Free Motion to Dismiss - Rule 12(b)(1) - District Court of Federal Claims - federal

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 1 of 46

FINAl REPORT
THE DIRECT IMPACT OF

THE MRGO ON HURRICANE STORM SURGE
Preformed under:

Contract No. 2503-05-39
Hydrodynamic Modeling Effort for MRGO Study

Prepared for:

State of Louisiana
Department of Natural Resources

February 2006
URS Ale: 1922750.001

UR
225.922.5700
In Association with:

Prepared by:

URS Corporation

7389 Florida Blvd.. Suite 300 Baton Rouge, LA 70806

,\""j

WORtDn7NDS

Defendant's Exhibit D

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 2 of 46

FINAL REPORT

THE DIRECT IMPACT OF THE MISSISSIPPI RIVER GULF OUTLET

ON HUCANE STORM SURGE
Performed under:
CONTRACT NO. 2503-05-39

Hydrodynamic Modeling Effort for MRGO Study

Preparedfor:
State of Louisiana Natural Resources

Departent of

Februar 2006

UR
7389 Florida Blvd.

Suite 300 Baton Rouge, Louisiana 70806 (225) 922-5700

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 3 of 46

TABLE OF CONTENTS

Executive Summary ........................................,............................................................... ES-1
Introduction and Background ................................................................1-1

Section 1

1.
1.2

1.
1.4
1.5
Section 2

objectives and organization ............................................................... I-I The MRGO and Vicinity ................................................................... 1-2 Local Hurricane Storm Surge Threat.................................................1-3
Previous studies of Local Hurricane Storm Surge.............................

1-5

Related Issues ....................................................................................1-8

The 2003 Corps Study.....................................................................:........ 2-1
2.1

2.2
Section 3

2003 ADCIRC Grid ....................................,...................................... 2-1 Summary of2003 Corps Study.......................................................... 2-2

Factors for Further Study......................................................................... 3-1

Section 4

Modeling of 124-Knot Synthetic Storm ........................................................4-1
4.1

Simulation Models ............................................................................. 4-1
QA Check.................. .................. ................................ ....................... 4-1 Simulation Results ............................................................................. 4-2 Discussion of Results ..,.................. ... ........................... ...................... 4-3

4.2 4.3 4.4
Section 5

Modeling of Hurricane Betsy........................................................................... 5-1 .
5.1

Simulation Results ................................"........................................... 5-1

5.2
Section 6

Discussion of Results........ .......................... ...................... ................. 5-1

Modeling of Hurricane Katrina.................................................................... 6-1
6.1

6.2 6.3
Section 7

Hurricane Katrina .............................................................................. 6-1 Simulation Results ............................................................................. 6-1 Discussion of Results ................. .......................... .................. ............ 6-2

Modeling of Levee Alignment Sensitivity .......................................................... 7-1
7.1

Simulation Model.............................................................................. 7-1

7.2

Simulation Results .................".......................................................... 7-2

UR

J:\roject\LNR\ 1 9227506-MRGO Siudynienm RepoI'NTERlM.RPT,doc \1 3-EP-)5\TR i

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 4 of 46

TABLE OF CONTENTS
7.3
Section 8

Discussion of Results... ......... ........................................ ..................... 7-3

Wave Run-Up Analysis ............................................................................ 8-1
8.1

8.2
8.3
Section 9

Location for Initial Calculation ..................................... 8-1 Wave Generation and Attenuation Calculation ................................. 8-2 Run-Up Calculation ........................................................................... 8-3
Selection of

Conclusions .................................... ............... ..................,...... ........... ...... 9-1

Section 10

Recommendations ..... ......... ............ ................... ................ ......... ...... ....... 10-1

Section 11

References.. ......... ............ .......................... .................... ...... .......... ........... 11-1

UR

1;\PioJe\LDNRl1922750-MRGO SIIHly\lnlerim ReporNNTERIM-RPT2.docllJ.EP-Q5\TR II

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 5 of 46

TABLE OF CONTENTS
TABLES

Table I
Table 2
Table 3

Saffir-Simpson Scale for Tropical Cyclone Characteristics ..........................1-4 Overview of Storm Surge Models ................................................................. 1-6 Nine Synthetic Hurricanes ................................................ 2-2 Characteristics of

Table 4

Difference in Maximum WSE (ft) Baseline MRGO versus Bayou La
Loutre Barrier ................................................................................................ 2-3 Summary ofURS Simulations....................................................................... 3-2
QA Check Comparison of Maximum Surge WSE URS/WorldWinds

Table 5

Table 6

Table 7

Table 8

ADCIRC Run versus 2003 Study Using the l24-Knot-Fast Storm............... 4-2 Difference in Maximum WSE for Baseline, MRGO Closure, and Bayou La Loutre Barrier Scenarios, 124-Knot-Fast Storm....................................... 4-3 Difference in Maximum Surge WSE for Baseline and Closure Scenarios
Hurricane Betsy ............................................................................................. 5-1

Table 9

Table 10

Difference in Maximum Surge Water Surface Elevations for Hurricane Katrina Simulation ofMRGO Baseline and Closure..................................... 6-2 Difference in Maximum Surge Water Sudace Elevations for Hurricane Katrina Simulation ofMRGO Baseline and Modified Levees...................... 7-2

FIGURES

Figure I
Figure 2

Project Area

Corps Estimate of Regional Inundation from Category 4 and Above Storm Surge

Figure 3

L WR Estimate of Regional Inundation from Hurricane Georges (1998)
Prior to Track Turn

Figure 4 Figure 5 Figure 6 Figure 7

Full 2003 ACIRC Grid Detail of2003 ADCIRC Grid for MRGO and Surrounding Area 3D Depiction of 2003 ADCIRC Terrain for MRGO and Surrounding Area Comparison of Surveyed versus 2003 ADCIRC MRGO Channel Near Shell
Beach, Plan

Figure 8

Comparison of Surveyed versus 2003 ADCIRC MRGO Channel Near Shell
Beach, Cross Section

Figure 9

Figure 10

Figure II
Figure 12
Figure 13

Tracks for Hurricane Simulations 3D Depiction of 2003 ADCIRC Terrain with Closed MRGO 24-Knot-Fast Storm, Baseline MRGO Maximum WSE for I 24-Knot-Fast Storm, Closed MRGO MaximumWSE for I

Difference in Maximum WSE for 124-Knot-Fast Storm,
Closed MRGO

Baseline versus

UR

l;\ojei:ILNR\1 927506-MRGO siuiln!e~m ReporlNTERRM-RPT2.doc \ 1 3-EP-oS\BTR 111

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 6 of 46

TABLE OF CONTENTS
Figure 14

Figure 15

Figure 16 Figure 17

Storm Surge Stage Hydrographs, 124-Knot-Fast Storm, Baseline versus Closed MRGO Maximum WSE for Hurricane Betsy, Baseline MRGO Maximum WSE for Hurricane Betsy, Closed MRGO

Difference in Maximum WSE for Hurricane Betsy, Baseline versus Closed
MRGO

Figure 18

Storm Surge Stage Hydrographs, Hurricane Betsy, Baseline versus Closed
MRGO

Figure 19

Storm Surge Current Speed Hydrographs, Hurricane Betsy, Baseline versus Closed MRGO
Hurricane Katrina Simulation

Figure 20
Figure 21

Figure 22 Figure 23 Figure 24 Figure 25 Figure 26 Figure 27
Figure 28

Maximum WSE for Hurricane Katrina, Baseline MRGO Maximum WSE for Hurricane Katrina, Closed MRGO Difference in Maximum WSE for Hurricane Katrina, Baseline versus Closed MRGO Storm Surge Hydrographs, Hurricane Katrina, Baseline versus Closed MRGO Storm Surge Current Speed Hydrographs, Hurricane Katrina, Baseline versus Closed MRGO Baseline 2003 ADCIRC Levees Modified Levees Maximum WSE for Hurricane Katrina, Modified Levees and Closed MRGO

