opinion on sirens in a blog post, I figure it’s time to do just that.

First, I am of the opinion that sirens serve an important role in public safety. In places like parks, golf courses, and common outdoor gathering areas, sirens are an excellent way of communicating a single message: “seek shelter”. Sirens are unable to communicate why shelter should be sought, when it is safe to come out (though some jurisdictions will re-sound sirens as an all clear), or what the threat is. Some siren systems lack battery backup, occasionally rendering them inoperable at the most inopportune times. Sirens also suffer, as do most alerting systems, of being stuck in a county-based warning system that no longer exists.

Although they are not intended to be indoor warning devices, some people still rely on them as their primary means of receiving severe weather alerts. When close enough to a siren, this can be fairly reliable, but it’s not always the case. My house is 1000 feet from the nearest siren and during tests it is quite audible (uncomfortably so when the windows are open). During heavy rain, it is not nearly as noticeable. I have little doubt that I would sleep through an overnight siren sounding if it happened during a heavy rain.

So what, then? Since the 1970’s, NOAA Weather Radio has served the country well. It provides both alerts and routine weather information around the clock. Unfortunately, it’s also stuck in the 20th century. County-based alerting simply cannot continue to dominate our warning dissemination systems. If the weather radio system were to send warning coordinates along with SAME codes, receivers could optionally determine if the alert needs to be sounded. In addition, most counties are served by a single transmitter. Each transmitter should have a redundant backup, located far enough away to be unlikely to fail from the same event (e.g. power outage, tornado), but still able to cover the assigned counties.

The current state of sirens in Tippecanoe County includes coverage of all areas I propose require it (and probably some that don’t). There are still, by my calculations (see note 1), approximately 25,000 people in the county who live outside the audible range of sirens. In order to cover the entire county’s land area, the initial investment would be $960,000 to $2.7 million (see note 2) with an annual maintenance cost of $115,200-324,000. (see note 3)

I wanted to look at the costs for all thirteen counties served by the WXK74 transmitter in Monticello, but it turns out the siren count information is not easy to find. WTHR in Indianapolis did some of the work for me, but the rest had to be independently researched. Sadly, some Emergency Managers don’t want to disclose even a count of the sirens in their county. As a result, I was only able to obtain authoritative siren counts for Cass, Clinton, Howard, and Tippecanoe Counties.

Using the same strategy as for Tippecanoe County, I calculated what it would take to get these four counties to 100% siren coverage. The recurring costs are $774,00 to $1.44 million. This after an initial investment of $6.42-12 million dollars. And remember, that’s just for four of the fourteen counties, only 30% of the land area covered by WXK74 (roughly 65% of the population).

What would happen if instead of sirens, we added a second transmitter site and bought every household in the covered area a $50 weather radio? The cost of the transmitter would be about $75,000 (see note 4). Buying the radios would bring the cost up to roughly $9.84 million, which is in the range of covering 65% of residents with sirens.

Of course, this is entirely academic. Sirens are funded at the county level, whereas weather radio is a federal project. It’s not easy to just move the money around. There’s also the alert-granularity issue that needs to be resolved.

Given the economics and the richness of information, it makes sense to push for more radios instead of more sirens. Sirens have their place, but those places are limited. The Indiana Department of Homeland Security guidelines would suggest that many local cities and towns should not have sirens at all (except for parks). Ultimately, weather alerting requires a defense-in-depth approach. Sirens are one layer in certain situations. Weather radios are another, more broadly applicable layer.

A third layer is the Wireless Emergency Alert (WEA) system that is being deployed. Unfortunately, it requires a relatively modern smart phone, so I expect the penetration rate is still fairly low. It, too, suffers from a lack of geographic granularity (although probably better than either weather radio or sirens), and sparseness of information (WEA messages are limited to 90 characters). There’s also people who don’t have their phone by their side at all times. Some people actually leave their mobile devices in other rooms sometimes? Quelle horreur! Disturbingly, WEA does not interrupt phone calls, meaning a long gab session will result in you receiving a warning after it has expired.

Given budgets and politics at all levels of government these days, I’m sad to say that I don’t see any of the existing deficiencies being resolved any time soon.

