In 2005, Hurricanes Katrina and Rita, aided by levee failures, collectively damaged over 200,000 homes, killed over 1,400 Louisiana

Transcription

1 Introduction The New Orleans Index at Ten Current and Future Flood Risk in Greater New Orleans David R. Johnson, RAND Corporation Jordan R. Fischbach, RAND Corporation Kenneth Kuhn, RAND Corporation In 2005, Hurricanes Katrina and Rita, aided by levee failures, collectively damaged over 200,000 homes, killed over 1,400 Louisiana residents, and displaced more than a million others. These events marked flood risk as the primary climatic threat to the vibrant culture of New Orleans and the rest of the Gulf Coast. 1,2 The average intensity of hurricanes is predicted to increase, 3,4 which emphasizes how critical it is to be prepared for the next major storm. Since 2007, over 160 miles of levees have been constructed, repaired, or upgraded in response to the crisis. The Greater New Orleans Hurricane and Storm Damage Risk Reduction System (HSDRRS) received $14.5 billion in federal and state investments to protect the metropolitan area: the US Army Corps of Engineers (USACE) has raised levee and floodwall elevations, built a massive new surge barrier, installed pumping stations, and constructed new canal closure structures. The Federal Emergency Management Agency (FEMA) has certified the upgraded HSDRRS as protecting New Orleans to at least the 100-year level meaning that there is less than a 1 percent chance of flooding occurring in a given year. 5 The new system was baptized by Hurricane Isaac in August 2012 and suffered no significant overtopping or breaches. 6 These risk reduction efforts have been guided by Louisiana s Coastal Protection and Restoration Authority (CPRA), which is tasked to produce a long-term coastal master plan for flood risk reduction and coastal restoration. 7 This essay presents estimates of flood risk in New Orleans associated with the current state of economic redevelopment and under the present-day configuration of HSDRRS. Previous studies outside of the CPRA master planning process have only examined intermediate stages of construction or alternative system designs, so this paper is the first to quantify the accomplishments of the post-katrina efforts to improve coastal defenses. We conclude that significant risk reduction, meeting a 100-year standard, has been achieved, but that risk may increase in the future unless levees are maintained or further upgraded. FIGURE 1: DEFINITION OF SPATIAL REGIONS REFERRED TO IN DISCUSSION

2 2 August 2015 History of the challenge Neither flooding nor flood control are new issues for New Orleans. Although the earliest levees and drainage canals were built in the early 1700s to tame the Mississippi River, the city has always been prone to both riverine and storm surge flooding. 8 Nearly 200 hurricanes have impacted the Louisiana coastline in the last 450 years. 9 Though Hurricane Katrina inflicted massive economic damage and loss of life, it is far from the first great flood to have struck the city. Modern flood control efforts have protected New Orleans from the Mississippi River, which has not flooded the city since The downside of this achievement is that the Mississippi s course has been locked in by 3,500 miles of levees. Over-bank flows have all but halted, diverting sediment into the Gulf and preventing it from counteracting natural land subsidence and erosion. 11 Four years before Hurricane Katrina, it was noted that land loss in coastal Louisiana could result in New Orleans being exposed to open water by Billions of dollars in CPRA s current Coastal Master Plan are dedicated to reversing land loss trends where possible, and rebuilding the coastline to serve as a natural barrier against storm surge. 13 Storm surge from a hurricane in September 1947 spurred increased federal investment in protecting New Orleans, culminating in the Lake Pontchartrain and Vicinity project, first approved in Plans were altered perpetually, however, after initial drafts were deemed too costly, levees failed during Hurricane Betsy, lawsuits halted various proposals, and momentum stalled. Sixty years later, most, but not all, areas were considered 90 percent complete, but the city was still highly vulnerable when Katrina struck in ,15 A diagram of the present-day HSDRRS layout (in bold black lines) is shown in Figure 1, colored according to the main polders and subsystems referred to in later discussion. The St. Bernard region consists of neighborhoods on the East Bank that lie south of the main outfall canal systems and the Inner Harbor Navigation Canal (IHNC). 16 New Orleans East consists of areas like Michoud that are north of the IHNC surge barrier and east of the Seabrook floodgate complex. New Orleans Main consists of portions of Orleans Parish west of the IHNC. East bank locations in other parishes are designated as St. Charles-Jefferson East Bank. Flood protection in a post-katrina New Orleans The magnitude of Katrina s destruction was enabled not only by the storm s intensity, but also by decades of indecision, errors in design, and insufficient funding for maintenance. 8,17,18 Recognizing this, CPRA s first Comprehensive Master Plan for a Sustainable Coast approved in 2007 outlined the environmental crises facing the region, defined the state s vision for coastal sustainability, specified the types of protection and restoration projects to be considered by future plans, and developed high-priority project concepts. 19 The plan supports a multiple lines of defense strategy for flood protection that combines many types of measures, including levee and floodwall systems, nonstructural mitigation efforts such as home elevations and floodproofing, and coastal restoration measures like marshland restoration and river diversions The first update to this plan, passed unanimously by the state legislature in May 2012, formalized the analysis and identified a longterm investment strategy. The 2012 Coastal Master Plan is a 50-year, $50-billion initiative that divides its funds between structural risk reduction projects (including further upgrades to HSDRRS), nonstructural risk reduction mechanisms, and coastal restoration measures. 13 All of this is above and beyond the extensive changes to HSDRRS already implemented: in some areas, levee elevations have already been raised more than 10 feet above pre-katrina levels. As noted earlier, FEMA has certified the current HSDRRS as providing defense against a 100-year storm surge for the portions of Orleans, Jefferson, St. Bernard, St. Charles, and Plaquemines parishes it is designed to protect. 5 No studies, however, have yet published estimates of the risk reduction benefits the system actually does provide. 24 Further, the FEMA certification does not necessarily mean that 100-year damage levels within HSDRRS are zero. 25 Measuring current and future flood risk under various assumptions about system failure, hurricane and other environmental characteristics, and development trends is one of the major tasks of the 2017 Coastal Master Plan. Initial results from this work are presented here. 26 THE COASTAL LOUISIANA RISK ASSESSMENT MODEL CPRA recognized that new modeling capabilities were needed to efficiently estimate flood risk under a wide range of future uncertainties and with various proposed projects in place. The Coastal Louisiana Risk Assessment (CLARA) model was designed to meet this need, and it accounts for many environmental, operational, and growth uncertainties. 27 It can, for example, easily change the assumed elevation of a particular levee segment and calculate the impact. This allows for examination of different assumptions about land subsidence or levee maintenance, or evaluation of projects that upgrade alignments to varying heights. CLARA produces estimates of flood depth exceedances at any specified return period, ranging from 10- to 2,000-year flood depths. In