Figure 29 Figure 30
Figure 31

Difference in Maximum WSE for Hurricane Katrina, Baseline versus
Modified Levees Storm Surge Hydrographs, Hurricane Katrina, Baseline versus Modified Levees Wave Set-up and Run-up Schematic of HPL Cross Section near Bayou Dupre, Station 673
Lake Borgne to HPL Levee § Bayou Dupre Transect for Wave Run-up

Figure 32
Figure 33

Analysis
Figure 34

Schematic of Wave Generation and Attenuation for Lake Borgne to HPL
Levee § Bayou Dupree

Figure 35

3D Depiction of Higher Resolution ADCIRC Terrain for MRGO and
Surrounding Area with Closed MRGO

ATTACHMENTS

Attchment I

Numerical Modeling of Storm Surge Effect ofMRGO Closure

UR

l;\oJect\LNR\1927506MRGO Sludfl'teem ReporNTERM-RP2.0$c \1 J-EP-05\TR iv

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 7 of 46

Exec_lVe Summa..
In March 2005 URS Corporation (URS) was tasked by the Louisiana Departent of

Natural

Resources (LDNR) to evaluate the impact of the Mississippi River Gulf Outlet (MRGO) on regional hurricane storm surge by examining the immediate and direct effects using a hydrodynamic model of certain selected storms. In October 2005, LDNR requested that Hurricane Katrina in this evaluation. The URS task follows a 2003 URS include modeling of US Ary Corps of Engineers (Corps) report on the effect of blocking the MRGO at Bayou La Loutre on hurricane storm surge.
The URS project team reviewed the 2003 Study results and identified seven additional factors for further study:
1.

The impact of complete closure (i.e., fillng in) ofMRGO;
The effect on surge across the entire study area; The influence on surge scour velocity; The impact on storm surge arrival and draining;
The impact of a severe storm;

2. 3. 4.
5. 6. 7.

The sensitivity of storm surge to levee alignment; and

The effect ofthe MRGO on levee wave run-up.

URS conducted a total of seven simulations using three hurricanes-a 124-Knot-Fast Synthetic Storm, Hurricane Betsy, and Hurricane Katrina-including comparisons ofMRGO Baseline versus Closure Scenarios. The simulations were conducted using the ADCIRC hydrodynamic model and the 2003 grid. MRGO closure was represented by filling in the channel to an elevation equivalent to approximately I foot above mean sea leveL. A levee alignment sensitivity simulation was conducted using the .levees along the south bank of the

GIWW and the MRGO, and the intermediate levee at the 40 Arpent CanaL. URS also
provided a wave run-up analysis.

URS understands that public concern for the MRGO's immediate and direct contribution to the MRGO's long-term impacts on storm surge is part of a broad interest in the full range of
the regional communities and environment. Especially given the effects of Hurricane Katrina, there is a justified interest in all measures that might protect human life and aid in restoring the economic, cultural, "and ecological resources of the area for generations to
come. This phase of work, however, only provides the findings and recommendations

UR

1;\Pajjct\LDNR\19227506-RGO $tudynel MRGO Repolnel Repoitdoc\9-FEB-\BR E8-1

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 8 of 46

Execute Summa",
associated with modeling the surge impact of the MRGO for hurricanes under present
conditions.
Major conclusions of

this study are:

.

The MRGO channel does not contribute significantly to peak surge during severe storms, when the conveyance of surge is dominated by flow across the lakes and marsh. Nor does the channel contribute entire surface of the coastal
significantly to wave run-up.

. .

Complete fillng of the MRGO-or blockage or partial filling-will not
provide significant immediate, direct mitigation of severe storm surge.

For a few locations outside the Hurricane Protection System (HPS) closure of the MRGO may reduce the peak surge for certin fast, low-to-moderate storms, when the surge is not dominated by flow across the open lakes and marsh, and may modestly delay the onset of surge.

. . .

For some storms and locations MRGO closure would slightly increase storm surge peaks and impair draining of storm surge following the storm passage.

MRGO closure would significantly reduce surge scour velocities at some
channel

locations, which is important to soft swamp and marsh organic soils.

Natural and man-made landform alignents (passes, ridges, levees, etc.) can create surge concentration under certain storm conditions. The effect of the "funnel" formed by levees along the GIWW and MRGO in concentrating surge was evident in Hurricane Betsy but not in Hurricane Katrina. However, the MRGO did not significantly impact the surge at the "funnel" for closure of either storm. Widening the funnel in the sensitivity simulation actually
resulted in an increase in surge at the IHNC.

The above findings on the role of the MRGO on storm surge imply that the surge conveyance of the MRGO is not an importnt factor in establishing near-term HPS requirements. Nearterm HPS requirements should be based on a thorough analysis of surge height recurrence

frequency-and those factors that can reasonably be expected to effect total surge heightsand the costs and benefits of alternative degrees of protection.

UR

l:\olectIlNR\19227506RGO SiudynBI MRGO RepolnBI RepDrldoc\9-FEB-R ES- 2

,

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 9 of 46

Execute Summarv
URS recommends that LDNR conduct further evaluations to better understand the long-term various closure scenarios of role of the MRGO on storm surges and the future implications of
the MRGO, including:

.

Develop an improved, high resolution ADCIRC grid of the MRGO and
surrounding area, with accurate representation of the channel, and regional topography and bathymetry corrected to updated NA VD-88 benchmarks.

. .

Conduct calibration studies using the improved grid for a range of tidal and
storm events.

Perform surge simulations using the improved grid to better resolve locations of impact, and degrees of impact (positive and negative), for various MRGO baseline and closure scenarios.

.

Evaluate the effect of various MRGO closure alternatives, subsidence,
erosion, and sea level rise, and restoration measures-such as controlling saltwater intrusion and introducing freshwater from the Mississippi River-Qn landscape. the long-term regional

.

Develop ADCIRC grids to represent long-term landscape scenarios and use them to study the future impact of natural processes, MRGO alternatives, regional wetland restoration alternatives, and other landscape changes on hurricane storm surge.

UR

1:\roJec:\LDNR\1921506-MRGO Stdynii MRGO Rep!IinBI RepDrdoc\9FEB-\8R ES- 3

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 10 of 46

SECTIONONE
In March 2005 URS Corporation (URS) was tasked by the Louisiana Department of

Introduction and Background
Natural

Resources (LDNR) to evaluate the impact of the Mississippi River Gulf Outlet (MRGO) on regional hurricane storm surge by examining the immediate and direct effects using a hydrodynamic model of certain selected storms. In October 2005, LDNR requested that

URS include modeling of Hurricane Katrina in this evaluation. This study was always intended as a preliminary evaluation of this specific issue, which-as this Introduction and Background show-is part of a broader, more fundamental, suite of concerns requiring careful examination. Nevertheless, a correct understanding of the MRGO's role in storm
surge conveyance is one importnt key to establishing a sound scientific and engineering approach to protecting and restoring St. Bernard Parish and eastbank Orleans Parish.
1.1

OBJECTIVES AND ORGANIZATION

Due to the extensive media coverage and public attention directed at the impact of the MRGO on storm surge, it is important to understand the objectives and limited scope of this phase of work. The URS task is a follow-on to earlier modeling performed the US Army

Corps of Engineers (Corps). In 2003 the Corps-as part of an assessment of options to
reduce salinity intrusion caused by the MRGO-valuated the impact of a saltwater barrier at Bayou LaLoutre on mitigating hurricane storm surge. The 2003 Study concluded that "the MRGO has minimal influence upon storm surge propagation." (A copy ofthe 2003 Study is included as Attachment I and is discussed in Section 2).
Since the 2003 Study only addressed blocking of the MRGO at Bayou La Loutre, the LDNR asked URS to evaluate the surge mitigation effects of totally filling the channeL. The URS task examines this question with the following limitations:

.

