US Building construction practices, revisited

In 1999, after the 03 May tornado outbreak of that year, I wrote a web essay on home construction practices and how that affects wind damage.  The recent December tornado events have re-awakened this topic, and it seems appropriate to offer some additional remarks after 16 years have passed with virtually no comprehensive change in construction practices.

The damage surveys I (and others) have done since then have continued to reveal not only the inadequacy of existing building codes, but also just how widespread violations of existing building codes are.  The existing codes remain, for the most part, pegged to a 90 mph standard for resistance to structural damage from winds.  The operational EF-scale puts 90 mph winds (a 3-second gust) at the low end of EF-1 tornadic winds (86-110 mph).  According to this standard, an EF-1 tornado (or anything stronger) is considered to be capable of initiating structural damage.  This isn't a very good standard for most of the United States east of the continental divide.  If a home is poorly anchored to its foundation (which is, unfortunately, all too often what is observed in American frame homes, despite such practices being below code standards), an 80 mph wind might well be able to slide it off the foundation, resulting in total loss of the home.

The reason for the widespread occurrence of code violations (in all buildings, including schools, not just homes) is simple.  Although structural enhancements can be added to new construction for about $1000 or so, the real issue is the time it takes to add those clips and strapping (see Fig. 1) to the frame and roof.  The added cost to the homeowner (passed on by the builder), amortized over a 3-year mortgage, is trivial.  For homebuilders and contractors, the key to profit is speed of construction.  In far too many cases, this means "shortcuts" are taken by the builders.  For instance, the code standard for attaching the wall plate to the foundation is the use of "J-bolts" embedded in the concrete, with washers and nuts tightened onto the threaded end of the J-bolt (see Fig. 1).

Figure 1.  An example of strapping used to attach a wall stud to the wall plate and the use of a washer and nut to attach the wall's bottom plate to the concrete foundation via a J-bolt.  Toe-nailing the stud to the bottom plate is poor practice (but acceptable by most current building codes) because it offers little resistance to forces acting to lift the wall.  The strapping uses nails (or screws) that are at right angles to lift forces, thereby creating much more resistance to those forces. 
Image courtesy of Tim Marshall. 

Surveys after tornadoes reveal too many examples where builders have installed the J-bolts, but failed to attach the washer and nut to the end!  That makes the J-bolt completely useless but of course it saves time!  Another common practice is for builders to be granted an "exemption" (by local governments) from codes requiring the use of J-bolts, allowing builders to use powder-driven cut nails for attaching the wall bottom plate to the concrete foundation.  This results in an extremely weak attachment (Fig. 2) when lift forces are applied to that attachment, as they are in tornadoes.

Figure 2.  A powder-driven cut nail left behind in the foundation after the wall bottom plate was torn away.  Note the damage to the concrete caused by the process of driving the nail through the board into the concrete.  Sometimes this shatters enough of the concrete to utterly negate the attachment of the nail but it would not be visible to the builder.  Image courtesy of Tim Marshall.

In my surveys, I've seen that shoddy, sometimes appallingly weak construction practices are widespread.  Buying an expensive home is no guarantee of high quality construction.  We see about as many code violations in expensive homes after tornadoes hit as we see in low-cost tract homes.  We also have seen that in many (not all) cases, the homes rebuilt after tornadoes are no better constructed than the destroyed homes they replaced.  The lessons learned from previous tornado event evidently are not being used widely to change construction practices.

If the standard for wind resistance to structural failure were raised in the tornado-prone parts of the US (everything east of the continental divide) to 120 mph (in the middle of EF-2 tornado winds - 111-135 mph), this would result in a substantial reduction in tornado damage.  Structural damage would primarily be associated with EF-3+ tornado events.  Flying debris in tornadoes from structural failures initiates considerable additional damage and increases the casualty risk to anyone caught in the tornado (Fig. 3).  Enhanced construction codes would thereby reduce tornado damage considerably and also lower the casualty risks.

Figure 3.  Damage from the 03 May 1999 tornado in Oklahoma City/Moore, OK.  Note the prevalence of broken 2X4 timbers within the debris - these are structural frame elements that may have been carried considerable distance by the winds.  FEMA image taken by C. Doswell.

Note also that even in the strongest tornadoes (EF-5), only a tiny fraction of the damage path experiences the strongest winds (at most only a few percent of the total damage area).  And EF-3+ tornadoes represent only about 10% (or less) of all tornadoes.  Limiting damage to parts of a tornado track with EF-3+ winds would represent a significant reduction in the amount of structural damage, and reduce casualties, as well.

At the very least, there's a need for more rigorous enforcement of building codes.  It would be an important advance if all buildings actually were built according to the existing building codes, to say nothing of additional benefits from enhanced code requirements.  We as a nation need to be benefiting  from our experiences with tornadoes, not ignoring the lessons they've provided.

