LRC Services

236 Red Oak Lane

Defuniak SpringsFL32433

850-830-8656

Bridwell

2898 Tropicana Lane

Crystal Beach Texas

View of 2898 Tropicana Lane on May 12, 2009

(Please note the steel bar and piling in background oriented East to West)

This is a post-storm report on the effects of Hurricane Ike on 2898 Tropicana Lane, Crystal Beach Texas for the period of September 12 through September 13, 2008.

The sources of data are as follows:

  1. National Weather Service
  2. NationalHurricaneCenter
  3. National Climatic DataCenter
  4. Internet from various websites (i.e. TexasTechUniversity)
  5. Photographs taken during my site visit

2898 Tropicana Lane prior to Hurricane Ike

It is my professional opinion the initial and primary cause of damage at 2898 Tropicana Lane, Crystal Beach Texas was caused by the following:

  1. Extremely high wind speeds several hours prior to the storm surge
  2. Microbursts and associated turbulence
  3. Tornadic activity
  4. Flying debris

High level view of 2898 Tropicana Lane

The site visit was conducted on May 12, 2009. The site has been partially cleared and bulldozed.

The National Hurricane Center has established the Bolivar peninsula near the landfall point for Hurricane Ike with winds of 110 miles per hour at landfall at 2:00AM on the morning of September 13, 2008. The Bridwell property was is in the right-front quadrant of the hurricane.

Another source of wind data comes from a commercial firm called “Weatherflow”. This company maintains a network of wind measuring equipment throughout the Houston area. Based on the data collected by Weatherflow, there were only 2 sites (Beaumont and Port ArthurTexas) with complete data records for Hurricane Ike. All other reporting stations filed incomplete reports, with maximum winds as high as 111 miles per hour (on Bolivar peninsula) with Hurricane Ike. The Weatherflow data is contained in the CD database.

Only one land based station reported sustained winds of hurricane strength, which was Hobby Airport (75 miles per hour with gusts to 92 miles per hour). Hobby airport is 50 miles northwest of Bolivar peninsula at a higher ground elevation. The lack of reporting stations with sustained hurricane force winds is primarily due to widespread power outages as Hurricane Ike approached shore. A statement issued by the National Weather Service post-storm report dated January 16, 2009 says:

“We can only speculate that if other observations did not fail during the hurricane that there would have been more reports of hurricane force winds.”

With a reported sustained wind of 110 miles per hour at landfall and gusts as high as 135 miles per hour (Hurricane Research Division paper, authored by Chris Landsea, dated April 21, 2006), there is no argument to the sustained winds at 2898 Tropicana Lane. It is my professional opinion that winds of this magnitude affected the Bridwell property many hours prior to landfall of Hurricane Ike.

NOTE: The reported winds from HobbyAirport were not the highest winds experienced at that location. The recorded sustained wind speed of 75 miles per hour was measured after a 3.5 hour power outage while the station pressure was dropping rapidly. The observation with the winds of 75 miles per hour occurred at 6:02AM, September 13, 2008, then the power was lost a second time for another 3 hours. A sentence from the National Weather Service post-storm report states:

While the wind speeds may be at a Category 1 level on the Saffir-Simpson Scale, these winds persisted for a several hours which would contribute to more extensive wind damage across the region than if Ike was moving faster or had a smaller wind field.

The statement above from the National Weather Service infers that it was the continuous hurricane force winds that contributed to the extensive wind damage caused by Hurricane Ike.

All sustained wind observations from land based stations must be used with caution. For the past 10 years, the National Weather Service has defined sustained winds as the “average wind over a 2 minute period”. This is different than the hurricane definition of sustained winds as averaged over 1 minute.

Statistically, there is a small percentage change between a 1 minute average versus a 2 minute average, but these statistics are based on all observations over a period of several years. Individual situations such as hurricanes can produce a wide variance of wind speeds resulting in a skewed 2 minute average.