The URS simulations examine the impact of closing the MRGO channelthat portion which extends southeast from the Gulf Intracoastal Waterway
(GIWW) to Breton Sound. The simulations do not look at possible impacts of filling or blocking the GIWW (e.g., at Paris Road).

.

As an economical first effort the task was limited to the use of three
"diagnostic" storms (one synthetic storm, Hurricane Betsy, and Hurricane

Katrina), rather than a comprehensive suite of storms with widely varying intensities, sizes, tracks, and forward speeds. URS would recommend further
simulations warranted by the results of this phase of work.

UR

t\ProJei;\LDNRII927506MRGO Sttdynterim ReporNNTERJM-RP2.docI1J.ËP-05UJTR 1- i

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 11 of 46

SECTION

ONE
.

Introduclion and Background

The past and future effects of the MRGO on regional landforms (e.g., wetlands) and the impact of regional landform changes on storm surge are
beyond the scope of this modeling task. This task examines only the

immediate and direct impacts of MRGO closure on hurricane storm surge
under current conditions.
The remainder of Section I provides background information for this report. The following

topics are addressed in the ensuing sections:
Section 2. Section 3. Section 4. Section 5. Section 6. Section 7. Section 8. Section 9.
Section 10.

The 2003 Corps Study

Factors for Further Study Modeling of I 24-Knot Synthetic Storm Modeling of Hurricane Betsy Modeling of Hurricane Katrina Modeling of Levee Alignment Sensitivity Wave Run-Up Analysis Conclusions Recommendations

1.2

THE MRGO AND VICINITY

The MRGO is a 76 mile man-made navigation channel bisecting the wetlands of St. Bernard Parish and connecting the GIWW in eastern New Orleans to Breton Sound and the Gulf of

Mexico (see Figure I). Construction of the 500-foot wide 36-foot deep MRGO was
completed in 1968 by the Corps at a cost of $92 million. The MRGO facilitated the

development of port facilities and expansion of commerce along the GIWW and Inner
Harbor Navigation Canal (IHNC). The Corps has routinely performed dredging (typically to
depths of 40 feet) to maintain the channeL.

The region surrounding the MRGO is dominated by low-lying coastal wetlands, including
cypress swamp, fresh-intermediate marsh, brackish marsh, and saline marsh. These wetlands

are tidally connected to the adjacent coastal waterbodies-Lake Borgne and Breton Sound. Aside from minor remnant ridges, these wetlands are typically less than 2 feet above the
local mean sea level (MSL) of the coastal

lakes and bays.

The New Orleans urban area developed into eastern Orleans Parish and into St. Bernard

Parish along the higher natural Mississippi River levee and swamp ridges, which commonly

UR

1:\Pro)ect\LRII92750-MRGO $bJdynlenm ReporlNTERJM-RPP,doc \1~EP-0TR 1-2

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 12 of 46

SECTIONONE

Introduction and Background

lie several feet above MSL. Development through the 20th Century expanded into adjacent swamps, which were drained by an extensive network of canals and pump stations.

Several small fishing vilages, including Reggio, Yscloskey, Shell Beach, Hopedale, and
Delacroix in St. Bernard Parish, arose near the coast. These communities, along with the

roadways leading to them, were established on low natural ridges of abandoned deltaic
distributaries at elevations a few feet above MSL.

1.3

LOCAL HURRICANE STORM SURGE THREAT

The more heavily developed portions of

the New Orleans area are surrounded by a Hurricane

Protections System (HS) of levees and floodwalls. Major upgrades to the regional HPS were begun by the Corps in the 1960s in the wake of Hurricane Betsy. The HPS was
typically designed to handle the surge from a "standard project hurricane" (equivalent to a fast-moving Category 3 hurricane). The HPS and floodgates are largely owned, operated, and maintained by local Levee Districts (e.g., Lake Borgne Levee District in St. Bernard
Parish).

The coastal fishing villages lie outside the HPS. Given their low elevation, these
communities are susceptible to significant damage from even minor tropical storm surge events. Moreover, evacuation in the face of a major hurricane is severely hampered by the
advancement of high water and inundation of

routes to safety.

Regional storm surge risks are magnified by the subsidence of the drained swamplands

within the levee system. Such areas have subsided many feet below MSL, resulting in topography that resembles a large "bowl". Gradual post-construction settlement of the levees also exacerbates inundation risk. Communities and key evacuation roads outside of the levee system are also subsiding. Ironically, it is only those healthy coastal wetlands
sustained by natural or simulated deltaic nourishment which maintain their elevation
(primarily through the accretion of natural detritus).
Table I summarizes

the basic characteristics of tropical cyclones based on the Saffir-

Simpson Scale:

UR

I,IP,oJecl\LRI1927S06-MRGO Sttclnterim Repr1NTERIM-RPT2,doc \1~EP-0\STR i - 3

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 13 of 46

SECTION

ONE
Table 1

Introduction and Background

Saffr-Simpson' Scale for Tropical Cyclone Characteristics

Tropical Depression Tropical Storm

"=39

39-73

t
2 3 4 5

74-95

" 980

3 to 5 6 to 8

96-110
111-130
131-155
greater than 155

979-965 964-945

9to 12

944-920
"=920

13 to18

19+

The probability that any hurricane will pass within 75 miles of New Orleans in any given
year is about 12.5 percent, or about once every 8 years. The odds of a major hurricane

(Category 3 and above) passing within 75 miles are about 3.2 percent per year, or about once every 30 years (see Sheets, Hurricane Watch, Appendix D, 2001).

Following close calls with Hurricanes Andrew (1994), Georges (1998), Isadore (2003), and Lily (2003) there was increasing research attention, public awareness, and governmental

concern for the threat to life, public health, and the regional economy posed by severe
hurricane storm surge in southeast Louisiana. For example, the LSU Center for the Study of Public Health Impact of Hurricanes began a five-year study, Assessment and Remediation of Public Health Impacts Due to Hurricanes and Major Flooding Events. The ability of levees
to withstand storm surge, problems of evacuation, and the catastrophic consequences of

inundating the "bowl," became front page news and a priority for federal, state, and local emergency planning and response offcials. In February 2001, in response to post-Hurricane Georges assessments, the Corps drafted an"unwatering" plan for New Orleans. In June 2002
the New Orleans Times-Picayune ran a multi-day feature: Special Report: Washing Away.
In July 2004 the Louisiana Offce of Emergency Preparedness in conjunction with the

Federal Emergency Management Agency (FEMA) conducted a planning exercise, which featured widespread inundation of the New Orleans area from a simulated Hurricane Pam, a
slow-moving, large Category 3 hurricane.. In October 2004 Category 4 Hurricane Ivan

threatened southeast Louisiana, precipitating a full-scale evacuation of hundreds of
thousands of people.

Hurricane Katrina, a large storm with Category 5 winds up until a few hours before landfall, struck southeast Louisiana at daWn on August 29, 2005. The hurricane center passed due north through the eastern flank of St. Bernard Parish. In the ensuing, hours, days and weeks

UR

1:\PlojectILDNR\19275O-MRGOStudynterim ReporlNTERlM.RPT,doc\1::EP-oS\TR 1-4

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 14 of 46

SECTION

ONE

Introduction and Background

the storm's impacts far exceeded the preparations of all governments, private organizations, and individuals. As of this date the total storm related fatalities in Louisiana have reached
1,100-128 in St. Bernard Parish and 286 in the Lower 9th Ward. The number of

area homes

that may ultimately have to be demolished may reach the tens ofthousands.
1.4
PREVIOUS STUDIES OF LOCAL HURRICANE STORM SURGE

Following Hurricane Betsy in 1965, the Corps calculated storm surge heights using a
standard project hurricane-without the advantages of to

day's complex computer models. In

the 1970 Flood Insurance Study for the Louisiana Gulf Coast, the Corps estimated the 10,
50, 100, 200, and 500 years storm surge elevations in St. Bernard Parish south of Lake Borgne at 7.0, 11., 12.2, 13.0, and 13.7 feet MSL. The Corps designed floodwalls and

levees in the MRGO-GIWW area to elevations of roughly 14 to 17.5 feet above NGVD-29. Pre-Katrina surveys have shown that some portions of the HPS were more than two feet below design grade (information provided by Lake Borgne Levee District)."
Beginning in the 1980s the FEMA and the National Oceanic and Atmospheric

Administration (NOAA) began to undertake more sophisticated numerical computer
modeling of storm surge (see Table 2). Throughout the 1990s models became more readily used by planning agencies, such as the Southeast Louisiana Hurricane Preparedness Study prepared by the Corps in 1994. Figure 2 shows Corps estimates of peak inundation for a Category 4 storm surge for southeastern Louisiana developed with the SLOSH modeL. ¡The inundation depths shown would not be produced from a single storm of a given track and
forward speed, but are the depths each location could experience from a severe surge

scenario for that area.)

