Expert Report of Dr. Patrick J. Fitzpatrick, Ph.D Case 1:06-Cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 2 of 40

Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 1 of 40

Exhibit D –

Expert Report of Dr. Patrick J. Fitzpatrick, Ph.D Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 2 of 40

The wind and surge of Hurricane Katrina on 2558 S. Shore Drive, Biloxi, MS

Dr. Pat Fitzpatrick Consultant meteorologist 180B Lakeview Drive Slidell, LA 70458

NASA satellite image from Terra’s MODIS sensor on August 28, 2005, at 12:00 PM. Hurricane Katrina was about 200 miles from southeast Louisiana at this time as a Category 5 hurricane. Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 3 of 40

Introduction

This report presents information about Katrina’s wind and storm surge elements. Section 1 outlines the ways hurricanes cause damage. Section 2 presents data and the time evolution of wind and surge which impacted 2558 S. Shore Drive. Section 3 summarizes this report’s conclusions.

1. Ways Hurricane Cause Destruction

Hurricane intensity is officially defined by sustained winds, not instantaneous winds. Sustained winds are the average speed over a period of time at 33 feet above the ground. In the Atlantic, this averaging is performed over a 1-minute period. The actual wind will be faster or slower than the sustained wind at any instantaneous period of time. Obviously, hurricane sustained winds are a source of structural damage. As winds increase, pressure against objects increases at a disproportionate rate, roughly the square of the wind speed. For example, a 50-mph wind causes a pressure of 5.5 pounds per square foot. In 100-mph winds, that pressure becomes 30 pounds per square foot. When the wind exceeds design specifications, structural failure occurs. Debris is also propelled by strong winds, compounding the damage.

To standardize intensity measurements worldwide, not only is time-averaging required, but a height level needs to be defined. The World Meteorological Organization states that official hurricane wind measurements represent 10-meter (33-feet) height. Since winds are rarely measured at this level, mathematical assumptions are required to normalize wind measurements to this height. The lack of observations, as well as the standardization of winds, creates additional uncertainty about the true nature of the hurricane wind field.

A sudden, brief increase in wind speed for a few seconds is known as a wind gust. Wind gusts in a hurricane are typically 20-50% higher than sustained winds. They initiate and amplify much of the wind damage. A hurricane with maximum sustained winds of 100 mph will contain gusts of 140 mph or more. Wind gusts are associated with precipitation downbursts or turbulent eddies near the ground.

Downbursts accompany thunderstorms in areas where heavy rainfall accelerates air to the ground. Collapsing cores of heavy precipitation are prevalent in many hurricanes. These downburst winds, when superimposed on the already strong hurricane wind field, produce localized regions of extreme winds and damage. Heavy convective rain can generate precipitation-induced downdrafts by transporting high-velocity air aloft towards the surface, then violently spreading outward along the ground. An example of downbursts by rain loading is shown in Fig. 1.

2 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 4 of 40

Strong Winds

Figure 1. Schematic of a downburst by rain loading top, resulting in extreme wind gusts after crashing into the ground (bottom). These heavy rain episodes are associated with radar returns greater than 50 dBz.

High-resolution imagery from radar imagery has shown that downbursts are also associated with turbulent eddies in hurricanes. These eddies manifest themselves as “rolls” due to vertical mixing of winds of different intensities. These rolls also transport fast winds aloft to the surface (Fig. 2).

Figure 2. Schematic representation of observed shear- and wind-parallel boundary layer rolls. High-momentum air (red) is brought to the surface in the downward legs of the rolls, while air slowed near the surface is brought aloft in the upward leg. The result is wind gusts.

3 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 5 of 40

Powerful wind entities also occur in isolated regions. As hurricanes make landfall, frictional ground interaction with the thunderstorms form columns of rapidly rotating air in contact with the ground, known as tornadoes. Tornadoes are especially prolific in the right front quadrant at landfall. Hurricane tornadoes tend to cluster near the hurricane core (within 100 km), and in the outer rainbands about 300 km from the center. Hurricane tornadoes are weaker with shorter tracks than the “supercell” variety seen in the Great Plains. But, they still contain winds up to 157 mph. On the day of landfall, tornadoes occur close to the center with proportionally fewer tornadoes in the outer rainbands. However, on the days following landfall, tornado occurrences show an increasing preference for the outer rainbands. Tornadoes in outer bands peak in the early afternoon due to a maximum solar heating, while inner-core tornadoes show no diurnal peaks. Large hurricanes produce more tornadoes than small hurricanes. Tornado frequency also increases in the more intense hurricanes. Finally, hurricanes with a slow or fast translation speed produce few tornadoes, while hurricanes with a motion between 8 and 33 mph contain tornadoes. Based on these composite studies, Katrina was a good candidate for tornado activity.

In addition, another phenomenon called mesoscale vortices, or sometimes mesovortices, was documented in Hurricane Hugo (1989) and Hurricane Andrew (1992). These are whirling vortices that form at the boundary of the eyewall and eye where there is a tremendous change in wind speed, and can be 5-10 times wider than a tornado (Figs. 3 and 4). The updrafts in the eyewall can stretch the vortices vertically, making them spin faster with winds up to 200 mph. An instrument which measures the vertical distribution of winds in hurricanes, called a dropsonde, was deployed in Hurricane Isabel (2003) and fell through a mesovortex, measuring winds of 241 mph.

Features associated with wind enhancement are summarized in Table 1.

4 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 6 of 40

Figure 3. Computer simulation depicting edge of hurricane eyewall breaking down into a series of mesoscale vortices.

5 mesovortices

Figure 4. Example of mesovortices in Hurricane Isabel.

5 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 7 of 40

Table 1. Destructive wind features in a hurricane.

Sustained wind The average wind speed over a period of time at 33 feet above the ground. In the Atlantic this averaging is performed over a 1- minute period. As winds increase, pressure against objects increases at a disproportionate rate, roughly the square of the sustained speed. For example, a 50-mph wind causes a pressure of 5.5 pounds per square foot. In 100-mph winds, that pressure becomes 30 pounds per square foot. Wind gust A sudden, brief increase in speed of the wind. The duration of a gust is usually less than 20 seconds. Peak 3-second gusts are the standard used. Wind gusts are typically 1.3-1.5 times larger than tropical-storm force sustained winds, and 1.2-1.4 times larger than hurricane-force sustained winds. Wind gusts are associated with downbursts and turbulent boundary layer rolls. Gusts usually initiate structure damage, and then amplify damage caused by the sustained wind. Downburst A strong downdraft that exits the base of a thunderstorm and spreads out at the earth's surface as strong and gusty horizontal winds that can cause property damage. Radar echoes greater than 50 dBz are often associated with downbursts. Tornado A rapidly rotating column of air that protrudes from a cumulonimbus cloud in contact with the ground, often (but not always) visible in the shape of a funnel or a rope. The right front quadrant of a hurricane often produces many tornadoes at landfall due to ground friction, but they can appear in any hurricane squall line. A radar can identify some – but not all – tornadoes using a “mesocyclone” algorithm which identifies rotating thunderstorms with symmetry. Boundary layer Violent but organized eddies which manifest themselves as rolls “rolls” due to vertical mixing of winds of different intensities. These rolls transport fast winds aloft to the surface. Mesovortex Whirling vortices that form at the boundary of the eyewall and eye where there is a tremendous change in wind speed. Winds may be up to 200 mph, especially in areas where winds are the same direction as the eyewall winds, and therefore extremely destructive. 5-10 times wider than a tornado, perhaps even larger in some cases. Some studies suggest they have a ratio 1/10th the diameter of the hurricane eye. They are believed to occur in major hurricanes (Category 3 or more). Also called mesoscale vortices.

6 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 8 of 40

Accompanying a landfalling hurricane is the storm surge, defined as an abnormal rise of the sea along the shore generated by an intense storm such as a hurricane. The storm surge is caused primarily by the winds pushing water toward the coast and wave breaking, which propels water further inland. A secondary contribution to surge is made by the reduced barometric pressure within the storm, which causes a dome of water level higher than the surrounding ocean. However, wind and wind-generated waves are the primary contributors to storm surge. The surge begins to rise gradually, then faster as the storm makes landfall. Despite some ill-conceived notions, it is not like a tsunami or a wall of water, but instead a steady increase in water levels. Typically the surge peaks after landfall, with a region experiencing tropical storm- and hurricane-force winds several hours before landfall.

Factors which impact storm surge elevation include: • Storm size: The larger the areal extent of tropical storm-force winds, the higher the water elevation • Storm central pressure: Lower interior atmospheric pressure increases the water level. Pressure is essentially the “weight” of the atmosphere. Water expands as pressure decreases, known as the inverse barometer effect. Since the pressure is lowest in the storm center, the higher pressure outside the hurricane also pushes water inward. The result of these two forces causes water to expand 3.9 inches for every 10-mb pressure drop. Overall, this is a small contribution to the storm surge. In Katrina, the inverse barometer effect contributed 2-3 feet of water to the surge. • Storm intensity: The maximum wind speed is the most important factor. The more intense the hurricane, the higher the water elevation. • Bathymetry: As the surface currents driven by the wind reach shallow coastlines, bottom friction impedes the seaward return flow near the bottom, causing water to pile up. Shallow areas with a gradual slope will experience greater storm surges than areas with a shelf that drops off rapidly near the coast. This is because water cannot sink and flow outward to the ocean, thereby causing more water to pile up offshore when the water is shallow. Because of Louisiana and Mississippi’s proximity to shallow water that gradually deepens offshore, these states are prone to high storm surges. • Speed of motion of the system: Because a slow-moving hurricane has a longer time to transport water onshore, slow systems are associated with higher storm surge values. Slower moving hurricanes can cause a storm surge 20-30% higher than fast moving hurricanes. Fast moving hurricanes cause the surge to “spike” over fewer hours with an overall lower surge. This process is most pronounced in shallow bathymetry. • Wave setup: Water levels can increase from onshore waves in windy conditions. Under normal conditions, waves that reach the coast break and water flows back out to the sea under the next incoming wave. In hurricane conditions, the water may not retreat in time before the next wave arrives, a situation called wave setup. This wave setup can be quite large and is most pronounced when deepwater is near the shore, because in shallow water waves break further offshore. Wind- induced surge enables waves to penetrate much further inland before they break.

