case-3

[NATURAL DISASTER]

Hurricane Andrew, at the time considered the most severe national disaster to date, was a small though intense hurricane,which struck southern Florida at 5 a.m. on Aug. 24, 1992. The National Weather Service estimated the maximum sustained wind speed to have been approximately 150 MPH, with peak wind gusts of almost 200MPH. On the Saffir-Simpson scale of hurricane intensity, with five being the most intense, Andrew ranked as a four, the same ranking Hurricane Hugo was given in 1989. The storm hit the coast near high tide and produced large but localized storm surges near southeastern Dade County, 15 miles south of Miami. The USGS estimated Hurricane Andrew's maximum storm surge at 16.2 feet.

The wind, flattened whole communities, destroyed thousands of homes and businesses, and left in its wake a wasteland of debris. Andrew killed 22 people in Florida and two in Louisiana with 35 hurricane-related injuries also reported in Louisiana. According to a survey by the Miami Herald, *61,000* residents lost their homes in the storm and 87 percent of the automobiles in South Dade County suffered some kind of damage. The workplaces of more than half of those employed were destroyed. More than *650,000* claims were filed by homeowners and businesses in the area (see endnote #1).

A.M. Best Company estimated that Hurricane Andrew had cost the primary insurance industry $13.7 billion in gross losses. After reinsurance, insurers were said to pay out a pretax net loss of $7.8 billion. These findings were released December 14th, 1992 in an A.M. Best survey, "Hurricane Andrew: A Postmortem," of more than 230 property/casualty insurance companies, representing 97 percent of the property insurance sold in Florida and Louisiana. Ultimately, A.M. Best believed gross losses from Hurricane Andrew would approach $15 billion to $16.5 billion (see endnote #2).

A ruling handed down in a Dade County, Florida lawsuit against State Farm had the potential of additionally pushing up the total cost of Hurricane Andrew claims. In December 1993, acting Dade Circuit Judge Bernard Jaffee said the insurer must pay its policyholders who had purchased guaranteed replacement cost coverage policies the amount ($30,000 => $40,000) it could cost to bring hurricane stricken homes up to code. Homes not in compliance and damaged in excess of 50 percent of their market value would have to be torn down and rebuilt to code. An estimated 3,000 hurricane-damaged homes were said to have been eventually brought up to the current standards.

A recent study by the National Association of Home Builders of structures in areas hit hardest by Hurricane Andrew found that weaknesses in doors, windows and roofing, the structural elements that cover openings, had led to the most costly damage. In South Florida, which has one of the strongest building codes in the country, experts estimated that between 25 and 40 percent of Hurricane Andrew losses were avoidable. Some of these losses were due to shortages in supplies which meant long waits for repairs. For instance, six months after the hurricane, about three-quarters of severely damaged homes still had not been repaired. In addition, a Dade county Florida Grand Jury report issued in December 1992 also confirmed that much of the preventable damage was due to lax code enforcement. As a result of these findings, local governments moved to strengthen those building codes. For instance, in Dade County, windows and doors have to withstand a tough impact test or be covered with shatter-proof shutters. In addition, the number of nails needed to secure a roof were increased. On a national level, the insurance industry also began development of a building code compliance rating system, similar in implementation to its fire protection rating system which dated back to 1916. Testing of this code enforcement grading system began in 100 communities across the nation in early 1995. Inspectors were to assess such things as the "size of the building code enforcement budget relative to the amount of building activity, the professional qualifications of building inspectors and past code enforcement levels." The program was phased in the beginning of 1995 starting with the states most prone to natural disasters. By the year 2000, every municipality will have received a grade. It is projected that communities will be regraded for building code enforcement every five years thereafter (see endnote #3).

Major infrastructures were paralyzed by the storm, most noteworthy being the power distribution grid. The storm knocked down 12,620 miles of power line and more than 634 miles of high power feeder lines. At least a dozen transmission substations, and thousands of transformers and power poles also were destroyed. Approximately 1.5 million people initially lost electrical service. Water supplies were also a major problem because many of the pumping stations were without power and many of the water mains were damaged by uprooted trees. Water treatment plants were also without power. The worst damage to the transportation infrastructure were the traffic signals which sustained heavy damage (see endnote #4).