- Today most fonnal references to elevation are given in the North American Vertical Datum of 1988, (NA VD-

88). In the New Orleans area references to the National Geodetic Vertical Datum of 1929 (NGVD-29) and
NA VD-88 are nearly equal. Elevations in either the NGVD-29 or NA VD-88 are not equivalent to local MSL.

An elevation of 0 feet in NGVD-29 or NA VD-88 is at about -1 foot local MSL. Thus, a reference to + i foot NGVD-29 or NA VD-88 is roughly equal to local MSL. A levee height which is accurately surveyed to 15 feet NA VD-88 is about 14 feet above local MSL. Accurate elevation detenninations in the New Orleans area must also be based on benchmarks that are valid at the time of survey.

UR

L:\ProJec:\LDNR\1927506MRGO SbJdýnlerim ReporNTERJM-RPT,doc\13-SE?-o5\TR 1-5

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 15 of 46

SECTIONONE
Table 2

InlrodUClion and Background

Overview or Storm Surge Models
Several different modeling techniques have been used to model storm surge in the coastal waters surrounding MRGO. These models are briefly reviewed.

SLOSH Model - A numerical-dynamic, tropical storm surge model, Sea, Lake, and Overland Surge from Hurricanes (SLOSH), was developed by Jelesnianski at the National Weather Service (NOAA) for real-time forecasting of hurricane storm surges on continental shelves, across inland water bodies, along coastlines and for inland routing of water. Overtopping of barriers such as levees, dunes, spoil banks, etc. is permitted. Also,
channel flow and flow through barier cuts are entertined. The model is two-dimensional, covering water

bodies and inundated terrain and is applied on a polar coordinate system. The SLOSH model does not address
wave set-up or wave run-up.

The SLOSH model is run to simulate the flooding caused by hurricanes. The model is designed for operational forecasting, and the model's input parameters that describe the hurricane are relatively simple and predictable.
The hurricane's position, size and intensity all enter as input to the modeL. Verification runs of the SLOSH model indicate that the accuracy of storm surges prediction is +/- 20 'Y.

FEMA Model - An overland flooding model has been developed by the Federal Emergency Management Agency (FEMA) to predict hurricane flood elevations for the National Flood Insurance Program. The model uses an explicit, two dimensional, staggered finite difference scheme to simulate the flow of water caused by
tides and wind systems. The inputs to the model include the bathymetr, coastline configuration, boundary

conditions, and bottom friction and other flow resistance coeffcients. Also required are the surface wind
velocity and atmospheric pressure distributions of the hurricane. The model predicts water level elevation and water transport everyhere in the modeled region. The model uses a rectagular grd to discretize the simulated region or the ocean and coast. The model grid expands during a simulation to predict the flooding of
low lying areas. Barriers and rivers which occur in the coastal zone have a controllng influence on flood
levels. Bariers can include roadways, levees and natural features such as cheniers. Rivers include channels,

canals and inlets. These features are typically much smaller in width than a typical grid cell, having widths that are about i 00 to 1000 feet. These features can be included in the computations as sub-grid scale elements. The FEMA model also does not address wave set-up or wave run-up.

ADCIRC Model - The ADvanced CIRCulation model was developed by Westerink and Luettich as a twodimensional depth integrated finite element hydrodynamic circulation code for ocean shelves, coasts, and estuaries. The finite element approach allows for modeling very large domains, with a flexible mesh that can provide coarse elements in open water far from areas of interest and fine resolution along areas of interest. Finite element also allows more accurate representation of interior features and model boundaries. The model incorporates recent developments in effcient finite element solution schemes and the code has been parallelized to run on commodity computer clusters. It includes wetting and drying algorithms and can represent hydraulic features such as levees, weirs, and culvert. The ADCIRC model does not address wave set-up or wave run-up.

UR

l;\oJect\LNR\1927506-MRGO SbJilnterlm RepolNTERIM-RPT.doc IIJ-EP-0TR i-6

c

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 16 of 46

SECTIONONE

Introduction and Background

Storm surge models were also used by university researchers. Figure 3 illustrates surge

inundation estimates, prepared by Dr. Joseph Suhayda using the FEMA model, of the
potential impact that the 1998 Hurricane Georges might have had on the New Orleans area had the storm not veered eastward (Louisiana Water Resources Research Institute, Louisiana State University).

These analyses fueled concern over hurricane storm surge risks and the need for further studies into the adequacy of levee protection. In 2002, the Corps completed a Hurricane Protection Reconnaissance Study to examine surge protection needs of the entire New Orleans area. The Corps recommended conducting an $8.6 million Feasibilty Study to examine alternatives for upgrading protection-including the option of providing Category 5
protection throughout New Orleans. As of August 29, 2005 the Feasibility Study was
awaiting funding. Following Hurricane Katrina, the Corps was authorized by Congress to

prepare a Category 5 hurricane protection technical report for south Louisiana.

As part of their ongoing pre-Katrina levee assessment, the Corps funded development of a sophisticated computer storm surge model using ADCIRC (Westerink and Luettich). The finite element model provides several advances over the traditional FEMA and SLOSH storm surge models of southeast Louisiana. By facilitating large domains the model can more accurately represent hurricane storm surge in the Gulf of Mexico. At the same time, the model allows for high resolution of critical south Louisiana features, such as the HPS and

the MRGO. Figure 4 ilustrates the ADCIRC Grid developed for the Corps. The grid
includes 600,331 elements and 314,442 nodes, with node spacing ranging from 15.5 miles in the mid-Atlantic to 330 feet in the New Orleans area. Simulations using this grid must be

done in I to 2 second time-steps-with one day of simulation requiring tens of bilions of
node-steps. Multi-day hurricane simulations can be completed in a matter of hours on super

computers utilizing parallel clusters of commodity processors (such as those typically
operated by the Corps and universities).
In 2004 researchers at the LSU Hurricane Center (Kemp, Mashriqui, and' van Heerden, in

conjunction with Westerink) utilized the parallel version of ADCIRC and the 2003 Corps Grid to simulate a very large, slow-moving Category 3 storm (referred to as Hurricane Pam) on the LSU supercomputer "Super Mike" for a multi-agency emergency planning exercise. The synthetic storm portrayed the catastrophic inundation of the New Orleans area resulting from massive storm surge (see http://hurricane.lsu.edu/).

UR

l:\roJecl\LR\1927506MRGO SIudyn\enm Repor1NTERJM-RPP2.doc\1::EP-0TR 1-7

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 17 of 46

SECTIONONE

Introduclion and Background

In addition to raising levees, the Corps, Levee Districts, and others began discussing

potential alternatives for mitigating hurricane storm surge by restricting the penetration of
surge at key locations. Such locations included the GIWW at Paris Road, which transmits

surge westward into the IHNC, and the Rigolets and Chef Menteur Passes, through which surge enters Lake Pontchartrain.
In the wake of Hurricane Georges and the Hurricane Pam exercise, one conveyance feature which received significant attention, particularly by residents and offcials of St. Bernard and

Orleans Parishes, was the MRGO. The conventional opinion has been that the MRGO
facil itates the transmittal of storm surge from Breton Sound into St. Bernard Parish and upward into the GIWW and IHNC, placing these areas under an increased threat.