For example, I have listed several 1 minute wind speed averages below to show how a 2 minute wind speed can produce a false picture of wind speeds. When a surface observation is recorded, the last 2 minutes of wind speeds are averaged and reported. Once the observation is taken, those previous wind speeds are no longer considered for the next observation.

Time 1 minute average wind speed Observation time Reported wind speed

0123L 83 mph

0124L 63 mph 0125L 73 mph

0254L 92 mph

0255L 71 mph 0256L 82 mph

In both cases above, the higher 1 minute wind speed average would never be reported. Not only does this system cause misleading information, but when gusts are estimated the result is based upon a lower number than what actually occurred.

Knowing this information, there had to be a higher 1 minute average than 75 miles per hour.

The Hurricane Research Division of the Atlantic Oceanographic and Meteorological Laboratory produces a post-storm analysis of winds using a variety of data sources. This product is called the H Wind and is an analysis for a snapshot in time of sustained winds associated with hurricanes.

When the Hurricane Research Division (HRD) releases its’ H Wind product, it is a compilation of data from a variety of sources. It is assumed that all sources report reliable, accurate and calibrated wind speed measurements. Some of the data sources are:

  1. Weatherflow – a commercial company that maintains a mesonet of wind measuring equipment across southeast Texas. There are no records available that provide any calibration records or accuracy requirements for this equipment. There is no explanation of how Weatherflow computes sustained wind speeds.
  1. Florida Coastal Monitoring Program – managed by the University of Florida. As with Weatherflow, there are no records available to insure accuracy of data and proper measurement of wind speed. I queried FCMP in the past on how they compute sustained wind speeds and I never received an answer.
  1. Moored buoys – this data comes from the National Oceanic Service using a 6 minute wind speed average.
  1. ASOS – this is the automated network used by the National Weaher Service that only reports a 2 minute sustained wind speed average.
  1. METAR – these are manual surface observations that also must report a 2 minute wind speed average.
  1. GPS SONDE – this is an instantaneous wind speed measurement recorded as the sonde falls through the atmosphere.

All of these sources of wind speed use different times to ascertain sustained wind speeds. Since the sustained wind speeds are not manipulated to reflect one minute averages, the result is a mixture of different time intervals. This is one reason why the product is considered “experimental”.

H wind analysis for 2:30AM, September 13, 2008

All of these listed data sources are part of the HRD H Wind product. The HRD website clearly states that the H Wind product is “experimental”. The H Wind output should only be used as a “tool” of any analysis, not as an objective result.

This H Wind product has been used extensively to depict sustained winds surrounding Hurricane Ike. Users of this product must understand how it is composed and the weaknesses associated with this analysis.

All H Wind products relevant to this case are listed in the CD database.

All this information means is that the sustained wind speeds reported for Hurricane Ike are not “true” one minute averages, but a mixed combination of different reporting procedures and averaging times. Therefore, this data must be used with caution and knowing it is lower than a real-time number.

Using the Federal Meteorological Handbook Number 1 (the “bible” of weather observations), (FMH-1) it states:

The National Weather Service has a published table on what the NWS considers representative of expected results from wind speeds. It is below:

As it can be seen, the National Weather Service “expects” widespread damage with winds greater than 56 knots (65 miles per hour). These are the official expectations of the National Weather Service for wind speeds.

There are no instructions in the FMH-1 to utilize the Saffir-Simpson Scale as a tool to estimate wind speeds based upon observed damages. There are numerous reports each year where wind speeds of 50 miles per hour and greater cause serious damage across the United States. As examples of how reliable this scale is, I placed some photos of damages caused by wind speeds of 50 and 55 miles per hour. These photos fit exactly into the 48-55 category of Table 5-1 above.