Continue reading

More thoughts on warning polygons

On Tuesday, Patrick Marsh wanted a distraction from his dissertation and embarked on an idle investigation of tornado warnings and impacted areas (my thoughts on what “impact” means are below). Using some very rough approximations, he calculated the percentage of warned persons who are impacted by a tornado. Even under the most generous set of assumptions, the verification by population is generally below 20%. It’s worth noting that 2011 (the most recent year that official tornado data is available) was the best year in the analysis, but there is no indication of a general improvement trend.

Despite some of the problems I’ve previously noted in the polygon warning system, it’s still better than warning entire counties. Still there’s a lot of room to improve the false alarm rate. Much of the population-based false alarm comes from warnings that have no tornado at all. The rest either comes from too-large warnings or not-small-enough warnings (“not-small-enough” warnings are small enough to be justifiable, but still larger than absolutely necessary).

It’s not always easy to shrink warnings. Only the supercell storms relatively close to a radar site seem suitable. In those cases, it’s possible to make the warning only a few miles wide, or the width of the mesocyclone with uncertainty added as you go downstream. This would minimize the area under the warning, but it got me wondering: would that be too small?

At the scale of a mile or two, how do you explain the warned area to the public? Storm-based warnings are already difficult to communicate quickly, and microwarnings would only compound the problem. Even in Lafayette, the 10th largest city in Indiana, the covered area might look something like:

...TEAL ROAD BETWEEN 4TH STREET AND 26TH ST...
...KOSSUTH ST BETWEEN 9TH STREET AND SAGAMORE PARKWAY...
...SOUTH ST BETWEEN FIVE POINTS AND FARABEE DR...
...18TH ST BETWEEN BECK LN AND FERRY ST...

And so on. Or maybe it would use neighborhoods and landmarks instead:

...LAFAYETTE COUNTRY CLUB...
...HIGHLAND PARK...
...JEFFERSON HIGH SCHOOL...
...FIVE POINTS...
...WALLACE TRIANGLE...
...COLUMBIAN PARK...

Either way, it’s much more complicated than a simple “LAFAYETTE”. Yes, it’s more detailed, does that help? First, it takes much longer to read the text. Secondly, can you count on people, especially those who are new to the area, to know the streets, neighborhoods, and landmarks well enough to quickly figure out if they’re affected or not? I suspect the answer is “no”.  Perhaps some day someone with the time, energy, and funding can look at this.

Sidebar: What does it mean to be “affected” by a tornado?

When Patrick commented on Twitter about his post from Tuesday, I remarked that the results depend on how “affected” is defined. His analysis was based on population, but that doesn’t necessarily convey all impacts. If my office is wiped out by a tornado but my house is untouched, I am still affected. You can expand this out even further and incorporate businesses that saw decreased revenue as a result of a tornado, even if they were not directly hit. Businesses that see increased revenue (e.g. home improvement stores) might also be included, even though the effect is a positive one. The broader (and, I would argue, more accurately) we define being affected, the more difficult it becomes to get accurate data.

Disseminating storm-based warnings

Earlier this month, I wrote about the format and wording of severe weather warnings, and how to effectively shape those warnings. In those previous posts, we came to the conclusion that county sections are pretty lousy ways to define warnings. Defining warnings based on counties leads to over-warning, and using county subsections are ambiguous to the public. Storm-based polygon warnings are the most accurate way to define warnings, but they come with their own problems. First, as I previously discussed, there’s the issue of having to shape around county boundaries. Secondly, they present some challenges in dissemination.

Storm-based warnings are easily transmitted visually (though they still require a basic level of geographical knowledge that I’m not sure we can assume), so they work well on TV, Internet images, and smartphone apps. They are really poor in text or audio formats. Warnings disseminated through audio or text must still reference vaguely-defined county regions. As a result, storm-based warnings lose some of their benefit immediately.

NOAA All-Hazards Radio’s Specific Area Message Encoding (SAME), which is used to selectively activate weather radio receivers, works on a county basis. As a result, the NWS obsoleted its own technology when it switched to storm-based warnings in 2007. On the other hand, county-based warning distribution like SAME has one distinct advantage: the ability to pre-warn. A common recommendation when programming weather radios is to have the radio activate for warnings for your own county and also the surrounding counties. This allows additional lead time in some events. The current warning paradigm does not allow for such a setup.