1.5

RELATED ISSUES

By way of background, there are the several environmental and economic issues which are related to concerns over the role of the MRGOin storm surge conveyance. While these this report, they are important to acknowledge. issues are not the subject of

.

The erosion of the MRGO banks. Bank erosion is estimated at up to 15
feet/year, widening the channel to as much as twice its original design in some places, and over time significantly increasing conveyance. Erosion is primarily due to ship wave action (Britsch and Ratcliff, 200 I). The Corps and

the LDNR are investigating methods to improve bank stabilization using articulated concrete mattresses, rock dikes, and other armoring techniques
(e.g., CWPPRA Project PO-32).

.

The increased salinitv in the upper estuaries. Increased salinity is occurring in part due to the introduction of saltwater via the MRGO channel, causing rapid

disappearance of surrounding fresh and brackish coastal wetlands. Some
locations have seen a 3 to 4 fold increase in salinity during the decades since the MRGO was opened. An estimated 11,000 acres of fresh/intermediate
marsh and cypress swamps have converted to brackish marsh and 19,000

acres of brackish marsh have converted to saline marsh. Land losses include fresh/intermediate marsh (3,400 acres), brackish marsh (10,300 acres), saline marsh (4,200 acres), and freshwater swamps (1,500 acres). The Coast 2050 Plan and the Louisiana Coastal Area (LCA) Feasibility Study acknowledged

the role of the MRGO in contributing to major coastal degradation and

UR

1;\?roJect\LDNR\19227506MRGO SbJdyntem ReporNNTE'RIM-RP2.doc\1:lEP-QS\TR 1-8

,

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 18 of 46

SECTIONONE

Introduction and Background

targeted reversing associated wetland loss in St. Bernard Parish as a critical

objective. Local offcials have asked if the MRGO's role in damaging
regional wetlands increases the long-term threat of hurricane storm surge.

Methods to control salinities in the MRGO, such as the installation of a
saltwater barrier (sil or gate) near Bayou La Loutre are being assessed by the Corps in a MRGO Re-Evaluation Study and by the Corps and LDNR in a MRGO Ecosystem Restoration Study.

.

The continued economic viability of a deep channel in the face of light cargo

traffic. An average of five cargo vessels per day utilize the MRGO (R.
Caffey, 2002). The Corps has been evaluating long-term alternatives

including modifying the MRGO to a shallow barge channel (MRGO
Reevaluation Study, in progress as of August 2005).

.

Regional rates of subsidence. sea level rise. and changing ocean climate.

Within St. Bernard Parish, benchmarks along the natural levee of the
Mississippi River are subsiding at a rate of 3 feet/century. Benchmarks along Paris Road near the GIWW (across former wetlands) are subsiding at 6

feet/century (Shinkle and Dokka, 2004). Natural geologic processes
contribute significantly to subsidence in the St. Bernard Parish area. In some locations-such as drained swamps-natural subsidence rates are exacerbated

by human intervention. Gulf Coast eustatic sea levels are estimated to be
rising at a rate of about I foot per century (R. Twilley, 2001). Moreover,
hurricane researchers theorize that the Atlantic Basin is experiencing

increasing frequency and intensity of hurricanes. Regional vulnerability to
storm surge may therefore be increasing.

UR

1:\leClILDNR\192750-MRGO S!udyn!e~m ReporlNTERIM-RPT_doc\1J-EP-05ITR i-9

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 19 of 46

SECTIONTWO

The 2003 Corps Studv

Concern over the contribution of MRGO conveyance to local hurricane storm surge inundation led the Corps in 2003 to expand their study of a Bayou La Loutre Saltwater
Barrier concept to include an evaluation of whether such a Barrier might mitigate storm
surge. The Corps conducted this assessment using the ADCIRC storm surge model (a copy

of the 2003 Study is included as Attchment I).
2.1

2003 ADCIRC GRID

The 2003 Grid was prepared by Dr. Joannes Westerink of Notre Dame for the Corps (also
referred to by the Corps as Grid S08). Figure 5 shows the density of ADCIRC grid nodes for the MRGO and surrounding area. In addition, the figure ilustrates the location of weirs used to represent elevated levees and roads incorporated into the modeL. Figure 6 depicts the the area, including the MRGO and levees, in a 3D oblique view. ADCIRC model terrain of
The terrain model appears to be a reasonable coarse representation of regional topography

and bathymetry. URS did not perform a detailed check of terrain values during this phase of work. We understand that several sources of topographic and bathymetric data were utilized and that accurate reconciliation of datums was not always possible. Certin features (e.g. portions of levees) are known to be off by a couple of feet. All ADCIRC simulations were conducted with the starting stil water surface at model elevation O.O-which we refer to as MSL. The surge results are not readily convertible to NGVD-29 or NA VD-88 due to the variety of datums reflected in the grid topographic and bathymetric elevations.
It is interesting to note that the ADCIRC grid used in the 2003 Study represented the MRGO channel cross-sectional area at roughly twice the Corps' most recent surveyed cross-sectional
area. Figures 7 and 8 present plan and cross-section comparisons of recent surveys versus
ADCIRC representations of the MRGO near Shell Beach. This comparison is typical of

the

entire surveyed and modeled channeL. The larger representation of the MRGO in the 2003 grid results from using a number of minimum-sized elements in representing the channel
width, which in needed to control numerical stability. Thus, the 2003 Study ADCIRC grid

signifcantly over-represents the conveyance of the MRGO. As shown in Figure 5, the 2003

Grid also has the alignment of MRGO slightly off, to the south, but the discrepancy in
overall channel

length and orientation with the coast is very minor.

The Bayou La Loutre Barrier was represented in model simulations by raising several grid nodes in the MRGO where it bisects the Bayou La Loutre ridge-in effect restoring the ridge the grid was not altered for the Barrier Scenario. across the MRGO. The remainder of

UR

1:\Pro¡e\LDNR\I927506-MRGOSiudynierimRepprtNTERIM-RPT.doc\IJ-EP-05ITR 2-1

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 20 of 46

SECTIONTWO
2.2
SUMMARY OF 2003 CORPS STUDY

The 2003 Corps Studii

In order to evaluate potential storm surge effects of the Barrier, the 2003 Study utilized nine

synthetic storms (see Table 3) and one historic storm, Hurricane Betsy (1965). The Betsy simulation represented the storm with top winds of 135 knots and a forward speed of about 20 knots at landfall, and included tides. The storm tracks are shown on Figure 9. All ten storms were simulated for the Baseline Scenario (without-Barrier) versus with-Barrier
Scenario.
Table 3

Characteristics of Nine Synthetic Hurricanes

."', To"Wi~~sd:

'.':i,'#~~~ràspt~iI¡I'\c,: " . .',.'. . . .' '. CelltråiP~.,siire (iib)Y
Fast, 20 Knots

.",.',',.",
934 934 934
955 955 955 989 989 989

124-Knot
I

24-Knot 24-Knot

Medium, 15 Knots
Slow, 5 Knots

I

100-Knot
100- Knot

Fast, 20 Knots
Medium, 15 Knots
Slow,
5

100-Knot

Knots

65-Knot 65-Knot 65-Knot

Fast, 20 Knots

Medium, 15 Knots
Slow, 5 Knots

I knot equals 1.55 miles per hour.

Hydrographs for the two scenarios (Baseline versus Barrier) were presented in the 2003
study for four locations:

. . .