Just who is responsible for your safety in tornadoes?

Media coverage of the late fall/early winter tornadoes that have been occurring this year includes their entirely too common efforts to seek out and give voice to those victims who say "the tornado struck without warning", despite the facts almost always showing precisely the opposite.  In most examples, timely forecasts and warnings were issued in advance of the event, sometimes literally days in advance!  I've seen this happen almost without fail every time we have major tornado impacts - over the course of my 40-year career, this has been an element in media coverage virtually in every example.

I consider this to be irresponsible reporting.  According to their clearly biased view of things, if people actually received a warning, that's not news.  It's only newsworthy when the perception is that the meteorologists dropped the ball and failed in their responsibilities.  The media should be ashamed for perpetuating the public misconceptions about the weather information meteorologists are providing for them.  So how does this false perception arise so frequently and consistently?  Why can such interview subjects be so readily found?

I've not done the surveys and research, but it seems pretty clear to me that when people interviewed by the media make the counterfactual "it struck without warning" statement, what they mean is that no one contacted them personally and told them that they needed to take shelter.   Sometimes they say that they didn't hear any tornado sirens, as if that's the only medium by which warning information can be conveyed.  Sirens can reach people who are outside in the vicinity of a siren - they're not the best and most important mechanism for disseminating tornado warning information!

What this mostly indicates to me is that some people - often those whose story is featured by the media - are simply not accepting any responsibility for their own personal safety.  In many cases, there's information about tornado threats that's available many hours, and even days, in advance of the storms.  The main point of providing this information is simply to let people know that there's a heightened tornado threat and that they need to maintain situation awareness during the time before storms develop and approach them.  Although such forecasts are not inevitably followed by major tornado events, they're sufficiently accurate to serve their intended purpose of helping people be prepared should the need arise.  Even in a major tornado outbreak, most people are not affected.  It's only the unlucky few who find themselves in a tornado's path.  But everyone should be responsible enough to keep up with the developing weather situation in cases of an enhanced threat level.

Once storms develop, the warnings are issued - in most cases, at least 15 minutes or so before the impact of the approaching tornado.  The warnings are not perfect and many of them turn out to be false alarms.  As noted, even when tornadoes do occur, the vast majority of people are not struck.  The state of the science simply won't permit perfectly accurate, extremely precise warnings and the number of perceived false alarms can only be reduced slightly by applying the best knowledge science has to offer.  The cases with the types of storms that produce the powerful tornadoes responsible for most fatalities are already handled pretty well by the forecast/warning meteorologists.  The primary problem with reducing false alarms is that it increases the likelihood of failing to warn for a tornado.  So what do people want?  Relatively frequent false alarms, or relatively frequent failures to issue a warning for an actual tornado?  Those are the only options.

It's not now possible, nor is it ever likely to be possible, to issue tornado warnings only for people who eventually will be struck by a tornado.  That's an ideal very far from the reality of what we meteorologists can do.  Even given that, however, it's evident that tornado forecasts and warnings have been saving lives here in the US since they began in 1952.  That means many thousands of fatalities have been prevented by a system that isn't even close to perfect!  While the forecasts and warnings can be improved with improved science, the existing products are not "broken"!

There's an asymmetry in the penalty function for tornado warnings.  There are ZERO tornado fatalities in a false alarm!  However, failing to issue a tornado warning for a fatality-producing tornado has a much higher penalty.  The result is a tendency to over-warn.  Tornado warnings are biased toward overforecasting tornadoes because meteorologists have a binary decision to make:  a warning forecaster either warns or s/he doesn't warn.  It's possible to reduce the bias for overwarning by issuing warnings with graded threat levels - in effect, a probabilistic threat forecast - rather than a yes/no forecast.  But such a system has yet to be implemented operationally, in part because of public demand for a virtually nonexistent certainty regarding what will happen in the weather.

In today's world, most people have many different options by which they can receive tornado forecasts and warnings.  Almost all of them require the user to make the decision to become situation aware.  People cannot simply assume zero responsibility for their own safety without running the risk of suddenly finding themselves in mortal danger.  Information about tornado hazards is readily available but people must make the effort to seek out that information without having to be told to do so.  They must plan for tornado hazards well in advance and take it on themselves to seek out information about what they can do to reduce the threat to their lives (and even their property).

The media need to become responsible for telling an accurate version of tornado events, rather than continuing to reprise the counterfactual scripted version of events that reinforces the myth of "it struck without warning".  The media have some responsibility here and to my mind, many of them are failing to carry out that responsibility.  The men and women who dedicate their lives to providing the public with the most accurate and timely weather information they can muster deserve our respect and admiration for their selfless efforts to inform.  They do not deserve to be portrayed as failing in their duties when the facts are clearly contrary to the media script.  Their failure to achieve perfection is far from being entirely their fault.