3 3 August 2015 DATA C E N T E R R E SE A RC H. ORG areas not enclosed by a ring levee system like HSDRRS, storms are run through a high-resolution hydrodynamic model. Flood depths from each storm are then aggregated to generate statistical exceedances. This relies on an assessment of the relative likelihood of observing storms with different characteristics, based on statistical analysis of the historic storm record To model risk within enclosed protection systems, the surge and wave behavior produced by storms along the system boundary is run through a Monte Carlo simulation of levee breaches in which the probability of failure is calculated as a function of the peak overtopping rate. 26,30 Flood depth exceedances are then derived from the probability distribution of flooding resulting from each storm and the likelihood of observing a real storm like those modeled. The model also projects direct economic damage associated with flooding. Damage exceedances are statistically aggregated to calculate expected annual damage (EAD). Damage calculations follow the general approach used by FEMA s Hazus model. 26,27,31,32 MODELED SCENARIOS AND UNCERTAINTY Some factors impacting flood risk are deeply uncertain, in the sense that the underlying probability distribution governing a parameter s value is unknown or not agreed upon. The impact of these deep uncertainties can be explored using scenario analysis. For example, the model can run scenarios with different assumptions about future sea level rise, land subsidence rates, and changes in the frequency and intensity of Atlantic hurricanes. This paper focuses on two such scenarios. One scenario represents current environmental conditions in The other is one plausible vision of the future in 2065, using assumptions from the 2012 Coastal Master Plan s Less Optimistic future scenario. 33 These two scenarios were used during the 2017 Coastal Master Plan s model development and testing phases to support model improvements. CLARA can make different assumptions about the probability of system failures from overtopping, backside scour and erosion, slope stability, or seepage. 26 The modeled probabilities of failure have a dramatic impact on the resulting flood depth exceedances within HSDRRS. In addition to a case where no failures are assumed to occur, the fragility curves used by CLARA are based on the Interagency Performance Evaluation Taskforce (IPET) study and the 2013 Morganza to the Gulf (MTTG) Reformulation Study;30,34 a Low and a High fragility case was developed using assumptions from each of the two studies. Except where noted otherwise, maps depict the MTTG Low scenario, which was chosen for being an intermediate case. Other uncertainties are incorporated into confidence intervals around estimates of average flood depth or damage values. These sources of uncertainty include inaccuracies in the hydrodynamic storm models, errors in measurements of ground elevations, variability in levee overtopping rates, and uncertainty in the estimated likelihood of different storms associated with having a small historic record of past storms. The combined effect of each of these factors is reflected in the reported 10th, 50th, and 90th percentile estimates. FIGURE 2: 100-YEAR FLOOD DEPTHS (90TH PERCENTILE ESTIMATE) 2015, MTTG LOW FRAGILITY SCENARIO