A noted Midwest tornado chaser, Mr. Tim Marshall, apparently agrees that the Saffir-Simpson Scale and even the Fujita Scale are misleading when engineers assess damages after-the-fact. Mr. Marshall was answering questions submitted to him concerning tornadoes that struck La Plata Maryland in April 2002.

In one of his responses, Mr. Marshall made the following comment:

Marshall: To me there was nothing tornado-resistant about the houses in La Plata. You're dealing with conventional building construction, and that's done by a building code. Well, a building code is a minimum, and if the goal is to get to the minimum, the chances are they're barely going to get there or they're not going to get there. What is the minimum? A wind of 80 to 100 miles an hour. We're finding that if you have winds greater than that, houses get into some serious trouble. And I mean catastrophic-type trouble.

Here is a noted specialist (Mr. Marshall) that assisted in the development of the Enhanced Fujita Scale saying that when winds get above 80 to 100 miles per hour, there will be catastrophic-type trouble.

Damage to storage facility 80 mph winds Wisconsin 1998

Note how winds blew building away but barely disturbed contents

Roof torn off – 55 mph winds (St. Mary’s Ohio September 2008)

Based on what the National Weather Services mandates for use to estimate wind speed by observable damage, winds of this magnitude began on the night of September 12, 2008, many hours before the storm surge arrived between 1:00AM and 2:00AM, September 13, 2008. Wind speeds of 50 miles per hour and higher were reported at HobbyAirport as early as 6:00PM, September 12, 2008

There were a number of other weather elements within Hurricane Ike that affected this specific property hours before landfall and at least 6 hours before the storm surge came ashore on Bolivar peninsula.

Beginning on the evening of September 12, 2008 the Bolivar peninsula and 2898 Tropicana Lane experienced severe thunderstorms, microbursts and the first mesocyclone that moved across the peninsula.

Before analyzing NEXRAD radar data for mesocyclones and tornadoes, you have to be aware of the limitations of NEXRAD for these elements. The NEXRAD experts at the NWS offices in Melbourne, Tampa and Jacksonville Florida stated: (this statement is pertinent to this case because of the NEXRAD analysis technique used for hurricanes)

The shallow depth and weaker detectable rotation of the TC (tornadic) mesocyclones greatly reduced the detection capability of the current WSR-88D mesocyclone algorithm when compared to identification of traditional supercells. Even with an increased recognition of favorable synoptic scale TC-tornado environments, many of these tornadoes still occur without official (NWS) warnings or with little (or negative) lead time.

There many other publications authored by research groups that talk specifically of NEXRAD limitations for tornadic mesocyclones associated with landfalling hurricanes. Some publications state that NEXRAD may miss as many mesocyclones as it detects.

There are two main reasons for this:

  1. Mesocyclones and tornadoes associated with hurricanes are much smaller than land based elements and are not detected by NEXRAD.
  1. The algorithms placed in NEXRAD are designed for the Midwest and must be changed at individual NEXRAD sites.

Based on the information I could find on the National Weather Service website, the Houston NEXRAD still maintains the original default algorithms for both mesocyclones and tornadoes. This means that the Houston radar would only detect these elements as found in the Midwestern United States and Great Plains area. This places the Houston NEXRAD at a huge and distinct disadvantage.

The image below is from the NWS Houston NEXRAD radar depicting the initial mesocyclone affecting Bolivar peninsula. The NEXRAD image shows the mesocyclone immediately offshore from Crystal Beach and moving in a northwesterly manner coinciding with the track of Hurricane Ike.

A mesocyclone is a large rotating disturbance in the lower atmosphere that can spawn tornadoes on the ground. Various research states that 30% to 50% of all mesocyclones develop tornadoes. These percentages indicate that Bolivar peninsula was hit by numerous tornadoes before the storm surge arrived and the eye made landfall.

Initial mesocyclone moving towards Bolivar peninsula.

This initial mesocyclone occurred at 10:31PM on September 12, 2008. Mesocyclones continued to move directly over Bolivar peninsula and the Bridwell property for several more hours.

