Opening Statement—NIST Joplin Tornado News Briefing
Nov. 21, 2013
Eric Letvin, Director of Disaster and Failure Studies
National Institute of Standards and Technology (NIST)
Thank you, Howard. Good morning.
I'm pleased to be here this morning to talk with you about the major findings and recommendations from NIST's technical investigation of the Joplin tornado and its impact on the community.
As Howard has explained, NIST has legislative authority to investigate building failures with the intent of recommending improvements to building codes, standards and practices. Our goal with these disaster studies is always fact-finding, not fault-finding. To make the best use of our resources, we use specific criteria to carefully choose which events to investigate.
The Joplin tornado of May 22, 2011, was clearly an unusually powerful and tragic event. The tornado caused 161 fatalities, more than 1,000 injuries, and damaged close to 8,000 buildings, making it the deadliest and costliest single tornado since official record keeping began in 1950.
NIST deployed a team of experts to Joplin three days after the tornado. We quickly concluded that the large number of deaths and the specific types of building and infrastructure damage seen with this tornado, made it especially important to learn as much as we could about what happened and how lives and property might be better protected from such tornadoes in the future.
Our draft report is very detailed and has 47 findings and 16 recommendations. I'm not going to be able to describe them all this morning. Instead I'd like to tell you briefly about a few particularly central findings and recommendations. I'll also briefly explain where additional research is needed to fill gaps in our knowledge.
The overarching conclusion of our two-year study is that death and destruction from tornadoes can be reduced.
Based on a significant body of research and observations from such events, our scientific understanding of tornadoes and their effects has matured substantially. It's time to begin developing and implementing nationally accepted standards and codes that directly address tornado hazards.
Through stronger construction methods, enhanced emergency communications, and greater access to tornado-safe shelters, we can reduce the tremendous human and economic cost of these events on our communities.
We reached this conclusion by studying the following aspects of the Joplin event: tornado characteristics; building and infrastructure performance; human behavior; emergency communications; and the impacts of each on protecting people from tornadoes. To our knowledge, this is the first scientific study to address the intersection of these distinct areas for a specific tornado.
Our study team included experts in structural and fire engineering, extreme wind, meteorology and severe storms, and sociology for analysis of human behavior and emergency response. We were fortunate to have essential assistance from a number of different organizations. We want to specifically thank the City of Joplin, the State of Missouri, the National Oceanic and Atmospheric Administration, especially the National Weather Service and the National Severe Storms Laboratory, and the Federal Emergency Management Agency.
Our data gathering and subsequent analysis efforts included:
- estimation of tornado wind speed and other characteristics based on observed damage;
- field surveys of 25 key buildings in the damage area shortly after the tornado, and subsequent review of building designs and structural analyses to determine their performance during the tornado;
- analysis of the emergency communications that took place prior to and during the tornado and the collection of experiences through interviews with 165 survivors, emergency responders, and family members of victims; and
- documentation of the pattern, location and causes of fatalities and injuries attributed to the tornado.
We constructed a detailed timeline of the event and created a computer model for estimating the wind environment. The model was optimized by comparing thousands of computer-generated tree-fall simulations to actual observations of fallen trees in Joplin. The best comparisons were used to estimate tornado wind speeds and directions.
We collected and examined design and construction drawings for some of the NIST-surveyed buildings to aid in developing an understanding of the sequence of failure for these structures.
Through speaking with tornado survivors, we also developed a detailed understanding of the public response before and during the tornado. We're grateful to many Joplin residents who provided information through face to face and telephone interviews, as well as to those who helped us recruit people to interview. To determine the effects of emergency communications on behavior, we studied the emergency communications systems in place in Joplin and Jasper County in May 2011. In addition, we conducted a qualitative analysis of other tornado warning systems throughout the United States.
We reviewed death certificates and obituaries, interviewed friends and families of victims, and collected news media and social media reports. We then combined this knowledge with analysis of the wind environment and building performance, and identified the factors that likely caused the impact–related deaths and severe injuries that resulted from the tornado—both outside and inside of structures.
We found that, in most cases, neither residential nor larger commercial buildings in Joplin adequately protected building occupants. A large majority of the fatalities, almost 84 percent, were due to impact-related causes associated with building failures.