The future of warning dissemination will be hitting people in the pocket. Mobile phones are the best way to reach a large (and growing) portion of the population. The Wireless Emergency Alert system is a good start. WEA will automatically send warnings to cell phones in affected areas (this also helps to address the issue of people driving, especially through areas they wouldn’t normally be), but it has some room for improvement. The character limit of WEA messages is 90, which is just over half the length of a traditional text message, resulting in a very information-sparse alert. In addition, it will be based on the tower‘s location, not the phone’s. This means that people will receive warnings that do not include them (or worse, will not receive warnings that do include them). Of course, it also requires that people have a WEA-capable device.

In the end, a multi-layered approach is required. Broadcast media must continue to remain a valuable partner. WEA and third-party smartphone apps should continue to get warnings to people’s phones. Weather radio technology should either be upgraded to support location-based alerting or be gracefully retired. Warning siren systems should be upgraded so that they can be sounded selectively (most systems still sound county-wide), or better yet scrapped entirely.

Effectively shaping warnings

Earlier this week, I wrote about the format and wording of severe weather warnings in order to most effectively communicate the necessary information to the public. In that post (and the excellent discussion that took place in the comments), I referred several times to the problems that can arise from the shape of warnings. Before I let loose on that, let’s set some historical context. From 1953-2005, the National Weather Service issued warnings on a by-county basis. This lead to over-warning unaffected areas.

In 2005, the NWS began a pilot program to issue warnings with a forecaster-defined polygon shape. In this way, the warnings could be issued such that they reflect actual threats and not political boundaries. All forecast offices began using these “storm-based warnings” in 2007, but the system still isn’t perfect.

Radar image showing two separate polygon warnings side-by-side

Sample polygon warning from WFO Indianapolis. Source: http://www.crh.noaa.gov/ind/?n=polywarn

Even though forecasters can issue warnings that are based on the atmosphere, they still must consider the political boundaries. Many warning dissemination systems (more on this in a future post) don’t support polygon warnings, so if a warning would normally clip a corner of a county, the forecaster must consider whether or not to cut a notch out of the warning. (I asked for clarification from a friend who is an NWS forecaster. He said there’s a setting that optionally excludes tiny slivers of counties. “We try to do our best to serve two masters between county based communication systems and scientifically based warnings. The main focus at all times, however, is getting information to people who are threatened as fast as possible and in as useful a manner as possible.”)

The problem is further compounded when a storm exists along the boundary between the County Warning Areas (CWAs) of two forecast offices. Current NWS practice does not allow a warning to extend outside an office’s CWA. Since CWAs edges are determined by the county borders, they are frequently uneven. A storm may clip a small portion of another office’s CWA, and the issuing forecaster must shape the polygon to avoid that portion. In order to include said portion, the other office must issue its own warning. While offices will often coordinate when storms are near a CWA boundary, I seriously doubt that any forecasters will take the time to make sure their polygons match precisely. The resulting discontinuity can be confusing to the public and makes absolutely no sense from a threat perspective. The storm does not respect political boundaries.

Radar image with severe thunderstorm warning polygons. The shape of the middle polygon is influenced by the boundary between county warning areas

A recent severe weather event in central Indiana. The warning in the middle of the image was issued by the NWS office in Northern Indiana. The southern extent of the warning is defined by the boundary between the Northern Indiana and Indianapolis county warning areas.

Removing the county-border issues from the polygon system still doesn’t lead to perfection. Polygons themselves suffer from some issues. Notably, they’re ripe for being over-large in order to improve verification scores, as Patrick Marsh posted earlier this week. The other key concern is that, as I mentioned above, some warning systems have no concept of polygons. NOAA All-Hazards Radio (also known as “weather radio”) is a prime example., as are outdoor warning sirens (some locations have the ability to sound sirens selectively, but it is by no means ubiquitous). Until these systems are modernized, even the best polygons will still lead to over-warning.

Effectively communicating to the public

One of the main challenges a meteorologist faces (and this is true for many professions. I eagerly await a post from Matt Simmons drawing a parallel in systems administration) is effectively getting a message to the public. This becomes especially important in times of severe weather when the timeliness and clarity of the message can literally be the difference between life and death. While the National Weather Service and the media do a good job, there’s a lot of room for improvement. Let’s consider the three questions a person needs answered:

  1. Am I threatened?
  2. If so, what do I do about it?
  3. When am I in the clear?

Now compare this to an actual warning.