IHNC

GIWW at Paris Road
Bayou Dupre

.
Comparison of

Shell Beach

the maximum water surface elevation (WSE) in MSL at each location for four graphs are of the nine synthetic storms plus Betsy is re-produced as Table 4. (The hydro

included in the copy of the report in Attchment I). Table 4 shows that there were two
reductions in maximum storm surge over 0.5 feet. These reductions represented about a 27
percent reduction of Baseline peak surge (in the case of the 65-Knot-Fast storm at Bayou

UR

I,IProjecl\LRII92750-MRGO Sludynterlm ReporNNTERJM.RPT.doc \13-EP-oTR 2-2

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 21 of 46

SECTIONTWO

The 2003 Corps Studv

Dupre) and 16 percent reduction (for the 124-Knot-Fast storm at Shell Beach). Reductions in the range of 0.3 to 0.5 feet were also seen at Shell Beach for the Betsy simulation and for the other three locations for the 65-Knot-Fast storm. The reductions in peak surge for the these the maximum surge less than 0.3 feet. In six of other 16 data points on Table 4 were all
was slightly increased by the Barrier.
Table 4

Difference in Maximum WSE (ft) Baseline MRGO versus Bayou La Loutre Barrier
, ,

Hu¡'ncåne
,', ,

"Range'
'1ft MSl,
"'5

Surge',

,, " i,

,',

IUC
"

"", GIWW ~
-0.19
-0.13 -0.33

'".,MRGO~
-0.16

,', ','

MRGO ~
Shell'Beach
-0.53
-0.26

Parislload " Baypu Dupre

124-Knot-Fast
124-Knot-Slow

-0.16
-0.11

"'10
"'3 "'4

-0.14

65-Knot-Fast
65-Knot-Slow
Betsy

-0.3
0.03

-0.54
0.02 0.02

-0.37
0.02
-0.3

0.02
-0.01

"'12

0.03

A positive value is an increase in surge associated with the Ban-ier and a negative value is a decrease in surge.

the hydrographs for the 124-Knot-Fast and Betsy storms for Shell Beach and Bayou Dupre reveal that the Barrier Scenario had a noticeable negative impact by impeding
Examination of

draining of storm surge.

The 2003 Study showed that, at most, blocking the MRGO would only slightly reduced peak surge, with the most reduction occurring for a fast storm. For the higher surge simulations, conveyance across the entire marsh appears to dominate the propagation of surge. The 2003 Study concluded that "the MRGO has minimal influence upon storm surge propagation."
This inference is buttressed by the very conservative representation of the channel cross

section in the 2003 model grid. Representing the MRGO channel in the 2003 grid at almost

twice the cross-sectional area causes the model to over-estimate the role of MRGO
conveyance. Thus, the reductions noted in Table 4 with the Bayou La Loutre Barrier simulations are probably over-predicted.

UR

1;\PoJect\LNR\19225D6-MRGO Siudynterlm ReporNNTERIM-RPT.doc \13-EP-0R 2-3

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 22 of 46

SECTIONTHREE

Faclors for Funher Siudii

The URS project team reviewed the 2003 Study results and identified seven factors that should be evaluated in order to establish a more complete and definitive set of findings with
regard to the immediate and direct impact of

the MRGO on storm surge:

8.

Complete closure of MRGO. Significant conveyance from Lake Borgne to

the GIWW, upstream of Bayou La Loutre via the MRGO, might stil be occurring in the 2003 Barrier Scenario. To better assess the role of the MRGO in contributing to storm surge, modeling of a complete closure
(fillng-in) of several

the MRGO should be performed. (This factor was also noted by
the 2003 Study.).

local St. Bernard offcials in their review of

9.

The potential for surge reductions throughout the area. The potential for

storm surge reduction should be assessed more systematically across the
surrounding area, and not be limited to the four locations.
10.

Possible impacts to surge scour velocity. Closing the MRGO may impact
surge velocities within the MRGO footprint and these should be examined.

11.

The timing of storm surge arrivaL. Impacts to the timing of surge can be
critical for the evacuation of areas. The evaluation of MRGO contribution to

surge threats needed to also address the timing of storm surge onset.
12.

Wave run-up analvsis. In addition to modeling these surge issues a wave runup analysis for the St. Bernard HPS should be performed to determine if regeneration of waves in the Baseline versus Closed MRGO has a significant impact. (ADCIRC only models the mean sea level during surge.)

These first five factors were incorporated into the URS evaluation of the Baseline versus
Closure scenarios for the hurricane simulations discussed in Sections 4 and 5..
13.

Assessment of impacts with a severe storm. None of the surges for the nine
synthetic storms used in the 2003 Study reached elevations that would

threaten to overtop the HPS. Therefore, URS recommended that a simulation

should be conducted with a storm that generated a surge approaching the
height of the MRGO levee. Hurricane Katrina struck just after the review of the initial simulation results and it was chosen for this assessment.

UR

l;\roJecmILR\1927SosMRGO Sludyn!erim ReportNTERIM-RPT2.docI13-EP.QS\BTR 3- I

'"

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 23 of 46

SECTIONTHREE
14.

Factors lor Funher StudY

Evaluation of the sensitivity of storm surge to levee alignent. While this was not initially a focus of the URS effort, results of initial simulations and concerns over the role of levee al ignents for Hurricane Katrina suggested
that a simulation to examine this issue was needed.

URS conducted a total of seven simulations and Table 5 presents a summary of the grids and

time-steps utilized. A one second time-step was used for the Hurricane Katrina runs to
provide added assurance of

model stability given the higher surge levels.
Table 5

Summary ofURS Simulatious
,

HU,rric8De

SÌlniiation '"
124,Knot-Fast
I (Svnthetic)

", S . ,', ',',

',

. . cena,nos.

Gnd

'" ,

'

Time Step
2 seconds
2 seconds

Baseline MRGO
Closed MRGO

2003 ADCIRC Grid
2003 ADCIRC Grid with MRGO Filed In

Betsy

Baseline MRGO
Closed MRGO

2003 ADCIRC Grid
2003 ADCIRC Grid with MRGO Filed In

2 seconds 2 seconds
1 second
1 second

Katrina

Baseline MRGO
Closed MRGO

2003 ADCIRC Grid

2003 ADCIRC Grid with MRGO Filled In
Height of

Modified Levees

2003 ADCIRC Grid with MRGO Filed In Levee on South Bank ofMRGO and GIWW Reduced Height of Interior Levee Increased

i second

UR

l;\roJect\LNR\1927S06RGO SbJiinlerim ReporNTERJM-RPT2.doc \1~EP-0TR 3-2

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 24 of 46

SECTIONFOUR

Modeling of 124-Knol S!lnlhelic Siorm

The first simulation pair was performed using the 124-Knot Fast synthetic hurricane from the 2003 Study. This storm was chosen as an initial test for five reasons:

.

Use of one of the previous storm scenarios would allow direct comparison of a Closed MRGO and a Bayou La Loutre Barrier simulation.
The 2003 Study showed that this storm produced one of the highest (albeit
stil modest) reductions of surge of any of the Barrier simulations-i.e., 0.53

.

feet at Shell Beach. This reduction was greater than that seen for Betsy.

.

The MRGO is probably a more important factor in surge conveyance for fast storms than for slow storms (due to the greater head differential created in the MRGO by fast storms).

. .
4.1

A strong, fast storm seemed most likely to create the greatest head gradient
along the MRGO.

The fast storm scenario was relatively quick to run.

SIMULATION MODELS

URS utilized the same parallelized version of ADCIRC and the same 314,442 node grid used
in the 2003 Study. Runs were performed on an 8-node cluster owned by WorldWinds, Inc. at

Stennis, Mississippi. (WorldWinds had previously run parallel ADCIRC to assess storm
surge scenarios on the Mississippi Gulf

Coast in 2004.)

To compare the Baseline MRGO scenario URS undertook two pairs of simulations-with the

MRGO channel represented as it had been in the 2003 Study-to a fully Closed MRGO scenario-in which the grid nodes within the footprint of the MRGO were changed to the the surrounding marsh (approximately I foot above MSL). Figure 10 ilustrates elevation of
the 2003 ADCIRC terrain configuration with the Closed MRGO.