4 4 August 2015 FLOOD DEPTH RESULTS IN 2015 (CURRENT ENVIRONMENTAL CONDITIONS) Figure 2 illustrates that CLARA s median assessment of current 100-year flood depths within HSDRRS is essentially consistent with FEMA s certification of 100-year protection within the city. Even the 90th percentile of estimates show virtually no flooding in New Orleans at a 1 percent AEP except for a few unpopulated locations within Violet Marsh and New Orleans East, and minor flooding in low-lying parts of the east bank in Jefferson Parish. 35 This finding holds across the different scenario assumptions regarding system failures. Generally speaking, the geographic distribution and overall extent of flooding is similar between fragility cases at the same return period, although depths may vary to a slightly greater degree. However, the distinctions between fragility cases are more readily apparent when expressed in dollar amounts rather than on a map, as discussed in the section on economic damage. In much of Greater New Orleans, the upgraded HSDRRS is found to provide greater than 100-year protection. Most West Bank communities, as well as communities on the East Bank not adjacent to Lake Pontchartrain (such as Harahan, Old Metairie, and the French Quarter and Central Business District), remain dry up to approximately a 0.25 percent AEP, or 400-year return period. Hurricane Katrina is often popularly described as a 400-year storm, so this is encouraging. However, large parts of the city do see flooding at the 400-year level, including the Lower 9th Ward, other portions of St. Bernard Parish, and much of Lake Forest and New Orleans East. Flooding is more extensive at the 500-year return period and beyond. Figure 3 shows CLARA s estimated 500-year flood depths; the 10th, 50th (median) and 90th percentile values are all shown to provide some indication of the range of uncertainty in flood depth estimates. In areas such as the Lower 9th Ward, the median 500-year flood depth estimates are actually less than the 90th percentile 400-year flood depth estimates. This illustrates clearly that there can be significant uncertainty about the probability of seeing any particular specified level of flooding. The 10th percentile estimates are fairly similar in the extent of flooding when compared to the other percentiles, indicating a fair degree of confidence about the geographic distribution of where flooding may occur at a 0.2 percent AEP. FIGURE 3: 500-YEAR FLOOD DEPTHS (10TH, 50TH, 90TH PERCENTILES) 2015, MTTG LOW FRAGILITY SCENARIO

5 5 August 2015 DATA C E N T E R R E SE A RC H. ORG There is, however, an important caveat to the previous discussion. CLARA s median estimate of the 400-year flood depths in 2015 predicts that relatively little flooding is experienced with a 0.25 percent AEP, provided that the upgraded HSDRRS performs as intended and no levee failures occur. Significant portions of the new HSDRRS have been armored to reduce the chance of failing during surge or wave overtopping, so this may provide greater than 100-year levels of protection at present. FLOOD DEPTH RESULTS IN 2065 (LESS OPTIMISTIC ENVIRONMENTAL CONDITIONS) Despite these estimates, flood risk remains a constant threat to the region, and that risk will only increase with future change unless further action is taken. The Less Optimistic environmental scenario from the 2012 Coastal Master Plan illustrates one plausible future in which land subsidence and sea level rise proceed at rates on the higher end of current projections, and hurricanes become more frequent and intense. Under the Less Optimistic scenario assumptions, some areas in New Orleans are projected to experience flooding in 2065 within HSDRRS even at the 50-year return period; median estimates of flood depths with a 2 percent AEP are up to one meter in West Lake Forest, and approximately half a meter in portions of Kenner and the uninhabited areas of Bayou Segnette. FIGURE 4: 100-YEAR FLOOD DEPTHS (90TH PERCENTILE ESTIMATE) 2065, MTTG LOW FRAGILITY SCENARIO Median estimates of the 100-year flood depths in 2065 show more extensive flooding in Kenner and West Lake Forest; the Lower Ninth Ward, Chalmette, and portions of Metairie near Kenner also flood at this exceedance. The 90th percentile estimates of 100-year flooding in 2065, for comparison with the 2015 values from Figure 2, are shown in Figure 4. At the 90th percentile, Gentilly, Westwego, and parts of the Upper Ninth Ward also are projected to flood. Flooding at the 500-year return period is predictably even more severe. Ninetieth percentile estimates, comparable to those from 2015 in Figure 3, are shown in Figure 5. While HSDRRS greatly reduces flood depths on the city interior, the results of another Katrina-like storm fifty years into a future without action would likely still be catastrophic. Figure 6 summarizes the risk of flooding in each spatial region by averaging flood depths by return period over all points within each region; median estimates are shown. The unweighted average is an inexact measure, given that grid points vary in the size of the area they represent, and unpopulated areas contribute less to economic damage estimates. However, the figure still provides useful insights into risk in each region. For example, it shows that all regions are currently protected up to approximately the 200-year return period (0.5 percent AEP), where depths start to increase in the New Orleans East and St. Bernard Regions. In 2065, New Orleans East starts to see flooding at higher frequencies than other regions, and only Orleans Main and the West Bank communities maintain roughly 100-year protection.

