What has Katrina taught us about how to better lessen risk from hurricanes?

Katrina taught us that we need to know more about hurricanes. We have all worked together to learn more and now we are all working to apply it. We have made improvements. We are continuing to work towards further improvements, and we will move forward to the 100-year level of protection – but we will not stop there. We know that multiple lines of defense are required, so we are looking at additional alternatives, such as structures/barriers in the Gulf and coastal wetlands restoration. Other actions can be taken in addition to the protection systems to lessen risk.  We all have a role to play in risk reduction. And all of these activities carried out in an integrated manner will give us the complete picture of risk management.

Who commissioned the IPET study and why?

The IPET was established by Lt. Gen. Carl Strock, the 51st Chief of Engineers, on Oct. 10, 2005, to determine the facts concerning the performance of the Hurricane Protection and Flood Damage Reduction System in southeast Louisiana in response to Hurricane Katrina.

IPET was tasked to address five basic questions on:

1. the System (what was the status of the protection system on 29 August 2005);
2. the Storm (what exact forces did Katrina put on the system);
3. the Performance (how did the system respond);
4. the Consequences (understanding the flooding and the losses); and
5. the Risk (what is the risk and reliability of the system). 

What does risk mean?

In the simplest of terms, risk is a measure of harm or loss associated with an action. For this particular situation, risk is the likelihood of loss of life or property as a result of flooding caused by hurricanes.

Explain the formula for risk? 

In the simplest terms, the risk assessment basically incorporates three main ingredients: the hazard, the protection system, and the consequences. 

The Hazard is the future probability of hurricanes, their surge and waves.  This gives us the water levels we can expect from future storms.  IPET used a new, advanced hurricane modeling method to look at storms ranging from 50 to 5000 years in possible occurrence.  We ran 152 possible future storms on the fastest supercomputers available to get the water levels we could expect across the complex Louisiana coastal area on the 350-mile protection system for the entire range of storms.  These storms had variables that included barometric pressure, maximum winds, storm size, how fast the storm was moving, and direction of storm track.

The Protection System involves determining the performance of levees, floodwalls, and other structures once we put the water levels from the storm hazard against them.  We looked at 134 reaches of levees and floodwalls that represented different areas of the protection system.  We also looked at 350 specific structures such as gates, pumps, transition points, etc.  For each reach or structure, we keyed in info on their elevation, design, construction, maintenance, soil foundation conditions, how levees would behave if overtopped, etc.  This gave us the reliability of the protection system from a system wide aspect, but we can also look at individual structures or areas.  This also gives us the expected flooding from larger storms.  We did today’s assessments on the pre-Katrina protection system (how it was on 29 August 2005) and current day (how the system looked on 1 June 2007). The New Orleans District will use these same tools to examine the 100-year protection system later this year. 

The Consequences involves the anticipated loss of life or property.  By putting the results of the Protection System reliability over population and property information, we can determine the expected levels of flooding and subsequent losses (life and property) from small and large storms.  We used pre-Katrina population and property information as our base information. We also assumed no evacuation on our population numbers.

What was involved in performing the risk analysis?

Risk assessment involved almost 63,000 hurricane hydrographs or water level records; 1,450 reliability relationships (protection system); and 68 consequence relationships (life or property).
Each of the 134 levees or floodwall reaches and each of the 350 specific gates, transitions and other features in the protection system had 152 storms run for the pre-Katrina and current day conditions. Each scenario estimated water entering a protected area by rainfall, overtopping and possible breaching. Chance of flooding to different levels was determined for each natural drainage sub-basin. Potential loss of life or property was determined for different flood levels by sub-basin. The chance of flooding and the losses are combined to estimate the risk.

How reliable is the risk modeling?

The IPET realized there would be uncertainties with the risk assessment. One of the reasons for the delay in the release of the risk information was a result of the thorough validation process that was done on all of the components of the risk assessment. This is the prototype risk assessment for the hurricane protection system, and IPET wanted it to be as reliable as possible. Uncertain information will be presented in a range, such as plus or minus one foot, for products that depict a chance of flooding. Any uncertainty factors will be stated in the risk assessment products.

How was the risk model tested to ensure accuracy?

The IPET ran Katrina through the risk model and got almost the exact same results (areas that flooded, damages, etc.) that Katrina actually produced. Also, the new hurricane modeling method used in the risk assessment was validated using information from six historic storms. These storms, when run through the hurricane model, produced the same storm results. These storms were: Rita, Katrina, Camille, Ivan, Betsy and Wilma.

What areas are covered by the risk analysis?
The risk assessment covers Jefferson, Orleans, St. Bernard, St. Charles and Plaquemines parishes. 

Did you model wind?

The IPET modeled large-scale wind fields to model hurricanes and surge and wave generation.  We did not model hurricane winds with respect to wind damage to structures. The damage we look at comes from flooding caused by hurricane storm surge and waves. We also added rainfall. We looked at threats that the hurricane protection system can counter. Wind is not one of them.

Did you take into account evacuation of people?

IPET risk assumed NO evacuation of people since we could not determine accurately who would evacuate. All risk to loss of life assumes NO evacuation. Therefore, evacuations would greatly reduce the risk of loss of life. Evacuation plans should always be considered for future storms.

Does the model factor in breaches and potential deeper flood depths they would cause?

Yes, the risk model includes overtopping, potential breaching, potential pump failure and hurricane-associated rainfall as its sources of water. The “current condition” maps reflect all system repairs and improvements in place as of June 1, 2007.

What will the risk products show?