7 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 9 of 40

• Track angle: Storms which make landfall perpendicular to the coastline produce larger storm surges than those which hit at an angle. Storms which make landfall at an angle have a smaller surge because some transported water experiences reflection and cross-current transport. • Local effects: The shoreline trajectory can enhance or weaken the surge through trapping mechanisms.

The storm surge is always highest on the side of the eye corresponding to onshore winds, which is usually the right side of the point of landfall. Winds are also fastest in the right front quadrant because storm motion (which averages about 10 mph but varies substantially) is added to the hurricane's winds. Because winds spiral inward, the storm surge is greatest along the eyewall but high water can impact other regions as well.

The total elevated water includes additional components - the astronomical tide, the steric effect, and the week’s previous wind patterns. The astronomical tide results from gravitational interactions between the earth and the moon and sun, generally producing two high and two low oceanic tides per day in most U.S. locations, but only one high and one low tide per day in Louisiana. Should the storm surge coincide with the high astronomical tide, the additional elevation will be added to the water level. However, tide ranges along the northern Gulf of Mexico are small, only contributing to one-foot of additional water at high tide, often less. Because warm water expands, water levels are naturally highest in the summer, known as the steric effect. In the Gulf of Mexico, this contributes about 0.52 feet of water in late summer. Finally, should the primary wind be onshore the previous week, this will also increase water levels 0.5—1 feet.

By definition, storm surge water levels do not include wave marks. Waves will be superimposed on the storm surge and will yield higher values on outside trees and structures that do not reflect water levels inside buildings. Because wave heights also vary based on proximity to the coast, outside water marks will vary and need to be considered when validating storm surge levels. Miles offshore in deep water, the waves will be large. However, as the depth decreases toward the shore, waves are impacted by the ocean floor and slow down while their period remains constant. As a result, the wavelength decreases and the amplitude increases. Eventually the wave will get too steep and break. New waves will be generated with less height, but as the depth continues to decrease, they will again break and reform as smaller waves. In theory locally generated shallow water wave heights can reach 73% of the water depth, but the distance traveled and time required to reach its potential maximum height (called the fetch) is too short near the shore; because the depth keeps decreasing, wave growth becomes disrupted and the wave will break again and again. In addition, shallow water waves also lose energy due to frictional interaction with the ocean floor. Frictional loss is even greater over flooded, vegetated land. In Mississippi, in the surf zone, wave heights will reach 1-4 feet on top of the surge. Further inland, the wave height will be less than 2 feet, reducing with distance from the coast or with land elevation.

The expected level of damage for a given hurricane intensity is described by the Saffir- Simpson Hurricane Scale. It was devised in 1971 by Herbert Saffir, an engineer in

8 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 10 of 40

Miami, for the World Meteorological Organization. Robert Simpson, the director of the National Hurricane Center, then added the storm surge portion. This scale classifies hurricanes into five categories according to central pressure, maximum sustained winds, storm surge, and expected damage (Table 2). Although all categories are dangerous, categories 3, 4, and 5 are considered major hurricanes, with the potential for widespread devastation and loss of life. Whereas only 21 percent of U.S. land-falling tropical systems are major hurricanes, they historically account for 83 percent of the damage. Note that the scale is not linear. A Category 3 such as Katrina hurricane causes 50 times as much damage as a Category 1, and a Category 4 is 250 times more destructive than a Category 1. The scale does not account for all factors affecting storm surge, and in fact the National Hurricane Center is removing any discussion of storm surge from a revised Saffir-Simpson scale. Recently, this author has developed a new surge scale based on the factors discussed on page 7, and it is available upon request (or see the presentation at the Interdepartmental Hurricane Conference website in Section 13 at http://www.ofcm.gov/ihc09/linking_file_ihc09.htm).

9 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 11 of 40

Table 2. The Saffir-Simpson Scale for Atlantic hurricanes. The scale does not account for all factors affecting storm surge, and the National Hurricane Center is removing any discussion of storm surge from a revised Saffir-Simpson scale. This author has developed a new surge scale based on the factors discussed on page 7, and it is available in Section 13 at http://www.ofcm.gov/ihc09/linking_file_ihc09 htm. Potential Damage Scale” provides a scale relative to a category 1 hurricane, where a category 1 is scaled as “1.” For example, a Category 3 hurricane typically causes 50 times as much damage as a Category 1 hurricane.

Category Central Maximum Storm Potential Damage pressure sustained surge Damage (approx) winds in feet Scale mb (mph) (approx) inches 1 > > 74-95 4-5 1 Damage primarily to shrubbery, trees, foliage, Minimal 979 28.9 and unanchored mobile homes. No real damage to building structures. Low-lying coastal roads inundated, minor pier damage, some small craft in exposed anchorages torn from moorings 2 965- 28.5- 96-110 6-8 10 Considerable damage to shrubbery and tree Moderate 979 28.9 foilage; some trees blown down and major damage to exposed mobile homes. Some damage to roofing, windows, and doors of buildings. Coastal roads and low-lying escape routes inland cut by rising water two to four hours before arrival of hurricane center. Considerable pier damage, marinas flooded, small craft torn from moorings. Evacuation of shoreline residences and low-lying island areas required. 3 945- 27.9- 111-130 9-12 50 Large trees blown down. Foliage removed from Extensive 964 28.5 trees. Structural damage to small buildings, mobile homes destroyed. Serious flooding at coast and many smaller coastal structures destroyed. Larger coastal structures damaged by battering waves and floating debris. Low- lying inland escape routes cut by rising waters 3-5 hours before arrival of hurricane center. Low-lying inland areas flooded eight miles or more. Evacuation of low-lying structures within several blocks of shoreline possibly required. 4 920- 27.2- 131-155 13-18 250 All signs blown down. Extensive damage to Extreme 944 27.9 roofing, windows, and doors. Complete failure of roofs on smaller buildings. Flat terrain 10 feet or less above sea level flooded as far as 6 miles inland. Major damage to lower floors of coastal buildings from flooding, battering waves, and floating debris. Major erosion of beaches. Massive evacuation: all residences within 500 yards of shore and single-story residences on low ground within two miles of shore. 5 < < > 155 > 18 500 Severe and extensive damage to residences and Catastrophic 920 27.2 buildings. Small buildings overturned or blown away. Severe damage to windows and doors; complete roof failure on homes and industrial buildings. Major damage to lower floors of all structures less than 15 feet above sea level. Flooding inland as far as 10 miles. Inland escape routes cut 3-5 hours before arrival of storm center. Massive evacuation of residential areas on low ground within 5-10 miles of shore.

10 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 12 of 40

Tornado damage also can also be categorized by a scale, known as the Fujita scale. The scale ranges from F0 to F6. The first three scales (F0, F1, and F2) have winds of 40-72, 73-112, and 113-157 mph, respectively. While useful for identifying and categorizing tornado damage and intensity, this also means major hurricanes, in general, have the wind devastation of an F0 or F1 tornado over a wide region! In fact, the National Weather Service issued tornado warnings throughout the entire impact regions since Katrina had F0 and F1-like winds. This scale has been updated using an Enhanced Fujita Scale.

2. The wind and storm surge of Katrina on S. Shore Drive

I. Katrina’s windfield

Katrina was a major hurricane when it made landfall in Biloxi. Because it was also an unusually large hurricane, Mississippi and Louisiana were exposed to hurricane-force winds for many hours, including several hours before landfall. Katrina’s hurricane-force winds extended 120 miles from the storm center, and tropical storm-force winds 230 miles outwards. Katrina also maintained a large eye, thereby providing a large areal- coverage of its most fierce winds. Satellite images, National Weather Service radar, airborne radar (from the Hurricane Research Division), dropsonde data, buoy data, and an Ingalls Shipyards’ anemometer provide intriguing insight into the three-dimensional structure of the hurricane. An outer-core band of strong thunderstorms from a second eyewall impacted the region. Eyewitness accounts also describe intense winds on the early morning of August 29, including the testimony of the director of the Civil Defense for Jackson County Butch Loper, Greg Verdes at 1022 Legion Lane in Ocean Springs and the Church’s located at 2545 S. Shore BLVD near the McIntosh residence. The Hurricane Research Division sustained wind analysis (a product known as HWINDS) is shown in Fig. 5, and depicts the wind evolution of Katrina’s landfall. One wind gust product from Nvision Solutions is shown in Fig. 6.

The Hurricane Research Division HWINDS analysis was used as guidance to determine the sustained winds at 2558 S. Shore Drive. This product accounts for open exposure along the coast, and frictional dissipation inland. Reconnaissance data, buoys (while functional), ASOS (while functional), NERRS, university portable towers, and Ingalls were also consulted to determine sustained speeds. One dropsonde recorded 76 mph winds at 47 feet at 6:12 AM 2.3 miles southwest of the McIntosh residence. Tropical storm-force winds began around 1:00AM August 29 on S. Shore Drive, with hurricane- force winds beginning right after 6AM. Peak winds occurred on S. Shore Drive between 9:00-10:00AM with 100 mph sustained winds associated with the inner eyewall. Hurricane-force, then tropical storm-force winds continued for another few hours, but of less intensity. In other words, S. Shore Drive was subject to tropical storm-force winds from conservatively 1AM to the late afternoon, and hurricane-force winds from 6AM to 11:30AM. The early morning winds are conservative; it’s likely the sustained winds were even stronger.