4.2

QA CHECK

URS first re-ran the Baseline I

24-Knot-Fast storm and compared results to the 2003 Study in

order to confirm that the model and grid were performing correctly on the W orldWinds

cluster. Results of the "QA Check" run are provided in Table 6 using a comparison of

UR

1:\Project\lDNR\192750-MRGO Siudynterim ReporNNTERIM-RPT.doc\13-EP-oSITR 4~ 1

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 25 of 46

SECTIONFOUR

Modeling 01

124-Knot Synthetic Storm

maximum surge WSE and show that the WorldWinds parallel version of ADCIRC was
performing properly.
Table 6

QA Check Comparison of Maximum Surge WSE URSlWorldWinds ADCIRC Run versus 2003 Study Using the 124-Knot-Fast Storm
,

, Location
IHNC
Paris Road

,' ,
'

,

UR
, , Maxium WSE
, , (ft MSL)
4.18
4.41

2003 Corps

. "

Maxium WSE

" ',, '

(ft MSL)
4.1

4.3

Bayou Dupre

3.85
3.31

3.9
3.5

Shell Beach

4.3
The I

SIMULATION RESULTS

24-Knot-Fast storm was then run with the Closed MRGO grid. Maximum WSE within the area surrounding the MRGO for the Baseline and Closed MRGO simulations for the 124Knot Fast storm are provided on Figures II and 12. Figure 13 depicts the difference between
the two simulations.

Table 7 compares the maximum inundation for the Closure Scenario at the six locations-the

four locations from the 2003 Study, plus Caernarvon and the MRGO Mouth-with
maximum inundations for the Baseline and Bayou La Loutre Barrier Scenarios. (The latter
data is taken from the 2003 Study.) Stage hydro

graphs for both simulations at the six

locations are included on Figure 14.

The Caernarvon location was added in order to examine a lag in storm surge peak reported by local St. Bernard Parish offcials for this area. The lag is clearly seen when comparing the hydrographs on Figure 14 and is attributable to the westerly driven emptying of the surge
from the Lake PontchartrainlBorgne waterbodies, combined with the southwestern

movement in the Gulf below New Orleans in the post-storm period.

UR

l;\aJecbUDNR\1927506-MRGO 5tudy~terlm RepOlNTERIM-RPT.do\1J-EP-05\TR 4-2

c

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 26 of 46

SECTIONFOUR

Modeling of 124-Knot Synthetic Storm

Table 7

Difference in Maximum WSE for Baseline, MRGO Closure, and Bayou La Loutre Barrier Scenarios,
124-Knot-Fast Storm
I, '.

MaximumWSE

I," Difference in " MRGO

Location

I,,
,

(ftl\fSI, "

" ,'.'MRGO
Closhre ,',,"

, ,Baseline

I, ""Çl~sure
-0.03 -0.08

", Barrier'' ',
(ft) "

,',

Max WSE

Difference in Max WSE

(percent), ,,", ,
MRGO' Closure
-0.7 -1.8
4.5
,

Barrier ','
-3.8
-4.3

IHNC
Paris Road

4.18
4.41

4.15 4.33

-0.16

Bayou Dupre
Shell Beach

3.85
3.31

4.02 2.70 2.76 3.99

0.17

-0.19 -0.16
-0.53

-4.2

-0.62
-0.03

-18.6 -0.9
0.7

-16.0

Caernarvan
MRGO Mouth

2.78 3.96

NA
NA

NA
NA

0.03

A positive value is an increase in surge associated with the Barer or Closure and a negative value is a decrease in surge.

4.4

DISCUSSION OF RESULTS

Figures II and 12 show similar patterns of maximum inundation for the Baseline and Closure Scenarios. Figure 13 highlights a few small areas for which the Closure Scenario has a lower surge peak:

. .

Along the MRGO north of Hopedale, upstream of the Bayou La Loutre ridge,
and

Sporadic, isolated pockets of marsh.

There are two areas for which the Closure Scenario exhibited higher surge peak:

. .

Near Chef

Menteur Pass, and Along the MRGO from the GlWW to just above Shell Beach.

Table 7 shows that at three of the four locations (IHNC, Paris Road, and Bayou Dupre),

simulation of full MRGO Closure produced less reduction in peak storm surge (one was actually an increase) than the Bayou La Loutre Barrier. The one location that showed a
greater reduction of peak storm surge with full closure, Shell Beach, saw a slight increase in the peak reduction from 16 percent to 19 percent.

UR

1;\Projei;\lDNR\1922150oMRGO Sludynterim ReporNNTERRM-RPT2,doc \13-EP-0TR 4-3

-,

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 27 of 46

SECTIONFOUR

Modeling of 124-Knol S!lnlhelic Siorm

The Baseline versus Closure hydrographs for the 124-Knot-Fast storm (Figure 14) provide further evidence of little difference in storm surge for these locations except Shell Beach.
The onset of surge at Shell Beach under the Closure Scenario is delayed-with the arrival of

a I foot surge lagging by 5 hours. However, the draining leg of the storm surge at Shell Beach is negatively impacted. Fifteen hours after the peak, the surge remains I foot higher
under the Closure compared to the Baseline Scenario. (Note: the 2003 grid does not include

Bayou Yscloskey. An improved grid which includes Bayous Yscloskey and Bayou La
Loutre might show different results.)

UR

l:\oJectILNR\1922750E-MRGO $ttdynlerim RepONTERIM-RPT.doç \13-EP-0\BTR 4-4

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 28 of 46

SECTIONFIVE

Modeling of Hurricane Beisy

To further assess the, effect of full MRGO closure, URS modeled the Baseline and Closure

Scenarios using the Hurricane Betsy simulation, but without tides. Not including tides shortened the model time by several days (and reduced simulation cost) and was not
expected to affect the relative comparison of scenarios. Hurricane Betsy was chosen because it produced the highest surge peaks of any of the 2003 Study storms. The same grids that 24-Knot-Fast Storm were used for Betsy. were used in the I
5.1

SIMULATION RESULTS

Maximum inundations within the area for the Baseline and Closed MRGO simulations for
Betsy are provided on Figures 15 and 16. Figure 17 depicts the difference between the two

simulations. Table 8 compares the maximum inundation for the Baseline and Closure
Scenarios for Betsy. Stage and current speed hydrographs are presented on Figures 18 and

19.
Table 8

Difference in Maximum Surge WSE for Baseline and Closure Scenarios
Hurricane Betsy
:':1':.

':: , '-':" :~: '¡,~/.: . '-':' : :" :::;1,;:

',:"'. ":',' "~i"~,'

. ' ,DirrereD1è(;:¡iö-:M~x"

'ijifferenc~,inN,åx "
. " ',' ", ":J,

,,::WS~!(ll) . .'.,
IHNC
Paris Road
10.13 10.35

WSE (Pér~él1tr.
3.1

10.44 10.80

0.31

0.45 0.08

4.3

Bayou Dupre

7.47
5.58

7.54 5.50
8.28 6.51

1.
-1.4
0.4 0.6

Shell Beach

-0.08

Caernarvan
MRGO Mouth

8.24 6.47

0.04 0.04

A positive value is an increase in surge associated with Closure and a negative value is a decrease in surge.

5.2

DISCUSSION OF RESULTS

Figures 15 and 16 again show similar patterns of peak surge for the Baseline and Closure
Scenarios. Figure 17 shows that the Closure Scenario produced no areas in which surge

reduction reached one foot. However, three areas of higher surge peaks occurred under the
Çlosure Scenario:

. .

Hopedale, the MRGO at the confluence with the GIWW, East of

UR

1:\Prale\LDNRI192750-MRGO SbJiinte~m RepoNTERIM-RPT.doc\13-EP-05ITR 5-1

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 29 of 46

SECTIONFIVE

Modeling of Hurricane Beisy

.