Risk products can show a variety of useful information for planners and the public to make informed decisions. Information products can be tailored to show or address specific concerns, such as overall risk by sub-basin, risk for specific features (gates, transitions, etc.), risk to people or property, etc. They will show depth of flooding and the annual chance for loss of life and property.  They will also show hurricane protection system areas that are more vulnerable than others.

What can I learn from the risk analysis?  How will it help me?

From the risk analysis we can see areas that have an increase of 5 to 6 feet of protection with the current conditions from what they were before Katrina. Other areas have 1 to 2 feet of increased protection. Using the risk information, planners can see what areas throughout the whole system or even specific structures within it that need improvement. Our intent is to provide information in these analyses for everyone – officials and the public – to make their own decisions. 

This information provides the ability to make more informed decisions, such as whether to raise a home, move or take other measures to increase personal safety. Risk and reliability analysis can also provide information about:

• How various future storms might affect different areas of the protection system;
• The average annual chance for flooding and to what depth;
• Possible weak points in the system;
• Improvements that provide the best protection; and
• Risky areas for property or people and which may need redevelopment limitations.

Explain pre-Katrina protection, current and 100-year levels of protection.

The pre-Katrina level of protection is the system as it existed on 29 August 2005 prior to Katrina’s landfall.  Current protection reflects the protection system as it existed on 1 June 2007.  The 100-year level of protection will protect the area from a 100-year storm, which has a 1-percent chance of occurring in any given year.
 
IPET is presenting risk assessment products related to the pre-Katrina and current conditions of the protection system.  Risk information for 100-year protection is not yet available. We expect to have that information later this summer.

Does this study prove that the City really needs “Category 5” protection?  Do you have a model that can show us how safe we’d be with Category 5 protection?

The Saffir-Simpson hurricane scale, the Category 1 to 5 system, relies on wind and barometric pressure to measure hurricanes. It is useful for weather forecasting, but woefully inadequate for coastal protection decisions. For instance, in 1969 Hurricane Camille was a Category 5 hurricane when it hit the Mississippi coast with 19- to 20-feet storm surge. Katrina was a Category 3 storm that hit the Mississippi coast with a 28-feet storm surge, the highest to hit a U.S. coastline. The IPET and other research show that storm size and intensity are more important than the Category 1 to 5 scale. For engineering decisions related to coastal protection, it is more realistic to use a probability for a certain destructive storm. A 100-year storm means that such a storm has a 1 percent chance of occurring in any given year. Katrina was a 400-year storm or a hurricane with a 0.25 percent chance of occurring in any year. Research shows that specifying a category of storm to define the threat and the appropriate level of protection is not realistic or sufficient. This risk study gives us the tools and information needed to adequately design the system needed for large storms that produce significant surge and waves.

How can you model the current population and property of the area now?

The risk assessment used pre-Katrina population and property information.

If a Katrina-like storm hit this summer, would New Orleans be flooded again? 

Katrina was approximately a 400-year-storm (one quarter of 1 percent or 0.25 percent chance of occurring in any year). Katrina, even if the levees had not breached, would have produced the worst flooding the greater New Orleans area experienced in recent history (approximately one-third of flooding experienced would have occurred from rainfall and overtopping). A 400-year type storm in the future would be very damaging and could cause significant flooding, especially from overtopping. There is less likelihood of catastrophic breaching because of the quality of the repairs made.

This looks like a great way to forecast damage to the area from specific storms.  How useful is the risk model in this respect?

The risk model is not useful to forecast damage from a specific storm. It is designed to provide a “big picture” look.

At 100-year elevations, most people would expect to be dry.  Will we see flooding from a 100-year storm? Can the 100-year level hurricane be run through the risk model to show what would happen?

A 100-year storm would cause some flooding in some areas since the entire protection system is not yet up to 100-year storm levels. The risk model uses 152 different storms to accurately reflect risk from a range of storms with differing tracks in order to provide a more accurate picture of risk. 

Are things better today?  What about with higher levels (100-year elevations) or even the proposals under LaCPR.  Will there still be risk?

Yes, things are better across the entire protection system. The system is stronger and offers improved protection from the pre-Katrina conditions. Where repairs have been made, levees are higher and made of clay, and failed I-walls have been replaced with “T” or “L” walls. These improvements have reduced the likelihood of catastrophic breaching and decreased the amount of potential flooding that could be experienced in certain areas. Continuing improvements ensure continued reduction in risk. The proposals under Louisiana Coastal Protection and Restoration (LaCPR) will address higher levels of protection, further reducing risk. But there will always be risk associated with living in a coastal area. The Administration is committed to providing 100-year level of protection to the New Orleans area, and the Corps is working towards that goal.

When will we get 100-year protection?

Public safety is the top priority.  The Corps has worked very hard. Even its harshest critics would say the Corps and its employees accomplished a Herculean task last year by completing the levee repairs in only 7 months. They worked 24/7 to accomplish this. The goal to repair and enhance the entire 350-mile hurricane protection system is by the end of the decade, but we want to caution people that the Corps wants to do the work right, and if that means taking extra time to ensure a better product, then that’s what we’ll support.

What will IPET do next to update or improve the risk analysis?

The IPET is finishing up its tasks and with the release of the Risk and Reliability report, Volume 8, will have completed its work. The IPET will address comments by the ASCE and NRC on Volume 8 and then finalize the volume later this year. Updates and improvements to the risk analysis will be made by the Corps and the New Orleans District in the future.

How will the risk model be used in future?

New Orleans District will use the IPET risk models to determine risk for 100-year elevations for the protection system and in the Louisiana Coastal Protection and Restoration study. We expect simplified versions of the risk model to be used by other coastal areas. A similar risk assessment has been initiated by the state of California for the Sacramento area.