11 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 13 of 40

Portable university wind towers, ASOS observations (while functional), and buoys (while functional) indicate 20-60% higher than the sustained winds, consistent with peer-review articles on hurricane wind gusts. The peak wind gust at S. Shore Drive is 120 mph, which is also consistent with radar and dropsonde wind data. This general area (Biloxi) received some of the strongest wind gusts on the Mississippi coast. Two dropsondes were deployed near Bay St. Louis and Gulfport around 6:00AM which recorded winds of 115 mph and 119 mph at an altitude between 500 and 1000 feet, three hours before landfall (and the peak sustained winds). Another dropsonde was south of Biloxi at 8:06AM which recorded winds between 129 mph and 142 mph at an altitude between 800 and 1200 feet, 90 minutes before landfall (and the peak sustained winds). Downbursts associated with severe squall lines can transport and mix these winds to the surface.

The first squall line containing a radar reflectivity of between 45-50 dBZ arrived at 5:45AM, signifying when such winds gusts could be transferred downward. Microwave imagery, which is strongly attenuated by hydrometeors (suspended water and ice particles, as well as precipitation), clearly shows this squall to be a well-formed curved band which is likely an outer eyewall. The SFMR instrument also shows this outer eyewall. This outer eyewall reached S. Shore Drive about 6:00AM, initiating peak wind gusts reaching 100 mph, with even stronger gusts likely in isolated regions. The inner eyewall reached S. Shore Drive around 9:00AM. At landfall, another dropsonde in Bay St. Louis showed winds of 155 mph at 1000 feet. This indicates that wind gusts between 130 and 140 mph were likely along open exposure areas in this region at this time. Inland the wind would decrease, but gusts were still at least 100 mph.

Ingalls’ data and the Dauphin Island CMAN station (18 miles east of Ingalls) provide additional information regarding wind gusts and the outer eyewall. Wind gusts at Ingalls reached over 100 mph at 6:00AM associated with the first squall line containing a radar reflectivity of 60 dBZ. The Ingalls data showed peak wind gusts of 117 mph at 7:50AM and 9:20AM. As will be discussed, an eyewitness account at the Jackson County EOC reports a wind gust of 137 mph. Dauphin Island showed wind peaks up to 103 mph, and matches the Ingalls’ gust trends well, providing confidence that both observations are credible. One expects that wind gusts in Biloxi in open exposure would be at least 20-30 mph stronger than either of these locations. In fact, a Texas Tech tower at Stennis airport measured a wind gust of 105 mph, and this was an inland location 8 miles from the Waveland beaches.

Based on this analysis, pre-landfall USGS tide gauge data, and other National Weather Service observations, a timeline can be established for the wind at 2558 S. Shore Drive, and is summarized in Table 4.

12 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 14 of 40

Figure 5. Hurricane Research Division winds from 5:30AM to 9:42AM for southeast Louisiana and Mississippi.

Figure 6. One estimate of wind gusts from Nvision Solution, Inc.

13 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 15 of 40

Radar imagery shows squall lines impacting S. Shore Drive starting after midnight on August 29, with reflectivity values greater than 50 dBz starting after 3:00AM. Recall that this kind of squall lines contains downburst activity. An example of the downburst activity in Katrina is shown in Fig. 7. Boundary layer rolls also contained wind gusts on the edge of these squall lines. Winds were very fast aloft, exceeding 160 mph at a height of 3 km (Fig. 8). Wind profiles measured from an instrument called a dropsonde – released by the hurricane reconnaissance flights – also show much faster winds between 120 and 140 mph between 500 and 1000 feet aloft. The microbursts and boundary layer rolls provide a mechanism to bring these very destructive winds to the surface as wind gusts.

Figure 7. Radar (right column) and a cross-section of the vertical line (left side), showing rain loading and the resulting downburst. Many instances of this phenomenon throughout the impact area have been documented by the University of South Alabama in Hurricane Katrina.

14 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 16 of 40

Figure 8. Airborne Doppler radar-derived wind speed cross sections obtained from a Hurricane Hunter aircraft at approximately (a) 3:00PM August 28 and (b) 5:00AM August 29. Radial distance from the center of the hurricane increases to the right, and both cross sections extend toward the east. Wind speeds are in meters per second (m/s) as indicated by color shade legend. Wind speeds extend down to the 300-m level. Note the broad and elevated wind maximum in the 2-4 km layer on August 29, which was not present on August 28 when the maximum winds were concentrated at the more typical location near 500 m.

15 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 17 of 40

II. Tornado activity in Katrina

Tornadoes are documented either by Doppler radar, post-storm surveys, or by eyewitness accounts (particularly trained “weather spotters”). Tornado documentation thus far in Katrina has been lacking from the National Weather Service and storm spotters. As a result, one is left with using Doppler radar as a tool to identify potential tornadoes at landfall.

Doppler radar measures radial velocity and precipitation intensity (reflectivity) which are input into automated algorithms. One algorithm attempts to identify mesocyclones, defined as rotation in a thunderstorm, typically around 2-6 miles in diameter, and associated with an existing tornado or potential tornado formation. The circulation of a mesocyclone covers an area much larger than the tornado that may develop within it. Doppler Radar includes the Mesocyclone Detection Algorithm which identifies circulations in thunderstorms that have the potential to spawn tornadoes. The software identifies Doppler velocity differences of 25– 75 ms−1 across core diameters of 2–8 km, with resulting azimuthal shear values of 5 × 10−3 s−1 to 2 × 10−2 s−1. It also looks for symmetry in the signal before being identified as a mesocyclone.

Unfortunately, the mesocyclone algorithm is also fraught with uncertainties. The software was developed for mid-latitude, inland severe thunderstorms, not hurricanes. Furthermore, hurricane mesocyclones tend to be small, shallow, and weak, resulting in a low probability of detection. Finally, many mesocyclones do not produce tornadoes. Actual observed mesocyclones (not radar-detected) show 10-30% of mesocyclones in the Great Plains produce tornadoes. It’s not known what this ratio is in radars, and what is valid in hurricanes. In summary, not all mesocyclones are detected in hurricanes, and only a small fraction of radar-detected mesocyclones spawn tornadoes. However, since an equal number are not detected, this implies the false signals and undetected signals cancel each other out, giving a good estimate of the total number of tornadoes.

With these caveats in mind, both the Slidell and Mobile Doppler radars show mesocyclone activity. The Mobile radar was operational throughout Katrina’s landfall, while Slidell quit transmitting data around 1400 UTC (9AM). The Mobile mesocyclone signals are shown in Figure 9. When one accounts for the uncertainties involving the ratio of tornadoes to mesocyclones, duplicate mesocyclones, and unseen mesocyclones, one could estimate several dozen tornadoes during Katrina’s landfall. A recent Geophysical Research Letter paper by Lee, Bell, and Goodman (2008) documented 23 supercells in Katrina near landfall capable of producing multiple tornadoes.

With regard to 2558 S. Shore Drive, several mesocyclones are seen in the region. While it’s not difficult to pinpoint a tornado impact at this location due to the surge obscuring a tornado’s path, this structure could have been impacted by a tornado. Nonetheless, sustained winds and wind gusts were strong in the region.

16 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 18 of 40

Figure 9. Mesocyclone signatures detected by the Mobile Doppler radar August 28 and 29. This algorithm does not detect mesocyclones with a low cloud base, so more mesocyclones are likely.

17 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 19 of 40

III. Mesovortices in Katrina

Little information is available on mesovortices in Katrina. However, NASA’s polar- orbiting Terra and Aqua satellites, which have a MODIS sensor with resolution of 250 meters, can see these in a clear eye. A MODIS image at 12:15PM on August 28 shows several eyewall mesovortices when Katrina was 200 miles from southeast Louisiana (Fig. 10). Unfortunately, because polar-orbiting satellites can only take one image in the same geographical region per day, no MODIS images are available when Katrina was off the Mississippi coast. Research and limited observation studies show that intense hurricanes such as Katrina commonly contain mesovortices. Therefore, conditions were favorable for mesovortices during the Mississippi landfall.

Figure 10. NASA satellite image from Terra’s MODIS sensor on August 28, 2005, at 12PM. This 250- meter resolution image, zoomed in on Katrina’s eyewall, shows swirls that are mesovortices. The hurricane was about 200 miles from southeast Louisiana at this time.

18 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 20 of 40

IV. Timing of wind and storm surge in Katrina at S. Shore Drive

Observations of Katrina’s storm surge life cycle generally do not exist because all tide gauges failed in the southeast Louisiana marsh and Mississippi during the brunt of the storm. The previous few days of water levels, as well the first few hours of the storm surge, were documented. Typically, one to two days before a storm such as Katrina makes landfall, the water increases 2-3 feet, known as the surge forerunner. On the day of landfall, water slowly increases, then rises faster as the eyewall makes landfall.

Despite the shortcomings of the gauges, they do provide a record of the wind and the surge before the eyewall comes onshore. They show unequivocally that tropical storm- force winds arrived several hours before the surge. A sample of Mississippi and Louisiana tide gauges are shown in Table 3, indicating that winds greater than 39 mph, and approaching hurricane strength, arrived between 4 and 8 hours before surge values of 8 feet occurred, less than would flood most homes.

Figure 11. Wind and water level values for tide gauges before operating failure at mouth of Pearl River (top left), Mississippi Sound (top right), and Ocean Springs (bottom).