027
WUUS53 KIND 011733
SVRIND
INC015-023-157-011830-
/O.NEW.KIND.SV.W.0131.120701T1733Z-120701T1830Z/

BULLETIN – IMMEDIATE BROADCAST REQUESTED
SEVERE THUNDERSTORM WARNING
NATIONAL WEATHER SERVICE INDIANAPOLIS IN
133 PM EDT SUN JUL 1 2012

This is just header information. The public rarely encounters it directly.

THE NATIONAL WEATHER SERVICE IN INDIANAPOLIS HAS ISSUED A

* SEVERE THUNDERSTORM WARNING FOR…
CARROLL COUNTY IN NORTH CENTRAL INDIANA…
NORTHWESTERN CLINTON COUNTY IN CENTRAL INDIANA…
NORTHEASTERN TIPPECANOE COUNTY IN WEST CENTRAL INDIANA…

Okay, so let’s assume I know what county I’m in. Am I in the right part of the county?

* UNTIL 230 PM EDT

The first question to be definitively answered is the third one.

* AT 131 PM EDT…NATIONAL WEATHER SERVICE DOPPLER RADAR INDICATED A
SEVERE THUNDERSTORM CAPABLE OF PRODUCING QUARTER SIZE HAIL…AND
DAMAGING WINDS IN EXCESS OF 60 MPH. THIS STORM WAS LOCATED 6 MILES
SOUTHEAST OF BROOKSTON…OR 8 MILES NORTHEAST OF LAFAYETTE…AND
MOVING EAST AT 40 MPH.

There’s some town names here and a description of the threat. Maybe I can figure out if I’m threatened by this or not.

* LOCATIONS IMPACTED INCLUDE…
DELPHI…
YEOMAN…
ROSSVILLE…
CAMDEN…
FLORA…
BURLINGTON…

THIS INCLUDES INTERSTATE 65 BETWEEN MILE MARKERS 173 AND 181.

Oh, okay! My city isn’t listed, so I’m probably in the clear. Folks in Delphi know that they are threatened. The first question is answered.

PRECAUTIONARY/PREPAREDNESS ACTIONS…

SEVERE THUNDERSTORMS PRODUCE DAMAGING WIND IN EXCESS OF 60 MILES PER
HOUR…DESTRUCTIVE HAIL…DEADLY LIGHTNING…AND VERY HEAVY RAIN. FOR
YOUR PROTECTION MOVE TO AN INTERIOR ROOM ON THE LOWEST FLOOR OF YOUR
HOME OR BUSINESS. HEAVY RAINS FLOOD ROADS QUICKLY SO DO NOT DRIVE
INTO AREAS WHERE WATER COVERS THE ROAD.

TORRENTIAL RAINFALL IS ALSO OCCURRING WITH THIS STORM…AND MAY LEAD
TO FLASH FLOODING. DO NOT DRIVE YOUR VEHICLE THROUGH FLOODED
ROADWAYS.

And here, basically at the end of the warning, is the answer to question 2.

LAT…LON 4073 8652 4070 8652 4069 8637 4063 8636
4034 8641 4045 8693 4057 8689 4057 8678
4066 8678 4067 8676 4073 8676 4074 8675
4074 8663
TIME…MOT…LOC 1733Z 280DEG 19KT 4052 8679
WIND…HAIL 60MPH 1.00IN

The latitude/longitude pairs define the shape of the warning (more on that in a bit). This can be used to provide really good answers to question 1.

Your typical severe weather warning answers all three of the questions we base our discussion on, but in the wrong order. Arguably, question 1 gets answered first, although it relies greatly on the geographical awareness of the public. I suspect many adults are aware of what county they live in (due to school districts, libraries, property taxes, etc), but if they work in a different county, do they know that county’s name? Can they name the surrounding counties (this is useful for preparedness. If the county to your west is under a warning, chances are good you might be soon)? The fact that subsections of counties (e.g. “northeastern Tippecanoe County”) are variable and undefined only add to the confusion.

The fact that the “call to action statement” (what NWS meteorologists call the answer to question 3) comes at the tail end of the warning means that a full minute may pass until it’s made clear what actions someone should take. In many cases, the lead time is sufficient that this additional time is acceptable, but in short-lead-time situations (for example, the Joplin, MO tornado of 2011) every second counts.