6 6 August 2015 FIGURE 5: 500-YEAR FLOOD DEPTHS (90TH PERCENTILE ESTIMATE) 2065, MTTG LOW FRAGILITY SCENARIO FIGURE 6: FLOOD DEPTHS BY REGION AND RETURN PERIOD SPATIALLY AVERAGED, MTTG LOW FRAGILITY SCENARIO, MEDIAN ESTIMATES DATA C E N T E R R E SE A RC H. ORG

7 7 August 2015 ECONOMIC DAMAGE EXCEEDANCES Damage estimates in CLARA are generally correlated with the extent and depth of flooding. Extensive damage occurs in areas with high probability of flooding, dense development, high-value assets, and low levels of nonstructural risk mitigation. To show the range of damage estimates across different fragility scenario assumptions, Figure 7 shows the 50-, 100-, and 500-year damage exceedances associated with two fragility cases: No Fragility and MTTG Low. Stacked bars show the contribution to median damage from different areas of Greater New Orleans, while the reference lines show the range of total damage from the 10th to 90th percentile. FIGURE 7: 50-,100-,AND 500-YEAR DAMAGE EXCEEDANCE ESTIMATES BY FRAGILITY CURVE AND REGION Note the different damage scales between the 2015 and 2065 panels of Figure 7. As discussed previously, significantly greater flood depths are predicted to occur by 2065, but at the same time, damage is compounded by an additional fifty years of economic development. These results reflect the same assumptions about economic growth that were used in the 2012 Coastal Master Plan analysis s baseline scenario. 36 As expected, 50- and 100-year flood damage is low in 2015 approximately $1.5 billion and primarily concentrated in Kenner (St. Charles-Jefferson East Bank). This might be explained more by the rainfall associated with higher-frequency storms overwhelming pumping systems than by storm surge overtopping the levees. Damage at the 500-year return period spikes dramatically in all cases, however, including the No Fragility case. This indicates that even if levees hold, the current design heights are projected to be insufficient to prevent significant overtopping at a 0.2 percent AEP. The median 500-year damage estimates in 2015, ranging from $9 to $22 billion, are significantly less than the damage projected to occur in 2065 under the Less Optimistic future without action. Even disregarding the 500-year estimate from the MTTG High fragility case as an outlier, damage generally jumps by an order of magnitude over the 50-year timespan.

8 8 August 2015 The geographic distribution of damage is also notable. In 2065, the large majority of damage occurs in New Orleans Main, compared to a more even distribution of damage among the regions in 2015 (except for the West Bank communities). This can be explained by examining the flood maps in the previous section; flooding in 2065 expands into the more densely populated areas of New Orleans Main as well as into commercial areas in the central part of the city. The distribution of uncertainty in damage estimates is skewed; 90th percentile estimates generally show greater deviation from the median values than do the 10th percentile estimates. In many combinations of return period and fragility scenario, the 90th percentile estimate of damage is 1.5 times greater than the median estimate or more. EXPECTED ANNUAL DAMAGE CLARA s estimates of EAD with the newly upgraded HSDRRS in place are summarized in Figure 8. Median values are broken out by asset type, with 10th and 90th percentile estimates for the totals indicated by lines. 37 At the median, EAD ranges from approximately $180 million to $250 million over the modeled fragility cases in In sharp contrast, IPET estimated EAD under the pre-katrina system to be approximately $650 million, and Hallegatte and colleagues estimated average annual losses of $600 million. 17,38 Notwithstanding differences in the value of exposed assets between 2005 and 2015, this is a significant achievement. Examining the breakdowns by asset type reveals that the main driver of current flood risk is damage to commercial properties. Commercial and industrial assets are harder to mitigate than residences: they are generally larger in square footage (possibly too large to elevate in place), and they can be expensive or difficult to floodproof effectively. Sales and wages lost during a post-flood repair/ reconstruction period further contribute to the increased damage and losses to commercial properties. Residential damage is split primarily between single-family homes and large multi-family residences. Single-family homes are smaller and easier to mitigate through elevation, but because they make up the bulk of housing stock, their EAD is larger than it would otherwise be. Manufactured homes are numerous (approximately 9 percent of residential structures in 2015) but less valuable. Small multi-family homes consist of shotgun-style apartments and duplexes, which have generally declined in number over the years. They represent only 3.5 percent of households and are typically small enough to mitigate, so they contribute a small amount to the total residential risk pool. Large multi-family structures are also small in number, at 6.5 percent of the housing stock, but they are more valuable on a per-unit basis and more difficult to mitigate, so they comprise a larger share of total risk. In 2065, a large increase in projected damage is clearly seen under the Less Optimistic scenario. In keeping with the shift over time of individual damage exceedances, EAD may jump by an order of magnitude over the next fifty years without further intervention. Projected patterns between asset types are similar between the two time periods. Of note is the greater proportional difference in 2065 between the MTTG Low scenario and the case in which no system failure is modeled. In a future without action, it is reasonable to expect levees to degrade if not properly maintained, resulting in a higher probability of breaches when confronted by storm surge. This illustrates the importance of proper maintenance of the system. FIGURE 8: EXPECTED ANNUAL DAMAGE ESTIMATES BY FRAGILITY CURVE AND ASSET TYPE