19 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 21 of 40

Table 3. Summary of wind and surge at three USGS Mississippi gauges (Ocean Springs, Mississippi Sound, and the mouth of the Pearl River). Two from Louisiana are also shown (Bay Gardene and Bayou La Loutre). Note that tropical storm-force winds occurred for several hours with surge insufficient to inundate most properties.

Wind (mph) Storm surge (feet) Locations Time 42 4.3 Ocean Springs 8/29 at 2:30AM 72 9.6 Ocean Springs 8/29 at 7:15AM 36 2.3 Mississippi Sound 8/29 at 12:00AM 53 5.9 Mississippi Sound 8/29 at 4:00AM 40 4.4 Bay Gardene 8/28 at 5:15PM 58 6.9 Bay Gardene 8/29 at 12:00AM 35 1.3 Bayou La Loutre 8/28 at 9:00PM 56 3.3 Bayou La Loutre 8/29 at 5:00AM 55 3.0 Mouth of Pearl River 8/29 at 12:00AM

The gauges are not designed to withstand the eyewall region at landfall, and do not present a complete picture of the surge cycle. Two gauges near Ocean Springs provide the longest surge record in Katrina before failing after 8AM. They show NAVD88 water levels of 3.8 feet at midnight, 6.9 feet at 4AM, 7.7 feet at 5AM, 8.2 feet at 6AM, 10.8 feet by 7AM, and 12.7 feet at 8AM. There is the possibility these may have a 1-ft bias, and this author is investigating. These gauges are about 8 miles east of McIntosh at the entrance of Biloxi Bay, and the surge in west Biloxi Bay will lag behind these gauges and also experience a lower peak.

Since observations are lacking of the entire surge cycle, three methods exist to document the complete storm surge: computer model simulations, post-storm high-water measurements, and eyewitness accounts. A computer model approximates time- dependent hydrodynamic equations which represent water flow driven by wind and pressure fields. It can be used to explore the qualitative evolution of the storm surge, to fill in data gaps, and to explore physical relationships. High water mark surveys are conducted by government agencies (such as the National Weather Service, the Army Corps of Engineers, and the USGS), and private companies such as URS and Haag Engineering. Usually the elevations are recorded relative to vertical datum NAVD88 (NAVD88 is roughly 0.5 feet more than mean sea level). They reflect either the stillwater elevation of the storm surge (areas outside the influence of breaking wave and wave runup, either far inland or inside buildings) or the stillwater elevation plus the wave runup component (areas in the wave swash zone - either breaking waves or wave runup). Stillwater elevation is recovered inside of commercial or residential structures as mud lines on walls or doors. The storm surge plus wave runup high water marks are generally found as debris or trash lines along coastal dunes, sloping terrain of the bay shoreline or the outside perimeter and exterior area of a structure. A high water mark of 18.6 feet is located near McIntosh, but because it’s an outside marker, it will be wave contaminated and the actual surge was near 18.0 feet, with wave action of 1 feet or less superimposed on the surge.

20 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 22 of 40

To assess the timeline of the surge versus wind, the U.S. Army Corps of Engineers ADvanced CIRCulation (ADCIRC) hydrodynamic model is used to simulate Katrina’s storm surge. ADCIRC was initially developed under the Dredging Research Program, a 6-year program funded by the Army Corps of Engineers, Office of the Chief of Engineers. The model was developed as a family of 2- and 3-dimensional finite element based codes with the capability of simulating tidal circulation and storm surge propagation over very large computational domains, while simultaneously providing high-resolution output in areas of complex shoreline and bathymetry. In addition to numerous Army Corps of Engineer applications, ADCIRC has also been used by many universities, including LSU and Notre Dame, and companies such as WorldWinds, Inc., and the URS Corporation. The latter companies have performed work for Louisiana Natural Resources Department for research on the storm surge in Mississippi River Gulf Outlet, storm surge simulations for NASA, and other applications.

The ADCIRC simulation provides a timeline of the surge evolution. East of the hurricane’s onshore winds, the surge can be seen moving up the Pearl River, Jordan River, and Biloxi River at 5AM. Marsh regions near Pearlington and Pascagoula begin to experience inundation. Islands offshore, the Louisiana marsh, as well as Dauphin Island in Alabama, are partially underwater. At 7AM and 9AM, this pattern continues, with surge values increasing along the Mississippi coast.

Partitioning the timing of surge versus wind timing is a crucial piece of information for many cases in this region. Eyewitness accounts and ancillary data suggest significant winds prior to, and during, storm surge inundation. In a 5/25/07 interview, Greg Verdes described his Katrina experience. Mr. Verdes was at his bait shop located at 1022 Legion Lane in Biloxi, which is southeast of the Washington Street drawbridge. He stated, “Everybody’s house was gone before the water came up……the wind came through and destroyed everything. The water came up to the higher level, from say thirteen to twenty feet, and it washed everything away.” Verdes described the destruction like “a domino effect” and the wind sounding like “twenty freight trains.” Jackson County Civil Defense director Butch Loper also described intense winds on the early morning of August 29 during court testimony, reporting a wind gust of 137 mph between 8:00AM and 8:30AM. When pressed on how he remembered the wind and the time, he stated that he instinctively checked when the roof blew off.

Most relevant to the McIntosh residence are the testimonies of Jo Ann Church and Michael Raymond Church, who stayed at their residence during Katrina. Both individually stated strong winds starting about 7AM (when a large limb fell in their yard and another tree had snapped), and stated the wind conditions “deteriorated quickly” afterwards. They both also noted that the water reached the street (elevated at about 14 feet) between 10:30AM and 11AM.

To elucidate the timing of wind and surge, data was output from ADCIRC for the grid node closest to 2558 S. Shore Drive. The land around S. Shore Drive is on a steep peninsula with an elevation of 12-14 feet NAVD88, and the McIntosh house slab is elevated 1-2 feet above the street. An analysis of the ADCIRC output shows inundation

21 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 23 of 40

of the land around 10AM, the road between 10:30AM and 11 AM, and peaking at 18 feet around noon. Animations show the noon peak is due to the wind shift converging water at the peninsula in west Biloxi Bay while east Biloxi Bay slowly drains. The water flow was from the southeast at 10AM, flowed from the south between 11AM and 1PM, then switches to the NNE as the surge retreated. Based on all available data and information, a time series of the surge is shown in Table 4.

The Interagency Performance Evaluation Task Force (IPET), a distinguished group of government, academic, scientists and engineers who studied Katrina peer-reviewed by the American Society of Civil Engineers, has reached the same conclusions regarding the timing of wind and surge along the Mississippi coast. They also conducted a storm surge simulation using ADCIRC. Their simulations show inconsequential surge elevations for any property below 13 feet on the Mississippi coast before 9AM, and the main surge component arriving between 10AM and 11AM. They also note that surge is a combination of wind-driven water south of Mississippi, as well water piled up against the Mississippi River levee system that moves northward with the storm. The following quotes appear in the IPET report:

• At 8AM, “The surge that built up against the lower Mississippi River levees is rapidly propagating in a northeasterly direction towards the Chandeleur Sound.” • At 9AM, “The surge originating along the levees of lower Plaquemines continues to propagate across Chandeleur Sound towards the Mississippi Sound in a northeasterly direction.” • At 10AM, “The surge that propagated from Southern Plaquemines Parish has now combined with the local surge being generated by the strong southerly winds and is dramatically increasing water levels between Bay St. Louis and Biloxi.” • At 11AM, “Surge along the State of Mississippi shoreline is spreading inland and continues to build up driven by the winds from the south reaching 29 feet.” • At 12PM, “Surge continues to propagate inland along the State of Mississippi shore.” • At 3PM, “Katrina has moved well inland. Surge along the State of Mississippi coast is subsiding.”

22 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 24 of 40

These reports are available at https://ipet.wes.army.mil. The surge is discussed in Volume IV.

Figure 12. ADCIRC simulation of Katrina’s storm surge from 5AM to 11AM. These values are relative to sea level. Subtract the structures elevation to compute the water inundation in a building.

23 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 25 of 40

Table 4. Summary of sustained winds, wind gusts, and storm surge for August 29, 2005 at 2558 South Shore Drive. Wave action less than 1 feet will be superimposed on the surge. Wind gusts of 100 mph likely began about 6:00AM. The street elevation is 14 feet. The structure was on a 1-2 foot slab.

Storm surge in Wind Biloxi Bay Time Sustained gusts relative to Storm surge relative Storm surge relative (Aug. 29) wind (mph) (mph) NAVD88 (feet) to street (feet) to slab (feet) 40 (east- 1:00AM northeast) 50 4 Street dry House dry 50 (east- 4:00AM northeast) 65 5 Street dry House dry 60 (east- 5:30AM northeast) 80 6 Street dry House dry

6:30AM 75 (east) 105 7 Street dry House dry

7:00AM 80 (east) 110 8 Street dry House dry

8:30AM 85 (east) 115 9 Street dry House dry 100 (east- 9:30AM southeast) 120 11 Street dry House dry 90 mph (east- Peninsula almost 10:00AM southeast) 110 12 entirely inundated House dry Street inundation 85 mph begins within 30 10:30AM (southeast) 100 14 minutes House dry 75 mph 11:30AM (southeast) 90 17 3 1-2

65 (south- 12:00PM southeast) 80 18 4 2-3

1:00PM 60 (south) 70 17 3 1-2 50 (south- 4:00PM southwest) 60 11 Street dry House dry

24 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 26 of 40

3. Conclusions

The following conclusions can be stated about Hurricane Katrina’s impact on 2558 S. Shore Drive on August 29, 2005:

• Tide gauges show tropical-storm force winds arrived several hours before significant flooding from surge.

• Computer models, eyewitness accounts, newspaper reports, and video show hurricane-force winds, tropical storm-force winds, and strong wind gusts occurred hours before the surge impacted S. Shore Drive. The Hurricane Research Division wind analysis HWINDS concurs with this assessment.