Along the MRGO near Proctor Point. area for the Closure Scenario;

Five of the six locations on Table 8 indicate increases,' albeit slight, in surge peak for
Hurricane Betsy Closure versus Baseline Scenarios. Only Shell Beach saw a decrease in the the i 24-Knot-Fast storm. peak, 1.4 percent, a much smaller reduction than the 19 percent for
The Hurricane Betsy stage hydro

graphs (Figure 18) for Baseline versus Closure Scenarios

again are nearly identical at each location, with the exception of Shell Beach. Similar to the 124-Knot-Fast storm, surge onset is slightly delayed as Shell Beach under the Closure Scenario, and the draining leg is negatively impacted.

Figure 19 provides a comparison of Betsy Baseline versus Closure current speed
graphs for the six locations. In the case of the MRGO closure scenarios this current is graphs are similar for four locations (IHNC, Paris Road, MRGO Mouth, and Caemarvon), but are significantly different for Bayou Dupre and Shell Beach. At Bayou Dupre the maximum current for Closure is lower by about 0.5 fps (on a Baseline peak flow of about 2.2 fps) and the duration over which the current exceeds I fps is cut by over half. At Shell Beach the maximum current for Closure is halved, from 7 to 3.5 fps. The Baseline Scenario depicted Shell Beach currents above 3.5 fps for 6 hours. This is a significant difference considering the scouring impact of the higher currents on the adjacent marsh. Moreover, actual velocities are higher due to added energy of wave action.
hydro for overland flow. The speed hydro

UR

l:'oJecmUDNRl1927506MRGO SIu'¥nler1m ReportNTERJM-RPT2.doe \13-EP-0TR 5-2

'-

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 30 of 46

SECTIONSIX

Modeling of Hurricane Katrina

The Hurricane Betsy simulation surge elevations did not approach the current HPS elevations, reaching 10.35 feet above MSL in the GIWW at Paris Road. URS therefore
recommended performing a simulation with a truly severe storm to confirm the preliminary conclusion. Following the surge impact of Hurricane Katrina on the study area on August

29, 2005, URS, at LDNR's direction, undertook to use Katrina as a diagnostic storm to further assess the role of MRGO channel conveyance on peak surge. URS conducted
Hurricane Katrina MRGO Baseline and Closure simulations using the same ADCIRC 24-Knot Fast Storm. Baseline and MRGO Closure grids used for the I
6.1

HURRICANE KATRINA

The track of Hurricane Katrina is shown in Figure 9. Katrina's forward speed was about 17.4 miles per hour as it was crossing St. Bernard Parish. The URS simulations used a
synthetic wind-field file prepared by Dr. Pat Fitzpatrick of WorldWinds, Inc.. Top winds

and central pressure on which the simulations were based are shown in Figure 20. The simulated top winds following landfall at Buras, Louisiana are more than 10 knots higher than the revised values provided by the National Hurricane Center in their December 20,
2005 report. The W orldWinds wind model does not take into account structural changes that the hurricane was undergoing at the time of landfall-including degradation of the southern eyewall. As a result the modeled Katrina is a more powerful storm than the actual one. The model does not include tides, and as with the earlier runs, uses a reference datum of MSL.

The URS ADCIRC Katrina simulations allow HPS overtopping but do not include HPS
breaches and failures. These simulations examine the relative impact of MRGO closure for a severe storm and are not meant as a detailed reproduction of actual Katrina events and conditions.
6.2
SIMULATION RESULTS

Table 9 compares the maximum inundation for the Hurricane Katrina Baseline versus
Closure scenarios. Figures 21 and 22 ilustrate the maximum water surface elevation (WSE)

for the MRGO Baseline and Closed scenarios. Figure 23 depicts the difference in peak surge between the two simulations. Stage and current speed hydrographs for selected locations are provided in Figures 24 and 25.

UR

l;\oJect\LNR\1927506MRGO Sludyntem ReporNNTERJM-RF.do. \13-EP-DTR 6-1

,

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 31 of 46

SECTIONS

ix
Table 9

Modeling 01 Hurricane Kalrina

Difference in Maximum Surge Water Surface Elevations for Hurricane Katrina Simulation of MRGO Baseline and Closure
,

Location

Maximum WSE (ft. MSL)

-c

'

,,',
DilTe,renee iii Max

"'

Baseline

Closure
16.1

' WSE (ft)
,

Differe'nce in Max WSE (perce~t)
-0.31

IHC
Paris Road

16.1

-0.05 -0.57 -0.03 -0.26

19.8

19.2

-2.86 -0.12 -1.24 -1.04
0.36

Bayou Dupre
Shell Beach

22.3

22.2
20.3
19.3

20.6
19.5

Caemaron
MRGO Mouth

-0.20
0.07

19.8

19.9

A positive value is an increase in surge associated with Closure and a negative value is a decrease in surge.

6.3

DISCUSSION OF RESULTS

As shown in Table 9 and in Figures 21, 22, 23, and 24, the Katrina simulations demonstrated
nearly identical patterns in surge WSE for the Baseline and Ciosure scenarios. The figures

clearly show that a massive wall of surge advanced entirely across the surrounding lakes and

marsh. There were only very slight reductions in peak WSE-amounting to less than 3
percent-with the closure scenario at Paris Road, Shell Beach, and Caernarvon (0.6, 0.3 and

0.2 ft, respectively). The peak surge was unaffected at the IHNC and Bayou Dupre by
closure, and was slightly increased at the Mouth of

the MRGO.

Figure 23 shows that the difference in peak surge for the two scenarios was less than 0.5 ft over the vast majority ofthe region. The 0.6 ft reduction at Paris Road is indicated on Figure 23 by the yellow area, which extends into portions of New Orleans East. The reduction in New Orleans East under the closure scenario is attributable to the slightly lower water levels in the GIWW at Paris Road, which resulted in slightly reduced overtopping of the GIWW
levee.

The stage hydrographs in Figure 24 further demonstrate that the Hurricane Katrina simulations did not produce any significant instances of surge increase or delay with the

MRGO closure scenario. Nor do the stage hydrographs show any evidence of impeded
drainage with the closure scenario.

The vertical scales on the individual current speed hydrographs in Figure 25 show that the rate of surge flow throughout the system is highly variable. Maximum velocities at the

UR

r:\ProJecl\LR\1927S0-MRGO Studfnlerim ReporfNTERIM-RPT.doc \13-EP-05\TR 6-2

~

Case 1:05-cv-01119-SGB

Document 27-5

Filed 10/04/2006

Page 32 of 46

SECTIONS

ix

Modeling of Hurricane Karrina

MRGO Mouth, Caernaron, Shell Beach, and Bayou Dupre are all in the range of 3.5 to 6 ftsec. The maximum velocity in the IHNC is about 1.5 ftsec under both scenarios (the simulations do not include breaching of the floodwalls). The highest velocity occurred at Paris Road-12.5 ftsec, which is reduced slightly to 11.6 ftsec under the closure scenario. The largest impact to velocity from closure occurs at Shell Beach, with an increase from 3.8
to 5.9 ftsec. Closure of the MRGO channel does reduce the duration of flow above I ftsec
at Bayou Dupre.

The modeled current speeds discussed here represent basic longitudinal flows (up and down

the channel), and do not include the additional energy of wave action. High surge
longitudinal flow will cause scour along soft, unarmored channel bank and bottom soils. The

HPS scouring observed as a result of Katrina was reportedly due to lateral flow, which
occurred when water levels exceeded the height of HPS structures and they were overtopped.

URS understands that there have been no major reports of longitudinal scouring of HPS structures to date. This is likely attributable to the fact that most of the HPS structures are
typically located hundreds of feet from the channel bank. High longitudinal currents do have

the potential to cause scour damage at structures supported by the channel bank and
bottom-such as bank revetments and piers for docks and bridges