Virtually all of the buildings where these fatalities occurred experienced maximum estimated wind speeds associated with EF-3 or higher on the Enhanced Fujita scale, or greater than 136 miles per hour. There were no community shelters in Joplin in May 2011 and residents had limited access to underground or tornado-resistant structures.
The outdoor emergency sirens alerted many people that an emergency was occurring. However, conflicting information from multiple sources caused confusion for many residents about the true nature of the hazard. As a result many people sought additional informational rather than immediately seeking shelter. A delayed or incomplete response by some was also caused by a lack of awareness of the tornado, inaccurate assumptions about Joplin's vulnerability to tornados, and by distrust or confusion about Joplin's emergency communications system.
These findings help explain why the fatalities and devastation were so great. With our findings in hand, the study team then looked at the current building codes, standards and practices to see what improvements might be possible.
Current U.S. model building codes include requirements to protect against many different types of hazards, including hurricanes, earthquakes, and floods. They do not include requirements to protect against tornado hazards, which include extreme wind speeds and impacts from wind-borne debris. The only exception where tornado hazards are considered is for safety-related buildings of nuclear power plants and for tornado-specific storm shelters or safe rooms.
To drive home this point . . . buildings in Joplin performed as we expected. They were not designed to withstand tornado winds and, not surprisingly, many failed.
In addition, many tornado-prone areas, including Joplin in May 2011, do not mandate the construction of shelters or safe rooms. As a result, many residents, particularly those living in multi-family apartments and nursing homes or those who found themselves in high-occupancy buildings such as hospitals, schools, or large commercial spaces when the tornado struck did not have sheltering options. There are also no nationally accepted standards for emergency communications to warn residents effectively of imminent tornado hazards.
To help address these needs, NIST developed 16 recommendations for improving how buildings and shelters are designed, constructed, and maintained in tornado-prone regions and for improving emergency communications.
I urge you to read the report available on the NIST web site for this full list.
Here are three key recommendations where implementation is likely to have the greatest impact:
- Develop national performance-based standards for tornado-resistant design of buildings and infrastructure, as well as design methods to achieve those standards, and require that essential facilities such as hospitals, be designed to remain operational in the event of a tornado.
- Install tornado shelters in new and existing multi-family residential, commercial, and other larger buildings, such as schools. Typically, the types of buildings we're talking about here are called "assembly occupancies" in the building codes. This includes places such as large restaurants, places of worship and theaters that accommodate 50 people or more at a time. As part of this effort, develop and implement uniform national guidelines to help communities design, install, and operate those shelters.
- Create national codes and standards for clear, consistent and accurate emergency communications and then ensure that emergency managers, the National Weather Service, and the news media in local communities have a joint plan for delivering those messages quickly and persuasively during tornados.
We believe that these recommendations are realistic, appropriate, and achievable within a reasonable period of time. We strongly urge state and local authorities to adopt and enforce model building codes and standards that currently exist and to update local codes as new standards aimed at improving resiliency to tornados are developed.
We'd also like to stress that NIST is assigning top priority to work vigorously with the building and relevant communities, including other federal agencies, to assure that there is a complete understanding of the recommendations and to provide technical assistance for implementing these recommendations into standards and codes.
Finally, we urge funding agencies to sponsor research on advanced measurement technologies that can provide faster warnings of dangerous winds, can quickly produce maps showing emergency managers which local areas are most likely to face the greatest hazards, and can "push" appropriate warnings to residents in those areas in time to seek shelter.
A single tornado affects a smaller geographic area than a hurricane and or an earthquake, but tornadoes happen much more frequently and kill more people in the U.S. each year than these other natural disasters combined. We know how to make structures that can withstand the majority of tornadoes, but so far we have not included protection against them in model codes and standards.
We can save lives, minimize disruption to our communities, and lower the economic cost of tornadoes. The time is right to develop and implement codes and standards that better protect our citizens and help communities recover more quickly from these powerful, but not invincible, natural forces.
At this point I'll be happy to take your questions.
Editor's Note-NIST would like to thank Missouri Southern State University for hosting the news briefing.