• The surge reached street level (14 feet) on S. Shore Drive between 10:30AM and 11AM, peaking at noon at 18 feet with waves superimposed at 1 foot or less. The structure is on a 1-2 foot slab, and experienced inundation of 2-3 feet between 11:30AM and 1PM. Tropical storm-force winds began at 1AM, and hurricane- force winds started right after 6AM. Peak sustained winds were 100 mph between 9 and 10AM.

• Wind gusts were 20-50% higher than the sustained winds. 100-mph wind gusts began around 6AM. Wind gusts peaked at 120 mph between 9 and 10AM.

• In addition, radar indicates several tornadoes in the vicinity of 2558 S. Shore Drive. Mesovortices, known to occur in Category 3 hurricanes or stronger, are also likely. Downburst activity is a certainty.

This report is based on current data, and subject to modifications from new information.

Report prepared by Dr. Pat Fitzpatrick: Digitally signed by Pat Fitzpatrick Pat Fitzpatrick DN: cn=Pat Fitzpatrick, c=US ______Date: 2010.01.29 11:45:16 -06'00'

25 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 27 of 40

Fee schedule for Pat Fitzpatrick

The fee for testimony and hourly work (including time for reports and meetings) is $200/hour.

Any travel over 50 miles round trip requires the government compensation mileage rate of 0.50/mile. Miles are calculated from mapquest.com from my address at 180 Lakeview Drive, Slidell, LA. Reimbursement for any meals is also required. For fairness, a maximum daily rate is used as established by the State of Mississippi; these rates vary by state, city, and time of year, but range from $36 to $46/day. The high cost meal allowance is calculated from: http://www.travel.msstate.edu/hcma/plhcma.php . Any lodging also requires reimbursement. In most cases, morning depositions more than 50 miles from my residence will require overnight lodging. Reimbursement for any data purchases is required.

Depositions need to be paid the day of testimony. Depositions fees are $200/hr for the first 3 hours. After the first 3 hours, the rate is $400/hr. In my experience, there is no reason for depositions to be longer than 3 hours for meteorology questions, and I also respectfully ask that any deposition not be an all-day affair.

Subpoena request for data costs $400, due in advance before the data is provided. Requests for data need to be made 6 weeks in advance of due date. No more than 2 datasets may be requested during this 6-week period per law firm.

Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 28 of 40

DR. PAT FITZPATRICK

PRIOR TESTIMONY

Depositions

2007 (1) Webster vs. USAA U.S.D.C. Southern District of Mississippi 1:05-cv-715

2007 (2) The Mississippi Division of the United Sons of Confederate Veterans, et al. vs. Charter Oak Fire Insurance Company, et al. U.S.D.C. Southern District of Mississippi 1:06-cv-731

2007 (3) Winters vs. State Farm U.S.D.C. Southern District of Mississippi 1:06-cv-00393

2007 (4) Norman Hasik and Donna Hasik v. State Farm Insurance Company U.S.D.C. Civil Action No. 06-6330, Sec. J, Mag. 1

2007 (5) McIntosh vs. State Farm U.S.D.C. Southern District of Mississippi 1:06-cv-1080

2007 (6) Maxus Realty Trust, Inc. v. RSUI Indemnity Company U.S.D.C. Western District of Missouri 06-0750-CV-W-ODS

2007 (7) Lisanby vs. USAA Circuit Court of Jackson County Mississippi 2007-00, 117 (3)

2008 (8) Lynn Patrick and Thomas Krafft v. State Farm Fire and Casualty Co. Circuit Court of Harrison County Mississippi, Civil Action No.: A2401-2006-140

2008 (9) Tosh vs. State Farm (court information unknown)

2008 (10) Wagner vs. Metropolitan Property and Casualty Insurance Company; Economy Premier Assurance Company; Rimkus Consulting Group, Inc; and James R. Kreimborg. U.S.D.C. Southern District of Mississippi 1:07-cv-969-LTS-JMR

2008 (11) Penthouse Owners Association, Inc vs. Certain Underwriters at Lloyd’s, London. U.S.D.C. Southern District of Mississippi 1:07-cv-568-LTS-RHW

2008 (12) Derose vs. Nationwide. U.S.D.C. CA No 1:07-cv-684-LTS-RHW

2008 (13) Abel vs. Nationwide Mutual Fire Insurance Company Circuit Court of Jackson County Mississippi 2006-00248 (1)

2008 (14) Bernard Smith vs. State Farm Insurance (court information unknown) Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 29 of 40

2008 (15) Phillip Remel v. State Farm Fire & Casualty Company U.S.D.C. Southern District of Mississippi 1:07-cv-126-LTS-RHW

2008 (16) Calvin Zar v. State Farm Fire & Casualty Company U.S.D.C. Southern District of Mississippi 1:07-cv-133-LTS-RHW

2008 (17) Michael Mokay v. State Farm Fire & Casualty Company Circuit Court of Hancock County 07-0390

2008 (18) Payment case (court information unknown)

2008 (19) Bishop v. State Farm Fire & Casualty Company

2008 (20) Campbell v. State Farm Fire & Casualty Company U.S.D.C. Southern District of Mississippi 1:07-cv-395-LTS-RHW

2008 (21) Herbert J. Ashe v. Citizens Property Insurance Corporation Case No. 2005 CA 000482

2008 (22) Bishop v. State Farm Fire & Casualty Company

2009 (23) Anthony Catania & Lakeshore Body Shop v. Universal Underwriters’ Group Circuit Court of Hancock County

2009 (24) Eleuterius v. Nationwide Mutual Fire Insurance Company U.S.D.C. Southern District of Mississippi 1:08-cv-342-LTS-RHW

2009 (25) David v. Nationwide Mutual Fire Insurance Company U.S.D.C. Southern District of Mississippi 1:08-cv-349-LTS-RHW

2009 (26) Frank Anthony deposition (court information unknown)

2009 (27) Nilson v. Nationwide Mutual Fire Insurance Company U.S.D.C. Southern District of Mississippi 1:07-cv-00990-LTS-MTP

2009 (28) Harold & Dolores Jackson v. Citizens Property Insurance Corporation Circuit Court for Escambia County, FL, Case No. 2005CA000298

Court testimony

2008 (1) Bishop v. State Farm Fire & Casualty Company Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 30 of 40

Fitzpatrick’s Publications

Many of these publications are available online at www.drfitz.net .

Peer-review Publications:

Pielke, R. A., L. R. Bernardet, P. J. Fitzpatrick. S. C. Gillies, R. F. Hertenstein, A. S. Jones, X. Lin, J. E. Nachamkin, U. S. Nair, J. M. Papineau, G. S. Poulos, M. H. Savoie, and P. L. Vidale, 1995. Standardized test to evaluate numerical weather prediction algorithms. Bull. Amer. Meteor. Soc., 76, 46-48.

Fitzpatrick, P. J., J. A. Knaff, C. W. Landsea, and S. V. Finley, 1995: Documentation of a systematic bias in the Aviation model’s forecast of the Atlantic Tropical Upper-Tropospheric Trough: Implications for tropical cyclone forecasting. Wea. Forecasting, 10, 433-446.

Fitzpatrick, P. J., 1997: Understanding and forecasting tropical cyclone intensity change with the Typhoon Intensity Prediction Scheme (TIPS). Wea. Forecasting, 12, 826-846.

Mostovoi, G. V., P. J. Fitzpatrick, and Y. Li, 2006: Regional accuracy of QuikSCAT gridded winds. Int. J. Remote Sensing, 26, 4117-4136.

Zhang, X., Q. Xiao, and P. J. Fitzpatrick, 2007: The impact of multi-satellite data on the initialization and simulation of Hurricane Lili's (2002) rapid weakening phase. Mon. Wea. Rev., 135, 526-548.

Eamon, C. D., P. J. Fitzpatrick, and D. D. Truax, 2007: Observations of structural damage caused by Hurricane Katrina on the Mississippi Gulf Coast. J. Perf. Constr. Fac., 21, 117-127.

Steed, C. A., P. J. Fitzpatrick, T. J. Jankun-Kelly, A. N. Yancey, and J. E. Swan, 2008. North Atlantic hurricane trend analysis using parallel coordinates and statistical techniques. Fifth International Conference on Geographic Information Science, Geospatial Visual Analytics Workshop, Sep. 23-26, 2008, Park City, UT. IEEE Computer Society.

Xiao, Q., X. Zhang, C. Davis, J. Tuttle, G. Holland, and P. J. Fitzpatrick, 2009: Experiments of hurricane initialization with airborne Doppler radar data for the Advanced Hurricane WRF (AHW) model. Mon.Wea. Rev., 137, 2758-2777.

Steed, C. A., P. J. Fitzpatrick, T. J. Jankun-Kelly , A. N. Yancey, and J. E. Swan, 2009. An interactive parallel coordinates technique applied to a tropical cyclone climate analysis. Computers and Geosciences, 35, 1529-1539.

Steed, C. A., P. J. Fitzpatrick, T. J. Jankun-Kelly, A. N. Yancey, and J. E. Swan, 2009. Tropical cyclone trend analysis using enhanced parallel coordinates and statistical analysis. Cartography and Geographic Information Systems, 36, 251-265.

Steed, C. A., J. E. Swan, T. J. Jankun-Kelly, and P. J. Fitzpatrick, 2009. Guided analysis of hurricane trends using statistical processes integrated with interactive parallel coordinates. In IEEE Symposium on Visual Analytics Science and Technology 2009, Oct. 11-13, Atlantic City, NJ. Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 31 of 40

Karan, H., P. J. Fitzpatrick, C. M. Hill, Y. Li, Q. Xiao, and E. Lim, 2010: The formation of two prefrontal squall lines, and the impact of WSR-88D radial winds in a WRF simulation. Wea. Forecasting, 25, 242-262.