The telecommunications carrier for the area Bell South, also suffered a great deal of damage to its buildings and other physical structures, but was able to recover functionality to it's customers quickly because buried fiber optic cable substantially minimized losses in services (see endnote #5). The storm also destroyed Homestead Air Force Base and crippled Miami International Airport, destroying hangars, aircraft, buildings and incomes from job losses (see endnote #5).

How were computer information systems and software used during Hurricane Andrew? In the reporting of storm tracks, the most important source of information to a meterologist is the computer which:

"builds a model of the weather system from observations of land, sea and air conditions taken by satellites, aircraft, balloons and weatherships. The processing of the observation data is too complicated to be done without a computer which creates a simulation of the weather system as it exists at a given moment and as it will look in the future, but it is not a totally accurate picture... (see endnote #6). What the computer actually works on is a grid or mesh of calculations of such variables as temperature, pressure, wind velocity and water vapour density at chosen points in the atmosphere. The finer the grid, the less risk there is of the computer overlooking some small but vital detail which could grow in significance as the real situation develops... the computer model, based as it is on a collection of snap-shots of the weather, cannot portray or forecast the weather with complete accuracy." (see endnote #7).

In the case of Hurricane Andrew, several factors affected the timeliness of prediction. On the one hand NOAA performed exceptionally prior to and during the storm. The hurricane forecast track error was 30 percent less than average with lead times on hurricane watches and warnings being 3 to 6 hours better than average. Hurricane watches were able to be issued with 36 hours of lead time in south Florida and 43 hours in Louisiana with hurricane warnings issued with 21 hours of lead time in south Florida and 36 hours in Louisiana. Unfortunately, the collection and dissemination of information, through the appropriate warnings and statements, was found to still need improvement. In addition, tracking of storm intensity forecasting also needed improvement, as noted by the Disaster Survey Team (DST) who found "that wind, not storm surge, was the major cause of direct deaths" from Hurricane Andrew (see endnote #8).

According to the GAO, disaster relief and recovery efforts were inhibited after Andrew due to several key points. It found that the federal government's strategy:

1. lacked provisions to comprehensively assess damage;

2. lacked provisions to assess the corresponding needs of disaster victims and to provide them with quick, responsive assistance;

3. did not have explicit authority to adequately prepare for a disaster even when there is a warning;

4. that state and local governments did not have adequate training and funding to enable them to respond to catastrophic disasters on their own (see endnote #9).

The GAO flatly stated that in Hurricane Andrew's case these combined factors directly contributed to "inadequate damage assessments, inaccurate estimates of needed services, and miscommuncation and confusion at all levels of government." FEMA's own Federal Response Plan was also cited for not being adequate for dealing with catastrophic disasters. Key reasons for the Plan's failure during Andrew included the "absence of provisions for rapid assessment of the disaster's magnitude and the lack of a specific functional responsibility to respond to the extraordinary requirements of a catastrophic disaster" (see endnote #10).

It was noted by the GAO that FEMA relies on state and local governments to identify the services needed from the federal government once they have determined they cannot adequately meet their own needs. Due to the overwhelming nature of Hurricane Andrew this system failed as the state and local governments were simply unable to specify the type, amount and locations to place needed services. For example, even though FEMA officials were aware by the second day after the disaster that the American Red Cross was unable to fulfill its mass care response role, and offered to provide the state with whatever assistence it requested, Florida still did not immediately request the significant amounts of mass care it needed "because it was under the mistaken impression that the state/local/volunteer level was doing an adequate job" (see endnote #11).

Some Findings and Recommendations from the
National Disaster Survey Report: (see endnote #12).
1.
The National Weather Service needed to staff the Warning Service Forecast Office in Miami with a Warning Coordination Meterologist as soon as possible to enhance the hurricane preparedness program in South Florida. The NWS should also adopt a plan to increase the number of forecasters capable of acting as hurricane specialists during emergencies.