Mesocyclones moving on Bolivar peninsula September 13, 2008 12:21 AM

Mesocyclones on Bolivar peninsula September 13, 2008 12:30AM

September 13, 2008 12:49AM

Using the collected NEXRAD data and knowing the limitations of NEXRAD to detect these phenomena, I estimate that there were over 60 mesocyclones that moved over the Bolivar peninsula. Based on this estimate and using the percentages from previous studies, this would place between 18 to 30 tornadoes on Bolivar peninsula as early as 10:43PM, September 2008.

There are several pictures I took during the site visit that show distinct rotation based upon how remaining structures and poles were oriented. This indicates tornadic activity.

At 7:00PM, September 12, 2008 there were measured winds of 115 knots (126 miles per hour) only 2000 feet above the surface along the edge of the Bolivar peninsula. This measurement was recorded by a rawinsonde observation. With the numerous convective cells over the Bolivar peninsula, it is very plausible that winds from 2000 feet above the surface were transported down to the surface causing gusts as high as 100 miles per hour. These winds would not be detected by NEXRAD because the radar beam would be located above the 2000 foot level over Bolivar peninsula.

At 7:43PM, September 12, 2008 the first wave of severe thunderstorms moved over the Bolivar peninsula. Though NEXRAD appears to show echoes between 35 to 45 dbz (an indicator of intensity), by blowing up the NEXRAD image I found several cells that moved directly over 2898 Tropicana Lanein the 45 to 49 DBZ range. Guidelines state that cells of 50 dbz or higher must be present for severe weather, but this guideline only states that individual situations must be looked at objectively and not with a broad brush approach.

925 millibar chart September 12, 2008 7:00PM

By correlating the Composite Reflectivity image and Vertically Integrated Liquid NEXRAD product with the Echo Tops data, it is shown that the top of this specific cell went from a height of 29,000 feet to less than 10,000 feet within 5 minutes, indicative of a microburst. In my experience, storm cells that collapse that much within a short period produce high gusty surface winds in the vicinity of 100 miles per hour. This microburst gust of 100 miles per hour was at least 6 hours before the surge came ashore between 1:00AM and 2:00AM, September 13, 2008.

Another microburst occurred at 8:28PM, September 12, 2008 with an estimated downward speed between 100 and 105 miles per hour. This specific weather element continued to occur throughout the evening of September 12, 2008 until the hurricane made landfall at 2:00AM, September 13, 2008

Depiction of microburst flow hitting the surface.

Using the VAD Wind Profile from NEXRAD, it shows that there were low level winds in excess of 65 knots (74 miles per hour) as low as 1000 feet above the surface. The strong convection of the individual cells within the spiral bands of Hurricane Ike would result in these stronger winds at 1000 feet to be brought to the surface, mix with the prevailing surface winds resulting in gusts as high as 50 miles per hour as early as 4:00PM, September 12, 2008.

Microburst schematic

A study by the famed research expert, Theodore Fujita, states that a microburst may be created and fueled by a low level jet stream affecting the specific thunderstorm cells. The VAD Wind Profile proves that a low level jet was present over Bolivar peninsula from the afternoon of September 12, 2008. Dr. Fujita also noted that the potential wind speed from a microburst is higher than 168 miles per hour.

I point this out because whenever the winds increase in speed, the associated wind force increases exponentially. For example, if the wind speed doubles, the associated wind force increases by a factor of 4, not 2 as one would think.

VAD Wind Profile for Houston NEXRAD

The graphic on Page 13 depicting the outflow of a microburst also shows the resultant turbulence as the downward winds hit the ground and spread out. People tend to forget that not only does a microburst create very strong surface winds, but also extreme low level turbulence capable of destroying large commercial airliners.

One commonly overlooked element of a microburst is that the resultant ground turbulence (though short-lived) is a horizontal vortex, similar to a weak tornado.