Sanyal, J., P. Amburn, K. Wu, S. Zhang, J. Dyer, P. J. Fitzpatrick, and R. J. Moorhead, 2010. On immersive virtual environments facilitating hurricane modelling and analysis. Submitted to Computers and Geosciences.

Fitzpatrick, P. J., N. Tran, Y. Li, and C. M. Hill, 2010. Quantifying the basic factors of local storm surge potential. Submitted to Nat. Hazards Earth Syst. Sci.

Fitzpatrick, P. J., Y. Lau, N. Tran, Y. Li, and C. M. Hill, 2010. A storm surge scale for evacuation planning. Submitted to J. Transportation Safety & Security.

Hill, C. M., P. J. Fitzpatrick, J. Corbin, Y. Lau, and S. Bhate, 2010: Summertime precipitation regimes associated with the sea breeze and land breeze circulations in southern Mississippi and eastern Louisiana. Accepted pending revisions to Wea. Forecasting.

Books:

Fitzpatrick, P. J., 1999: Natural Disasters: Hurricanes. ABC-CLIO. 288 pp.

Fitzpatrick, P. J., 2005: Hurricanes: A Reference Handbook. ABC-CLIO. 412 pp.

M.S. Thesis and Ph.D. Dissertation:

Fitzpatrick, P. J., 1995. Understanding and forecasting tropical cyclone intensity change. Ph.D. dissertation. Colorado State University, 380 pp.

Fitzpatrick, P. J., 1992. A numerical study of mesoscale convection in a rotating tropical atmosphere. M. S. thesis, Texas A&M University, 117 pp.

Encyclopedias:

Fitzpatrick, P. J., 2010: The debate over tropical cyclones and anthropogenic climate change. World History Encyclopedia. A. J. Andrea, Ed., ABC-CLIO, accepted.

Fitzpatrick, P. J., 2010: Global warming. World History Encyclopedia. A. J. Andrea, Ed., ABC- CLIO, accepted.

Technical reports:

White, T. D., B. McAnally, D. Truax, H. Cole, C. Eamon, L. Zhang, P. Gullett, P. Fitzpatrick, Y. Lau, S. Bhate, Y. Li, 2006. Coast in the eye of the storm: Hurricane Katrina, August 29, 2005. Mississippi State University. 86 pp. GRI #6001.

Fitzpatrick, P., G. Mostovoi, and Y. Li, 2002. Validation and comparison of a mesoscale model and NOGAPS in the Bay of Bengal and Arabian Sea for March-August 2001. MSU technical report submitted to the US Naval Oceanographic Office, MSSU-COE-ERC-02-11, 88 pp. Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 32 of 40

Herndon, D., and E. Valenti, 2002. High-resolution atmospheric model comparison over Navy operational areas in CONUS. WorldWinds’ technical report submitted to the Central Navy Meteorology and Oceanography Command, 46 pp.

Fitzpatrick, P., G. Mostovoi, Y. Li, M. Bettencourt, and S. Sajjadi, 2002. Coupling of COAMPS and WAVEWATCH with improved wave physics. Final Project Report, Contract No. N62306- 01-D-7110, 10 pp.

Fitzpatrick, P. J., 1996. Understanding and forecasting tropical cyclone intensity change. CSU bluebook technical report number 598. 346 pp.

Valenti, E., P. J. Fitzpatrick, and Y. Li, 2005. Web-Based hurricane storm surge and flood forecasting using optimized IFSAR bald earth DEMs. SBIR Phase II Final Report, WorldWinds, Inc.

Steed, C. A., P. J. Fitzpatrick, T.J. Jankun-Kelly, A. N. Yancey, and J. E. Swan II, 2008. An interactive parallel coordinates technique applied to a tropical cyclone climate analysis, Naval Research Laboratory, Stennis Space Center, MS, NRL/MR/7440--08-0126, 25 pp.

Steed, C. A., P. J. Fitzpatrick, T.J. Jankun-Kelly, and J. E. Swan II, 2008. Visual analysis of North Atlantic hurricane trends using parallel coordinates and statistical techniques, Naval Research Laboratory, Stennis Space Center, MS, NRL/MR/7440--08-9130, 18 pp.

Conference papers and posters:

Fitzpatrick, P. J., N. Tran, Y. Li, and C. M. Hill, 2009. A proposed new storm surge scale. Oceans ’09 MTS/IEEE, Biloxi, MS, October 26-29.

Hill, C. M., P. J. Fitzpatrick, and Y. Lau, 2009. Examination of the tropical cyclone environment through comparison of COSMIC with other satellite data. European Geosciences Union General Assembly, Vienna, Austria, April 19-24.

Karan, H., P. J. Fitzpatrick, C. M. Hill, Y. Li, Q. Xiao, E. Lim, and J. Sun, 2009. The formation of multiple prefrontal squall lines and the impact of NEXRAD radial winds in a WRF simulation. European Geosciences Union General Assembly, Vienna, Austria, April 19-24.

Hill, C. M., P. J. Fitzpatrick, X. Fan, V. Anantharaj, M. Masutani, L. P. Riishojgaard, and Y. Li, 2009. An observing system simulation experiment to evaluate CrIS / ATMS observations in modeling a mesoscale weather event. 89th Annual Meeting of the American Meteorological Society, January 20-24, Phoenix, AZ.

Riishojgaard, L. P., and F. Weng, M. Masutani, T. Zhu, H. Sun, C. M. Hill, V. Anantharaj, P. J. Fitzpatrick, R. M. Errico, S. J. Lord, Y. Han, J. Woollen, D. Groff, and T. J. Kleespies, 2009. Evaluation of GOES-R and NPOESS instrument in Joint OSSEs. 89th Annual Meeting of the American Meteorological Society, January 20-24, Phoenix, AZ.

Fitzpatrick, P. J., 2009: Hurricanes and climate: The debate over tropical cyclones and anthropogenic climate change. MMS Gulf of Mexico Information Transfer Meeting, January 6-8, New Orleans, LA.

Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 33 of 40

Fitzpatrick, P. J., 2009: Unmanned aircraft hurricane reconnaissance. MMS Gulf of Mexico Information Transfer Meeting, January 6-8, New Orleans, LA.

Fitzpatrick, P. J., C. M. Hill, N. Tran, Y. Lau, Y. Li, and H. Karan, 2008. The impact of Louisiana’s levees and wetlands on Katrina’s storm surge. 2nd Annual Northern Gulf Institute Conference, May 13-14, Biloxi, MS.

Sanyal, J., P. Amburn, S. Zhang, P. J. Fitzpatrick, and R. J. Moorhead, 2008. Visualization of numerical model output of Hurricane Lili’s rapid weakening in a virtual 3D environment. 2nd Annual Northern Gulf Institute Conference, May 13-14, Biloxi, MS.

Sanyal, J., P. Amburn, S. Zhang, P. J. Fitzpatrick, and R. J. Moorhead, 2008. Applying immersive visualization techniques to analyze model outputs: A case study of Hurricane Lili. IEEE Visualization 2008, October 19-24, Columbus, OH.

Sanyal, J. P. Amburn, S. Zhang, J. Dyer, P. J. Fitzpatrick, and R. J. Moorhead, 2008. User experience of hurricane visualization in an immersive 3D environment. 4th International Symposium on Visual Computing, December 1-3, Las Vegas, NV.

Masutani, M., J. S. Woollen, R. Errico, Y. Xie, T. Zhu, H. Sun, J. Terry, R. Yang, S. Greco, N. Prive, E. Andersson, T. W. Schlatter, A. Stoffelen, F. Weng, O. Reale, L. P. Riishojgaard, G. D. Emmitt, S. J. Lord, Z. Toth, G. J. Marseille, V. Anantharaj, K. Fielding, G. McConaughy, S. J. Worley, C. F. Shih, M. Yanaguchi, J. C. Jusem, C. M. Hill, P. J. Fitzpatrick, D. Devenyi, S. S. Weygandt, S. A. Wood, Y. Song, E. Liu, D. N. Groff, M. Hart, G. Gayno, A. Da Silva, M. J. McGill, D. Kleist, Y. Sato, and S. A. Boukabara, 2008. Progress in joint OSSEs – Three joint OSSE nature runs and simulations of observations. 88th Annual Meeting of the American Meteorological Society, January 20-24, New Orleans, LA.

Anantharaj, A., G. Mostovoy, and P. J. Fitzpatrick, 2008. Impact of soil moisture initialization on numerical weather forecasting over the Mississippi Delta region. 88th Annual Meeting of the American Meteorological Society, January 20-24, New Orleans, LA.

Xiao, Q., X. Zhang, C. A. Davis, J. D. Tuttle, G. J. Holland, and P. J. Fitzpatrick, 2008. Impacts of airborne Doppler radar data assimilation on the forecasts of Hurricane Katrina (2005). 88th Annual Meeting of the American Meteorological Society, January 20-24, New Orleans, LA.

Hill, C. M., P. J. Fitzpatrick, H. Karan, and Y. Lau, 2008. An examination of COSMIC In the tropical cyclone environment. 28th Conference on Hurricanes and Tropical Meteorology, April 28-2 May, Orlando, FL.

Fitzpatrick, P. J., C. M. Hill, Y. Lau, Y. Li, J. Corbin, and H. Karan, 2008. A theory on the expansion of Hurricane Katrina’s wind field. 28th Conference on Hurricanes and Tropical Meteorology, April 28-2 May, Orlando, FL.