2.
It was found that there was a need for better protected, self contained facilities for the National Hurricane Center and all National Weather Service offices, especially at critical National Centers where full backup procedures are difficult to implement.

3.
It was determined that there needed to be more emphasis on building preparedness and awareness efforts by both state and local emergency management, especially in areas determined to have high vulnerability. It was also noted that safety and preparedness materials were in short supply. Therefore, NOAA and the Dept. of Commerce were encouraged to increase their support for developing, printing and the distributing of preparedness materials. FEMA and the National Weather service was also encouraged to increase the number of annual workshops to train coastal emergency management officials.

4.
Since lead times for evacuation may not be more than 24 hours, it was found that it "may not be practical or even possible to evacuate all inland residents. Therefore, planners need to develop a "refuge of last resort methodology, as appropriate, for occasions when critical saturation points are reached in the flow of evacuation traffic." A proven orderly plan needed to be developed for the safe and timely evacuation of entire metropolitian areas which take into consideration regional complications which can compromise orderly evacuation.

5.
The National Meterological Center was found to need to formulate plans for the intensive interactions with the media that are required when hurricane s threaten the US coastline. Television meterologists are to be encouraged in their involvement in the encouraging of evacuation from threated areas. Media in general should be encouraged to be pro-active and weather conscious in order to promote an essential link with the public to convey important information, and Emergency Broadcast System reports. Efforts should be made to include key radio as well as television providers, including non-english speaking stations. For example, it was noted that WQBA, the primary EBS station for Latin American residents in Miami, did not activate EBS during Hurricane Andrew, making an arbitrary decision that as the hurricane threat was "already well known" there was no need to activate (see endnote #13).

6.
There needed to be better education of coastal emergency managers to teach them about the scientific reasoning behind the designation of hurricane watch and warning areas. In addition there needed to be better communication between the National Hurricane Center and emergency management in regards to watch and warning designation. Emergency managers also needed to be taught how to use all tools available to them in their decision making processes, including PC-based software specifically designed for that purpose. Specific programs include, a commercially available program entitled "Enhanced Graphic Decision System (GDS), which helps users account for forecast uncertainties, and HURREVAC, which performs basic calculations and features an atlas of map screen displaying possible areas which could be flooded in various storm scenarios (see endnote #14).

7.
The National Hurricane Center, National Meterological Center and the Hurricane Research Division were found to need to develop better models and operational techniques to forcast tropical cyclonic intensity changes. Currently the National Hurricane Center's forecast operations have (7) numerical models for track prediction which include" both statistical and dynamic types with some incorporating both concepts in their design." The NHC depends largely on the use of these model outputs, but it is "always the forecaster's judgement and experience that ultimately determines NHC's official track forecast." But unfortunately there is only one operational model that is specifically used to predict the rate of strengthening, the SHIFOR (Statistical Hurricane Intensity Forecast) model. It was the SHIFOR model which failed to estimated the true rate of strengthening of Andrew. Therefore, it is stressed that more models are needed (see endnote #15).

8.
Computers: excessive heat build-up contributed to the failure of the IBM 4381 mainframe computer at the National Hurricane Center. This computer drives the McIDAS VDUC, which "injests and displays real-time interactive satellite, gridded and lightning data." It was found that the National Weather Service needed to install a stand-alone air conditioning system for the National Hurricane Center independent of the leased commercial facilities, to prevent the chance for overheating of critical communications and computer equipment.

Results from the US Army Corp of Engineers
(USACE) After Action Report:

1.
Under the 1988, Public Law 93-288, amended by Public Law 100-707, and renamed as "The Robert T. Stafford Disaster Relief and Emergency Assistance Act (93-288)", the Department of Defense has designated the USACE as the operating agent "for planning, preparedness, and response under the Emergency Support Function (ESF) #3, Public Works and Engineering." These tasks include "technical advice and evaluations, engineering services, construction management and inspection, emergency contracting, emergency repair of wastewater and solid waste facilities and real estate support" (see endnote #16).