9 9 August 2015 Implications for future policy The prospect that risk may increase by an order of magnitude over the next fifty years if no further action is taken serves to underscore the importance of an ongoing commitment to flood risk reduction investment in coastal Louisiana. As shown by comparing CLARA s estimates of current risk to studies of previous incarnations of HSDRRS, New Orleans is in significantly better shape than it was 10 years ago with respect to flood risk. According to the best available scientific knowledge, USACE has succeeded in its goal of providing 100-year protection to the metropolitan area. That said, initial analysis in support of CPRA s 2017 Coastal Master Plan suggests that additional measures must be taken to maintain that level of protection over the long term. Given the expense of structural defenses, any commitment to a specified level of protection should include plans for maintaining that standard long into the future, in order to maximize total benefits. Now that the immediate plans for the reconstruction of New Orleans s defense system have been implemented, decisionmakers should consider whether a 100-year level of protection is sufficient for ensuring the city s sustainable and vibrant future. Several studies, using different methods and different assumptions, have concluded that greater protection standards are not only justifiable but optimal in New Orleans CPRA s 2012 Coastal Master Plan is a commendable step towards safeguarding New Orleans s future. The plan calls for land-building measures that would slow land loss, moderating the natural increase in flood risk associated with sea level rise and land subsidence. It also recommends several structural projects directly related to HSDRRS improvements. Though nonstructural measures are generally more cost effective in other parts of the coast with less structural protection, the plan does not restrict nonstructural funds from being used in New Orleans wherever they may be competitive. The 2012 Coastal Master Plan is expensive, estimated at a cost of $50 billion over the next 50 years. Construction has already begun on several high-priority projects, but projections of expected damage in 2065 emphasize the importance of following through with the current plan to ensure that funds are committed to its full execution. The 2017 plan will provide updated estimates of the effectiveness of risk reduction projects recommended in 2012 and give a better picture of what a sustained pledge might achieve. Through the master planning process, Louisiana has shown a laudable commitment to objective scientific analysis of its many options for risk reduction and coastal restoration. Hundreds of projects have been proposed over the past decades, and the state now has a framework in place to evaluate their costs and benefits using a comprehensive, systems-based approach. 42,43 This approach must be continually relied upon to alleviate the political gridlock which stymied past efforts, leading to the conditions which allowed Hurricane Katrina to wreak such havoc. To the extent that CPRA determines that projects are best funded elsewhere in the state, city planners and other local stakeholders should be prepared to identify alternative funding sources to maintain sufficient protection for New Orleans without implementing policies at the expense of other coastal communities. Questions of who should pay for flood protection continue to have implications on system readiness and future maintenance; residents of St. Bernard Parish recently voted down a property tax increase requested by the Lake Borgne Basin Levee District. 44 Conclusions The $14.5 billion spent over the last decade to upgrade New Orleans s protection system has required substantial state and federal commitments to the city and was the result of a vibrant local, state, and national discussion about the national economic value of New Orleans and its unique culture. Post-Katrina upgrades to the New Orleans flood protection system have successfully reduced risk by a substantial amount in a cost-effective manner, but without continuing maintenance and assessments of evolving future conditions, those gains may be swiftly lost. Analysis of the projects considered for inclusion in CPRA s 2017 Coastal Master Plan should demonstrate the value of long-term protection efforts, but it is also important to maintain the sense of urgency instilled by Hurricane Katrina to ensure that these plans come to fruition.