Fitzpatrick, P. J., Y. Lau, S. K. Bhate, and C. M. Hill, 2007. A comparison of COSMIC temperature and moisture profiles against dropsondes, AIRS, Terra, and Aqua near Atlantic tropical cyclones of 2006. 61st Interdepartmental Hurricane Conference, March, New Orleans, LA.

Lim, E., Q. Xiao, J. Sun, P. J. Fitzpatrick, Y. Li, and J. L. Dyer, 2007. The impact of Doppler radar data on rainfall forecast: a case study of a convective rainband evening in Mississippi Delta Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 34 of 40

using WRF 3D-Var. 18th Conference on Numerical Weather Prediction, June 25-29, Park City, UT.

Fitzpatrick, P. J., N. T. Tran, Y. Lau, C. M. Hill, and W. A. Shaffer, 2007. A sensitivity study of the storm surge from Hurricane Katrina. Annual Northern Gulf Institute Conference, May 16-17, Biloxi, MS.

Majure, L. C., G. N. Ervin, and P. J. Fitzpatrick, 2007. Storm-driven maritime dispersal of prickly pear cacti (Opuntia species). International Cactoblastis cactorum Conference, May 7-10, Phoenix, AZ.

Lim, E., Q. Xiao, J. Sun, P. J. Fitzpatrick, Y. Li, and J. L. Dyer, 2007. The impact of Doppler radar data on rainfall forecast: a case study of a convective rainband event in Mississippi Delta using WRF 3D-Var. 22nd Conference on Weather Analysis and Forecasting, June 25-29, Park City, UT.

Steed, C. A., P. J. Fitzpatrick, T. J. Jankun-Kelly , and J. E. Swan, , 2007. Practical application of parallel coordinates to hurricane trend analysis. IEEE Visualization 2007, October 28-November 1, Sacramento, CA.

Masutani, M., E. Andersson, J. Terry, O. Reale, J. C. Jusem, L. P. Riishojgaard, T. Schlatter, A. Stoffelen, J. Woollen, S. Lord, Z. Toth, Y. Son, D. Kleist, Y. Xie, N. Prive, E. Liu, H. Sun, D. Emmitt, S. Greco, S. A. Wood, G.-J. Marseille, R. Errico, R. Yang, G. McConaughy, D. Devenyi, S. Weygandt, V. Anantharaj, C. Hill, P. Fitzpatrick, F. Weng, T. Zhu, and S. Boukabara, 2007. Joint OSSEs. 18th Conference on Numerical Weather Prediction, June 25-29, Park City, UT.

Anantharaj, V. G., P. J. Fitzpatrick, Y. Li, R. L. King, and E. Johnson, 2006. An Analysis of MODIS Landuse Data on a Gulf Coast Sea Breeze Simulation. Proceeding of the 10th Symposium on IOAS-AOLS, 86th Annual Meeting of the American Meteorological Society. 10IOAS - 8.3.

Fitzpatrick, P. J., Y. Lau, S. Bhate, Y. Li, E. Valenti, B. Jacobsen, and J. Lawhead, 2006. Storm Surge Issues of Hurricane Katrina. 60th Interdepartmental Hurricane Conference, Mobile, AL.

Fitzpatrick, P. J., Y. Lau, S. Bhate, V. Anantharaj, X. Zhang, and Q. Xiao, 2006. The Impact of Multi-Satellite Data in a 4DVAR MM5 Simulation of Hurricane Lili’s (2002) Rapid Weakening. 60th Interdepartmental Hurricane Conference, Mobile, AL.

Fitzpatrick, P. J., X. Zhang, and Q. Xiao, 2006. The Impact of Multi-Satellite Data on the Initialization and Simulation of Hurricane Lili’s (2002) Rapid Weakening Phase. 10th Symposium on Integrated Observing and Assimilation Systems for the Atmosphere, Oceans, and Land Surface, American Meteorological Society.

Fitzpatrick, P. J., X. Zhang, Q. Xiao, N. Tran, S. Bhate, and Y. Lau, 2006. The impact of Terra, Aqua, TRMM, AVHRR, and dropsonde data on Hurricane Lili simulations. 10th Symposium on Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 35 of 40

Integrated Observing and Assimilation Systems for the Atmosphere, Oceans, and Land Surface, American Meteorological Society.

Fitzpatrick, P. J., E. Valenti, and Y. Li, 2006. XM Satellite Marine Weather - The successful result of Stennis Space Center’s technology cluster. NOAA Symposium on the Public/Private Sector Partnership, American Meteorological Society.

Anantharaj, V. G., P. J. Fitzpatrick, Y. Li, S. Bhate, and Y. Lau , 2005. Incorporation of MODIS Land Data into Mesoscale Weather Prediction Models. Enterprise for Innovative Geospatial Solutions (EIGS) Research Symposium, Stennis Space Center, MS.

Anantharaj, V. G., R. L. King, Y. Li, and P. J. Fitzpatrick, 2005. Influences of MODIS land use data on high resolution numerical weather. EIGS Research Symposium, Stennis Space Center, MS.

Fitzpatrick, P. J., X. Zhang, Q. Xiao, N. Tran, S. Bhate, and Y. Lau , 2005. Applications of Four Dimensional Analysis to Examine Intensity Variations of Hurricane Lili using Aqua, Terra, and other Satellite Data. Enterprise for Innovative Geospatial Solutions (EIGS) Research Symposium, Stennis Space Center, MS.

Zhang, X., Q. Xiao, P. J. Fitzpatrick, and Y. Kuo, 2005. Four-dimensional Variational Data Assimilation of Satellite Retrieval Data for Hurricane Lili (2002). 6th WRF/15th MM5 Users’ Workshop.

Mostovoi, G. V., and P. Fitzpatrick, 2004. Numerical Modeling of Rising Thermal: Application for Cumulus Clouds Parameterization. Proceedings of the 14th International Conference on Clouds and Precipitations, Bologna, Italy. Vol. 2. 1386-1388.

Fitzpatrick, P. J., A. Mehra, T. Haupt, J. Corbin, and other authors, 2002. The Distributed Marine Environment Forecast System (DMEFS). Preprints, 2002 AMS Conference, Orlando, FL.

Mostovoi, G. V., P. J. Fitzpatrick, and other authors, 2002. The Effect of Averaging Surface Roughness Values on Coastal Winds for Different Mesoscale Model Resolutions. Preprints, 2002 AMS Conference, Orlando, FL.

Sajjadi, S. G., M. Bettencourt, P. J. Fitzpatrick, and G. Mostovoi, 2002. Sensitivity of a coupled tropical cyclone/ocean wave simulation to different energy transfer schemes. Tropical Meteorology and Hurricane Conference.

Veeramony, J., P. J. Fitzpatrick, D. Dmitry, D. Herndon, N. Tran, and E. Valenti, 2002. The Incorporation of SAR Interferometry Elevation Data into an ADCIRC Simulation of Hurricane Camille. Tropical Meteorology and Hurricane Conference, San Diego, CA.

Mostovoi, G. V., P. J. Fitzpatrick, P. J., and Y. Li, 2001. Comparison of 9-km Wind Forecasts Versus 27-km Wind Forecasts During the Northern Gulf of Mexico Littoral Initiative. Preprints, Ninth Conference on Mesoscale Processes, Ft. Lauderdale, FL.

Fitzpatrick, P. J., S. Mehta, V. Budamgunta, U. Reddy, R. Mahecha, N. W. Scheffner, and G. Easson, 1999. Operational Modeling and Validation of the Hurricane Storm Surge Using ADCIRC, 1999 AMS Conference.

Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 36 of 40

Fitzpatrick, P. J., N. Tran, Y. Li, S. Mehta, and E. Valenti, 1999. Validation Study of Operational MM5 Forecasts on the Gulf Coast Using FDDA in a Commercial Environment. Preprints, Third Conference on Coastal Atmospheric and Oceanographic Prediction and Processes, New Orleans, LA.

Scheffner, N. W., and P.J. Fitzpatrick, 1998. Real-Time Predictions of Surge Propagation. 5th International Conference on Estuarine and Coastal Modeling.

Blumberg, A., J. Blaha, P. Fitzpatrick, and others, 2001. The Northern Gulf of Mexico Littoral Initiative, 2001. International Conference on Estuarine and Coastal Modeling Conference, Ft. Lauderdale, FL.

Fitzpatrick, P. J., S. Mehta, V. Budamgunta, R. Mahecha, Y. Li, and D. Zhang, 1999. Analysis and comparison of a tropical cyclone boundary layer database against model simulations. Preprints, 1999 AMS Conference, Dallas, TX.

Fitzpatrick, P. J., S. Mehta, Y. Li, J. K. Lee, E. Valenti, and E. Barker, 1998. Improved MM5 and QLM forecasts using scatterometer data. Preprints, 1999 AMS Conference, Dallas, TX.

Fitzpatrick, P. J., 1995: Forecasting tropical cyclone intensity change in the West Pacific. Proceedings of the 21st Conference on Hurricanes and Tropical Meteorology, Miami, FL, April 22-28, AMS Conference, 94-96.

Fitzpatrick, P. J., 1993: Some observational and theoretical processes related to rapidly intensifying tropical cyclones. Preprint, Twentieth Conference on Hurricanes and Tropical Meteorology, San Antonio, TX, May 10-14, 516-519.

Finley, S. V., P. J. Fitzpatrick, and others, 1995: A systematic bias in the Aviation's forecast of the Atlantic TUTT: Implications for tropical cyclone forecasting. Proceedings of the 21st Conference on Hurricanes and Tropical Meteorology, Miami, FL, April 22-28, AMS Conference.