2.
There were a total of (80) lessons learned by the USACE with resulting recommendations. Categories included were: administration; contracts; Directorate of Military Support (DOMS); emergency management; human resources; information management; individual mobilization augmentees (IMA); internal review; logistics management; prime power; construction standards; real estate; and resourcemanagement.

3.
The commander's assessment stated that "never before has the Federal government been confronted by a disastrous situation in which it was depended upon, so heavily, for assistance in saving life and property." The aftermath of Hurricane Andrew "caused events and responses that will shape and reshape the way the Federal government responds to major disasters in the future" (see endnote #17).

4.
Under the administration category , recommendations were made to:

a) create a file of administrative policy issues applicable to disaster response efforts;

b) standardize forms and have "dummy" examples of those forms to help automate the forms preparation process;

c) create updated standard organization charts with names and phone numbers of key personnel;

d) and create standardized in-processing procedures to inform individuals of current disaster situations (see endnote #18).

5.
Under the category of contracts recommendations were given to:

a) standardize letter contracts as much as possible;

b) letter contracts chould not be a primary contracting tool and should be used as little as possible;

c) definitization schedules should be established ASAP after or in conjunction with the award of contracts;

d) modify databases to also include local vendor names, phone numbers, and 24 hour point of contacts (POC);

e) give preference to local labor contracts;

f) make sure to first to wait until FEMA has determined the direction disaster relief should take before assigning contractors;

g) there needs to be "an automatic delegation of authority for the Contracting Officer at the Recovery Office to execute contracts within pre-determined limits";

h) that FAR 5.303(a) be amended to add an exception which would waive the Congressional notification requirement during a Presidentially declared emergency (see endnote #19).
t/dt><

6.
Under the category of DOMS the following recommendations were made:

a) that the Director of Military Support messages "which assign the support command relationships also must identify for the supported commander the funding source to fulfill requests for Corps of Engineers support;

b) that a coordinated link be created between the Federal Coordinating Officer (FCO), the Defense Coordinating Officer (DCO) and the Emergency Support Function (ESF) #3 team at each Disaster Field Office, to cut down on confusion;

c) that a Federal Response Plan training be developed for military personnel "to provide instruction on the ESF concepts and the military's coordination role";

d) that all the military specialty units be identified whose primary missions relate to ESF functions;

e) that DOMS and FEMA should establish clear policies on the extent of military involvement in disaster response activities (see endnote #20).

7.
Under the category of Emergency Management these recommendations were given:

a) that there is a need to identify individuals who are capable and available for serving on emergency response teams;

b) that there should be made an extensive study on the effectiveness of mobile command centers;

c) that repeated lessons learned from previous disasters be noted and utilized in the Corps of Engineers Corrective Action Program;

d) that there is a need for one automated system based on standard, readily available software, to be adopted for Corps wide use;

e) that there was a need to establish guidelines for the reporting of data and information, with exact requirements clearly defined, with written communication used to the "maximum extent possible", with all other information being transmitted electronically;

f) careful consideration needed to be given as to what information and at what frequencey the information is really needed at a headquarters level, with efforts toward the development of standardized reports;

g) that a common reporting program be created that could access other data bases;

h) that a checklist of "standard" or essential elements of information requirements be developed that could be responded to in a written format;

i) prior to a storm event have military aircraft standing by, prepositioned outside the storm area, "with sufficient personnel to perform aerial reconnaissance as soon as weather conditions permit";

j) there needs to be a field manual created that will outline the basic steps in responding to specific kinds of disasters;

k) for each activity performed toward disaster relief, a Disaster Mission Folder should be made with a checklist of all required actions (chronologically) listed necessary to the completion of each task;

i) that a matrix needed to be developed that showed all the support agencies, their available manpower, and the various skills and/or equipment that can be provided for response/recovery operations (see endnote #21).