10 10 August 2015 Endnotes 1. Knabb, R.D., D.P. Brown, and J.R. Rhome, Tropical Cyclone Report: Hurricane Rita (18-26 September 2005), in Tropical Cyclone Reports, NOAA Administration, Editor 2006, National Hurricane Center: Miami, FL. 2. Knabb, R.D., J.R. Rhome, and D.P. Brown, Tropical Cyclone Report: Hurricane Katrina (23-30 August 2005), in Tropical Cyclone Reports, NOAA Administration, Editor 2011, National Hurricane Center: Miami, FL. 3. Knutson, T.R., et al., Tropical cyclones and climate change. Nature Geoscience, : p Hill, K.A. and G.M. Lackmann, The Impact of Future Climate Change on TC Intensity and Structure: A Downscaling Approach. Journal of Climate, (17): p Holder, K., FEMA Accredits Hurricane and Storm Damage Risk Reduction System (HSDRRS), 2014, US Army Corps of Engineers. 6. U.S. Army Corps of Engineers, Hurricane Isaac With and Without year HSDRRS Evaluation, U.S. Army Corps of Engineers, Editor 2013: Washington, D.C. p Act Number 8, Rogers, J.D., Development of the New Orleans Flood Protection System prior to Hurricane Katrina. Journal of Geotechnical and Geoenvironmental Engineering, (5): p Shallat, T., In the wake of Hurricane Betsy, in Transforming New Orleans and its environs: centuries of change, C.E. Colten, Editor 2000, University of Pittsburgh Press: Pittsburgh. p Much of the modern infrastructure protecting the city at the time of Katrina was spurred by a major flood in 1927, in which levees in Plaquemines Parish were intentionally destroyed to relieve upstream pressure on levees in the city. 11. Morton, R.A., et al., Historical Subsidence and Wetland Loss in the Mississippi Delta Plain. Gulf Coast Association of Geological Societies Transactions, : p Fischetti, M., Drowning New Orleans, in Scientific American Louisiana Coastal Protection and Restoration Authority, Louisiana s Comprehensive Master Plan for a Sustainable Coast, State of Louisiana, Editor 2012, State of Louisiana,: Baton Rouge. 14. Fischbach, J.R., Managing New Orleans Flood Risk in an Uncertain Future Using Non-Structural Risk Mitigation, 2010, Pardee RAND Graduate School: Santa Monica, CA. 15. Woolley, D. and L. Shabman, Decision-making chronology for the Lake Pontchartrain & Vicinity Hurricane Protection Project, U.S. Army Corps of Engineers, Editor 2008: New Orleans. 16. Some neighborhoods like the Lower 9th Ward are included in the St. Bernard subsystem, as defined here, even though they do not fall within St. Bernard Parish. 17. Interagency Performance Evaluation Taskforce, Performance Evaluation of the New Orleans and Southeast Louisiana Hurricane Protection System, U.S. Army Corps of Engineers, Editor 2006: New Orleans. 18. Shallat, T., Structures in the Stream: Water, Science, and the Rise of the U.S. Army Corps of Engineers 1994, Austin, TX: University of Texas Press. 19. Louisiana Coastal Protection and Restoration Authority, Integrated Ecosystem Restoration and Hurricane Protection: Louisiana s Comprehensive Master Plan for a Sustainable Coast, State of Louisiana, Editor 2007, Coastal Protection and Restoration Authority of Louisiana: Baton Rouge, LA. 20. U.S. Army Corps of Engineers, Louisiana Coastal Protection and Restoration Technical Report, U.S. Army Corps of Engineers, Editor 2009: New Orleans, LA. 21. National Academy of Sciences, Reducing Coastal Risks on the East and Gulf Coasts, T.N.A. Press, Editor 2014: Washington, DC. p Jacobsen, B., Hurricane Surge Hazard Analysis: The State of the Practice and Recent Applications for Southeast Louisiana, S.L.F.P.A.-. East, Editor 2013: Baton Rouge, LA. 23. Lopez, J.A., The Multiple Lines of Defense Strategy to Sustain Coastal Louisiana. Journal of Coastal Research, 2009(SI(54)): p The Interagency Performance Evaluation Taskforce (IPET) modeled flood risk under the configuration of HSDRRS that existed in 2005 when Hurricane Katrina struck, as a means of analyzing whether the levee failures and ensuing devastation were to be expected under the circumstances. The study later evaluated risk reduction under several interim versions of HSDRRS as the system existed in 2007 and as it was projected to exist in 2011 but the levee heights modeled under these configurations still varied significantly from what exists today in many areas. The Louisiana Coastal Protection and Restoration (LACPR) study performed by USACE examined the upgraded HSDRRS as it was envisioned in 2010; LACPR s flood risk model also assumed in all cases that levee systems could not breach or suffer other operational failures. If Hurricane Katrina has taught us anything, it is that levees can fail. 25. Johnson, D.R., Improved Methods for Estimating Flood Depth Exceedances within Hurricane Protection Systems, in Improving Flood Risk Estimates and Mitigation Policies in Coastal Louisiana under Deep Uncertainty, D.R. Johnson, Editor 2013, RAND Corporation: Santa Monica. p Fischbach, J.R., et al., 2017 Coastal Master Plan: Model Improvement Plan, Storm Surge and Risk Assessment Improvements (Subtask 4.9), CPRA, Editor 2015: Baton Rouge, LA. 27. Johnson, D.R., J.R. Fischbach, and D.S. Ortiz, Estimating Surge-Based Flood Risk with the Coastal Louisiana Risk Assessment Model. Journal of Coastal Research, 2013(Special Issue 67). 28. Resio, D.T., White Paper on Estimating Hurricane Inundation Probabilities, Interagency Performance Evaluation Taskforce, Editor 2007: New Orleans. 29. Toro, G.R., et al., Efficient joint-probability methods for hurricane surge frequency analysis. Ocean Engineering, (1): p U.S. Army Corps of Engineers, Final Post Authorization Change Report: Morganza to the Gulf of Mexico, Louisiana, 2013: New Orleans, LA.