Other publications:

Fitzpatrick, P. J., Y. Li, and G. Mostovoi, 2002. COAMPS High-Resolution Weather Forecasts for Mississippi and Louisiana Coasts. NAVOCEANO MSRC Navigator magazine, 5-6,21. Also: http://www.navo.hpc.mil/Navigator/fall02 Feature1.html

Bettencourt, M. T., S. G. Sajjadi, and P. Fitzpatrick, 2002. A Distributed Model Coupling Environment for Geophysical Processes. NAVOCEANO MSRC Navigator magazine, 7-9. Also: http://www.navo.hpc.mil/Navigator/fall02 Feature2.html

URS Corporation, 2006. The Direct Impact of the Mississippi River Gulf Outlet on Hurricane Storm Surge. CONTRACT NO. 2503-05-3, Prepared for State of Louisiana Department of Natural Resources. (Consultant participation through WorldWinds, Inc.)

Lau, Y., S. K. Bhate, and P. J. Fitzpatrick. 2008. Visual data analysis for satellites. NASA Tech Briefs, 32, 18,20.

Anantharaj, V, and P. J. Fitzpatrick. 2008. MODIS - Atmospheric data handler. NASA Tech Briefs, 32(12), 26-27.

Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 37 of 40

Tran, N., and P. J. Fitzpatrick. 2008. Data assimilation cycling for weather analysis. NASA Tech Briefs, 32, 65.

Anantharaj, V., Y. Li, and P. J. Fitzpatrick. 2009. Incorporation of MODIS land cover data into COAMPS. NASA Tech Briefs, pending publication.

Fitzpatrick, P. J., N. Tran, and Y. Li. 2007. Handling input and output for COAMPS. NASA Tech Briefs, 31, 45.

Valenti, E., and P. J. Fitzpatrick. 2006. Forecasting of storm-surge floods using ADCIRC and optimized DEMs. NASA Tech Briefs, 30, 28-30.

DVD:

The Winds of Katrina, 2008, Bill Hudson & Associates.

Oral Presentations as author or co-author (list available upon request):

2009 – 16 presentations 2008 – 14 presentations 2007 – 17 presentations Other years – Typically 10-20 annually

Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 38 of 40

fitz_vita_2010_hurr Patrick J. Fitzpatrick Geosystems Research Institute, Mississippi State University BLDG 1103, Room 233 Stennis Space Center, MS 39529 Work: (228) 688-1157, Fax: (228) 688-7100, Cell: (985) 788-9486 Email: [email protected] Website: www.drfitz.net College: Texas A&M University, Meteorology, B. S., 1988 Texas A&M University, Meteorology, M.S., 1992. Advisor: Dr. Dusan Djuric Colorado State University, Meteorology, Ph. D., 1995. Advisor: Dr. Bill Gray

Employment:

05/2001-Present: Associate Research Professor, Mississippi State University

01/1998-01/2006: Chief Scientist, WorldWinds, Inc.

01/1996 - 05/2001: Assistant Professor, Jackson State University

Summer 1996: Visiting Researcher, Department of Atmospheric Science, Colorado State University

07/1994 - 11/94: Hurricane Forecaster, Shell Oil Company

01/1992-12/1995: Graduate Assistant, Department of Atmospheric Science, Colorado State University

09/1989-12/1991: Graduate Instructor, Department of Meteorology, Texas A&M University

09/88-08/89: Graduate Assistant, Department of Meteorology, Texas A&M University

Activities: 05/2001-Present: Mississippi State University activities include: hurricane intensity research; hurricane modeling; storm surge; development of operational forecasting capabilities; validation studies of forecasts; development of diverse data assimilation capabilities (Multivariate Optimum Interpolation, 4DVAR, FDDA) and quality control algorithms; coupling weather models and ocean wave models; development of new ocean wave growth algorithms; development of web portal project for models called the Distributed Marine Environment Forecast System; Observing System Simulation Experiments; Visualization Methods. 05/2006-07/2009: Expert testimony and reports in legal cases involving Hurricanes Katrina and Ivan.

01/1998-01/2006: Leads R&D effort for WorldWinds, Inc. with support provided by NASA's Mississippi Space Commerce Initiative (MSCI), a Page 1 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 39 of 40

fitz_vita_2010_hurr partnership between Stennis companies. WorldWinds provides weather data and performs operational runs of numerical weather models (COAMPS and MM5) on the company's Linux cluster. These forecasts and data are transmitted by XM Radio to marine interests. WorldWinds is also developing storm surge software using the ADCIRC model with a new elevation database based on IFSAR data. 05/1999-Present: Volunteer hurricane consultant, Mississippi Emergency Management Agency (MEMA). Participates as advisor during hurricane threats. 05/1999-Present: Author of hurricane book: "Hurricanes: A Reference Handbook." This book contains chapters on: hurricane structure and forecasting; a chronology of historic events and achievements in hurricane science; biographies of influential hurricane researchers; a description of important organizations (i.e., the National Hurricane Center, the National Weather Service, the American Red Cross, etc); tables, letters, government documents, mitigation issues; a comprehensive list and description of hurricane publications; electronic sources for hurricane information; and a glossary. The second edition was released in November 2005.

01/1996-05/2001: Assistant Professor, Jackson State University. Taught most undergraduate meteorology classes, undergraduate general science classes on technology, and a UNIX system administration class. Conducted mesoscale modeling and hurricane research. Upgraded facilities with a state-of-the-art computer lab.

Some Publications Relevant to Hurricanes:

Fitzpatrick, P.J., 2005: Hurricanes: A Reference Handbook. ABC-CLIO. 412 pp.

Fitzpatrick, P. J., 1999: Natural Disasters: Hurricanes. ABC-CLIO. 288 pp.

Sanyal, J., P. Amburn, K. Wu, S. Zhang, J. Dyer, P. J. Fitzpatrick, and R. J. Moorhead, 2009. On immersive virtual environments facilitating hurricane modelling and analysis. Submitted to Computers and Geosciences. Steed, C. A., P. J. Fitzpatrick, T. J. Jankun-Kelly , A. N. Yancey, and J. E. Swan, 2009. Tropical cyclone trend analysis using enhanced parallel coordinates and statistical analytics. Cartographic and Geographic Information Science, 36, 251-265. Xiao, Q., X. Zhang, C. Davis, J. Tuttle, G. Holland, and P. J. Fitzpatrick, 2009. Experiments of hurricane initialization with airborne Doppler Radar data for the Advanced-research Hurricane WRF (AHW) model. To be published in Monthly Weather Review. Zhang, X., Q. Xiao, and P. J. Fitzpatrick, 2007: The Impact of multi-satellite data on the initialization and simulation of Hurricane Lili's (2002) rapid weakening phase. Mon. Wea. Rev., 135, 526-548.

Valenti, E., and P. J. Fitzpatrick. 2006. Forecasting of storm-surge floods using ADCIRC and optimized DEMs. NASA Tech Brief., 30, 28-30. Page 2 Case 1:06-cv-00433-LTS-RHW Document 409-6 Filed 01/29/2010 Page 40 of 40

fitz_vita_2010_hurr Eamon, C. D., P. Fitzpatrick, and D. D. Truax, 2007: Observations of structural damage caused by Hurricane Katrina on the Mississippi Gulf Coast. J. Perf. Constr. Fac., 21, 117-127. White, T. D., B. McAnally, D. Truax, H. Cole, C. Eamon, L. Zhang, P. Gullett, P. Fitzpatrick, Y. Lau, S. Bhate, Y. Li. 2006. Coast in the Eye of the Storm: Hurricane Katrina, August 29, 2005. Mississippi State University technical report. 86 pp. Fitzpatrick, P. J., 1997: Understanding and forecasting tropical cyclone intensity change with the Typhoon Intensity Prediction Scheme (TIPS). Wea. Forecast, 12, 826-846. Fitzpatrick, P. J., and others, 1995: Documentation of a systematic bias in the Aviation model's forecast of the Atlantic Tropical Upper-Tropospheric Trough: Implications for tropical cyclone forecasting. Wea. Forecast, 10, 433-446.

Sajjadi, S. G., M. Bettencourt, P. J. Fitzpatrick, and G. Mostovoi, 2002: Sensitivity of a coupled tropical cyclone/ocean wave simulation to different energy transfer schemes. Tropical Meteorology and Hurricane Conference, San Diego, CA

Veeramony, J., P. J. Fitzpatrick, D. Netchaev, D. Herndon, N. Tran, and E. Valenti, 2002. Assimilating IFSAR interferometry DEM data into an ADCIRC simulation of Hurricane Camille. Preprints, 2002 Tropical Meteorology and Hurricane Conference, San Diego, CA.

Fitzpatrick, P. J., Y. Li, and G. Mostovoi, 2002. COAMPS High-Resolution Weather Forecasts for Mississippi and Louisiana Coasts. NAVOCEANO MSRC Navigator magazine, 5-6,21. Also: [2]http://www.navo.hpc.mil/Navigator/fall02_Feature1.html

Bettencourt, M. T., S. G. Sajjadi, and P. Fitzpatrick, 2002. A Distributed Model Coupling Environment for Geophysical Processes. NAVOCEANO MSRC Navigator magazine, 7-9. Also: [3]http://www.navo.hpc.mil/Navigator/fall02_Feature2.html

Fitzpatrick, P. J., S. Mehta, V. Budamgunta, R. Mahecha, Y. Li, and D. Zhang, 1999: Analysis and comparison of a tropical cyclone boundary layer database against model simulations. Preprints, 1999 AMS Conference, Dallas, TX.

Fitzpatrick, P. J., 1995. Understanding and forecasting tropical cyclone intensity change. Ph.D. dissertation. Colorado State University, 380 pp. Fitzpatrick, P. J., 1992. A numerical study of mesoscale convection in a rotating tropical atmosphere. M. S. thesis, Texas A&M University, 117 pp.

Page 3