It’s not bad to give people more information than they need, especially when it reinforces your point. One example comes from a derecho that hit Louisville, KY on July 13, 2004. This storm caused widespread damage through southern Indiana and central Kentucky. In order to make it clear that this was an exceptionally dangerous storm, one of the forecasters at the NWS office in Louisville made mention of “hurricane force winds” in the warnings. Although this wasn’t strictly necessary, it helped to make the danger more clear in the mind of the public. Extra information can be valuable, but it should never get in the way of the main point.

So how would I re-format warnings?

<headers>

THE NATIONAL WEATHER SERVICE HAS ISSUED A <threat> WARNING FOR

<counties>, including <cities>

<call to action, don’t die, etc.>

<THIS WARNING EXPIRES AT <expiration time>

<More explanation, supporting info>

The media has an advantage here, as they can format their presentation of the warning however they’d like. The NWS is stuck with a defined format. It’s worth noting that decades ago, the public did not receive NWS products directly; they were filtered through the broadcast media first. In modern times, All-Hazards Radio, weather websites, and mobile apps are putting more NWS products directly in front of the public. So far, the NWS has not updated warning text to fit this new model.

One thing you may have noticed is that my version doesn’t mention the office issuing the warning. I had this discussion with a recent meteorology graduate last week. What use is knowing what office issued the warning? Most people probably don’t know what NWS office serves them. Furthermore, a warning may be issued by a backup office if the primary office is unavailable (e.g. if the office staff is taking shelter because they’re about to get hit by a tornado). Adding the issuing office does nothing to answer or reinforce the three questions. It’s not extra information, it’s extraneous.

Note: I was going to add some comments about the shape of warnings, but this post is long enough and that rant won’t be short. Look for it in the next few days.

Increased complacency about severe weather benefits no one

The number of meteorologists in the United States is very small.  According to the Bureau of Labor Statistics, less than 10,000 people are employed as atmospheric scientists in non-faculty positions (anecdotal evidence suggests that the number of people holding meteorology degrees is significantly higher. To wit: of the 12 people in my graduating class, four are meteorologists).  With such a tiny fraction of the population trained in atmospheric science generally, and severe storm meteorology specifically, it should come as no surprise that the public knows relatively little about severe weather.  With the small number of meteorologists, a heavy reliance is placed upon the media and local officials to convey information.

However, while the media and local officials may get more exposure to weather information, they do not necessarily understand it any better than the rest of the general public.  This leads to newspapers reporting that a “local tornado warning was issued” (only the National Weather Service issues tornado warnings officially, and causing confusion about this does not help the public interest) after a “funnel cloud on the ground” was sighted (a “funnel cloud on the ground” is more properly known as a “tornado”, but in this case it was more likely a mere “scary-looking cloud”). It leads to emergency managers sounding warning sirens when the greatest threat is heavy rain and sub-severe winds.  And it leads to confusion and eventual complacency for the public.

Meteorologists have enough trouble fighting complacency as it is.  The most recent data from the National Weather Service indicates that 76% of tornado warnings are false alarms.  This is not because of incompetent meteorologists.  It is a limitation of available observation systems (radar), of the understanding of tornadogenesis, and of the (quite reasonable) belief that it’s better to overwarn than to miss a tornado.  Additionally, since tornadoes are often relatively small and short-lived events, it may be that some of these false alarms are not so, but there are no reports thus the warning remains unverified.  The upshot of all of this is that it’s very easy for the public to not take warnings seriously.

I can, perhaps, understand the reason the Tippecanoe County Emergency Management Agency (TEMA) decided to sound the sirens last Saturday.  A street festival was about to begin in downtown Lafayette, and many people were moseying down Main Street.  The wind and rain had already begun clearing the streets before the sirens sounded, and no one seemed to be in any additional hurry when they heard the beautiful wail.  It can argued that the sirens were sounded appropriately in that case, but the public mindset is that the sirens are “tornado sirens”, so sounding them for non-tornadic events (especially events that posed such a dubious threat) does a disservice to the public because it increases complacency.  In this specific case, the sirens added nothing helpful, and thus should have remained silent.

Were this an isolated incident, I would not have felt compelled to write this post, but TEMA during the Mark Kirby era has been quick to sound the sirens.  In my circle of meteorological friends, there are two common consequences to rainfall: 1) the Indianapolis radar goes out of service, and 2) the tornado sirens are sounded in Tippecanoe County.  If I’ve associated the sirens with rainfall, surely there are others in the county who have done so as well.  So who benefits from sounding the sirens so much? No one.