11 11 August Federal Emergency Management Agency, Multi-Hazard Loss Estimation Methodology, Flood Model: Hazus-MH MR4 Technical Manual, 2009, Federal Emergency Management Agency: Washington, D.C. 32. CLARA does not model indirect spillover effects, such as disruptions in national GDP attributable to reductions in refining capacity, or non-monetary losses like casualties. Damage estimates include the replacement cost or actual cash value of damaged structures, along with the value of structural contents (and inventory, in the case of commercial or industrial establishments). The model also includes categories like damage to roads, vehicles, and agricultural crops, as well as losses directly linked to property damage and the reconstruction period, like lost rents, sales, and wages. The replacement cost of a structure is assessed as a function of its purpose: buildings are categorized into, for example, ten representative types of commercial structures, six types of industrial facilities, etc. Assumptions are made about each type s average square footage and replacement costs per square foot. The value of contents and inventory are assumed proportional to the structural value. Damage is assessed as a proportion of the structure s replacement cost and depends on the depth of flooding relative to the top of the building foundation. 33. See Table 3 of Johnson, Fischbach & Ortiz (2013) for the specific model parameter values associated with the Less Optimistic scenario. 34. Interagency Performance Evaluation Taskforce, Performance Evaluation of the New Orleans and Southeast Louisiana Hurricane Protection System, U.S. Army Corps of Engineers, Editor 2009: New Orleans. 35. In all figures, flooding external to HSDRRS is excluded from the maps for visual clarity. 36. The model projects an average annual growth rate of 0.67 percent (approximately the rate of growth experienced in the decade preceding Hurricane Katrina), and it assumes that the distribution of assets between urban and rural areas remains the same as in current conditions. 27 Structural assets and vehicles are assumed to grow directly in proportion with population growth, 26 while roads and agricultural crops are modeled as remaining constant. 37. The Other asset type includes public facilities, agricultural crops, roads, and vehicles. 38. Hallegatte, S., et al., Future flood losses in major coastal cities. Nature Climate Change, : p Dijkman, J., A Dutch Perspective on Coastal Louisiana Flood Risk Reduction and Landscape Stabilization, N.W. Partnership, Editor 2007, United States Army Corps of Engineers: Delft, The Netherlands. 40. Jonkman, S., et al., Risk-based design of flood defence systems: a preliminary analysis of the optimal protection level for the New Orleans metropolitan area. Journal of Flood Risk Management, : p von Winterfeldt, D., Using Risk and Decision Analysis to Protect New Orleans Against Future Hurricanes, in On Risk & Disaster: Lessons from Hurricane Katrina, R.J. Daniels, D.F. Kettl, and H. Kunreuther, Editors. 2006, University of Pennsylvania Press,: Philadelphia, PA. p Groves, D.G., C. Sharon, and D. Knopman, Planning Tool to Support Louisiana s Decisionmaking on Coastal Protection and Restoration, 2012, RAND Corporation: Santa Monica, CA. p Louisiana Coastal Protection and Restoration Authority, Louisiana s Comprehensive Master Plan for a Sustainable Coast, Appendix E - Decision Support Tools - Planning Tool, State of Louisiana, Editor 2012, Coastal Protection and Restoration Authority of Louisiana: Baton Rouge, LA. 44. Schleifstein, Mark. St. Bernard tax defeat means higher flood risk, flood insurance rates, levee leaders warn. New Orleans Times-Picayune, May 4, New Orleans, LA.

12 12 August 2015 Glossary ANNUAL EXCEEDANCE PROBABILITY (AEP) N-YEAR RETURN PERIOD The probability of a quantity, such as flood depths or flood damage, occurring or being exceeded in a given single-year period. A statistical measure equivalent to an annual exceedance probability, the n-year return period has a 100 percent AEP. For example, an event with a 100-year return period is an event with a 1 percent AEP. EXCEEDANCE The quantity in question, with reference to a given AEP or return period. For example, if five feet of flooding has a 0.2 percent AEP in a particular area, then one would say that the 500-year flood depth exceedance is 5 feet. EXPECTED ANNUAL DAMAGE A statistical measure summarizing the entire probability distribution of damage that could occur in a geographic area. It represents the average amount of damage an area might experience on an annual basis if conditions were to remain constant. N-YEAR FLOODPLAIN FRAGILITY CURVE The geographic region for which flooding is projected to occur at the n-year return period. The probability of a levee breach or other system failure, as a function of the rate of water overtopping the protection features. NOTE The U.S. Geological Survey (USGS) and other agencies have moved in recent years toward framing risk in terms of annual exceedance probabilities and no longer favor the use of 100-year return periods and other year-based terminology. This is largely due to misconceptions about what a 100-year return period actually means. It does not mean that a 100-year flood will occur regularly every 100 years, or only once in a 100-year period. A common interpretation is that over the span of a 30-year mortgage, there is a 26 percent chance of a home experiencing a 100-year flood. Return periods and AEPs can be used interchangeably, but bear in mind that they reflect assumptions about conditions at the time they are calculated; for example, hurricanes are projected to increase in intensity in the future, so the statistical results shown in 2065 are based on different assumptions about the relative likelihood of different storms than the results from 2015.

13 13 August 2015 Acknowledgments from the Authors This work has been funded by the Louisiana Coastal Protection and Restoration Authority under the 2017 Coastal Master Plan s Master Services Agreement. Coordination of the process has been managed by The Water Institute of the Gulf. In particular, the authors wish to thank collaborators Hugh Roberts and Zach Cobell at ARCADIS; Mandy Green, Melanie Saucier, Mark Leadon, and Karim Belhadjali at CPRA; Denise Reed at The Water Institute; and Gary Cecchine at RAND Corporation for their intimate involvement in and support of this project. We also wish to thank Lauren Mayer (RAND Corporation), Doug Bessette (Penn State University), Ann Carpenter (Federal Reserve of Atlanta), and Bob Collins (Dillard University) for their helpful comments in reviewing this paper. For more Information David R. Johnson RAND Corporation Santa Monica, CA djohnson@rand.org Jordan R. Fischbach RAND Corporation Pittsburgh, PA jordanf@rand.org Kenneth Kuhn RAND Corporation Santa Monica, CA kkuhn@rand.org

14 14 August 2015 About The Data Center The Data Center is the most trusted resource for data about greater New Orleans and Southeast Louisiana. Since 1997, The Data Center has been an objective partner in bringing reliable, thoroughly researched data to conversations about building a prosperous, inclusive, and sustainable region. About The New Orleans Index at Ten Collection The New Orleans Index at Ten collection includes contributions from The Data Center, the Brookings Institution, and more than a dozen local scholars. The aim of this collection is to advance discussion and action among residents and leaders in greater New Orleans and maximize opportunities provided by the 10-year anniversary of Katrina. The New Orleans Index at Ten: Measuring Greater New Orleans Progress toward Prosperity analyzes more than 30 indicators to track the region s progress on economic, inclusive, and sustainable growth. Essays contributed by leading local scholars and the Brookings Institution systematically document major post-katrina reforms, and hold up new policy opportunities. Together these reports provide New Orleanians with facts to form a common understanding of our progress and possible future. The New Orleans Index series, developed in collaboration with the Brookings Institution, and published since shortly after Katrina, has proven to be a widely used and cited publication. The Index s value as a regularly updated, one-stop shop of metrics made it the go-to resource for national and local media, decisionmakers across all levels of government, and leaders in the private and nonprofit sectors. Acknowledgments from The Data Center Many thanks go to Erin Kenzie for editorial assistance, and Southpaw Creative for design. The Data Center wishes to thank blue moon fund, Foundation for Louisiana, Greater New Orleans Foundation, JPMorgan Chase Foundation, Walton Family Foundation, and Zemurray Foundation for their generous support of The New Orleans Index at Ten. Additional gratitude goes to Baptist Community Ministries, GPOA Foundation, Entergy Corp., Institute for Mental Hygiene, Methodist Health Systems Foundation, RosaMary Foundation, Patrick F. Taylor Foundation, United Way of Southeast Louisiana for their support of the ongoing work of The Data Center. For More Information Allison Plyer Executive Director and Chief Demographer The Data Center allisonp@datacenterresearch.org Nihal Shrinath Data Manager and Research Associate The Data Center nihals@datacenterresearch.org Disclaimer The New Orleans Index at Ten collection represents studies and reports on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to The Data Center, its trustees, or its funders.