8. Under the category of Human Resources the following recommendations were made:

a) that it was necessary to develop a "comprehensive education, training and exercise program to be implemented throughout the Corps at all levels to provide personnel with the appropriate information and latest changes in established guidance, policies and procedures for disaster response under the Federal Response Plan (FRP)";

b) that it was recommended that theResponse Planning Group establish a centralized database which would be equally accessible to all offices concerned (see endnote #22).

9) Under the catagory of Information Management, these recommendations were made:

a) have a GIS software package and knowledgeable users on hand, as well as spacial data bases for major metropolitan areas available for downloading at disaster sites;

b) place emphasis on the standardization of compatible, easily integrated software, as well as develop a standardized Emergency Management Office Data Base Management System - systems should contain standardized hardware, software and local area network connections - there should be a standardized GIS system, used by all response agencies, for maintaining day to day evaluations of progress.

c) make sure alternative methods of communication are provided to the organizations involved in response actions (see endnote #23).

In sum, it appears that the overall lessons learned through the Hurricane Andrew experience had to do most with a need for better overall communication between governmental offices, the Army Corp of Engineers, emergency management planners in the state and local levels, and a need for increased dissemination of information to the public. Emergency management policies and infrastructures had simply not been fully tested and proven prior to the disaster showing a lack of overall readiness. Computer enhanced communication systems were either insufficient, lacking backup systems, or non exhistent. Policies for public training programs and dissemination of published emergency information materials were insufficient based on the needs of the population. Such similar problems would also be encountered by governmental agencies in the following year, which we will examine in our last case of the "Great Flood of 1993."

Footnotes for Case #3:

1.
Ruth Gastel, "Catastrophes: Insurance Issues" Insurance Information Institute Reports
(Insurance Information Institute, February 1994)
2.
"A.M. Best releases Hurricane Andrew Survey Results"
Financial News(PR Newswire Assoc. Inc., 12/14/92)
3.
Ruth Gastel, "Catastrophes: Insurance Issues" February 1994
4.
Richard Korman, "Natural Disasters: Andrew exposes safety gaps"
ENR: The McGraw-Hill Construction Weekly vol.229 no.10
(New York: McGraw Hill, 9/7/92) p.8

5.
Ann Lindstrom, "Potential for Disasters keep telcos alert: Be Prepared!"
TELEPHONY vol.223 no. 13 (Kansas : Intertec Publishing Corp., 9/28/92) p.17
6.
Donald E. Fink, "Storm Cripples South U.S. Airline, Military Facilities"
Aviation Week and Space Technology 8/31/92 p. 27
7.
Dr. Forest W. Horton, Jr. and Dr. Dennis Lewis. Great Information Disasters:
Twelve prime examples of how information mismanagement led to human misery,
political misfortune and business failure. (London: Aslib, the Association for
Information Management, 1991) p. 112
8.
Natural Disaster Survey Report - Hurricane Andrew: South Florida and Louisiana August 23-26, 1992
(US Dept. of Commerce/National Oceanic and Atmospheric Administration [NOAA], November 1993) p. xvi
9.
GAO/T-RCED-93-4 "Disaster Management: Recent Disasters Demonstrate the Need
to Improve the Nation's Response Strategy" United States Accounting Office Testimony
Before the Subcomittee on VA, HUD and Independent Agencies, Committee on Appropriations,
U.S. Senate. 1/27/93 p.1
10.
GAO/T-RCED-93-4 p. 3
11.
GAO/T-RCED-93-4 pp.4-5
12.
Natural Disaster Survey Report - Hurricane Andrew. pp. xvii-xxviii
13.
Natural Disaster Survey Report - Hurricane Andrew. p. 65
14.
Natural Disaster Survey Report - Hurricane Andrew. p. 37
15.
Natural Disaster Survey Report - Hurricane Andrew. p. 51
16.
"US Army Corps of Engineers: Headquarters USACE After Action Report for Hurricanes Andrew and Iniki "
(USACE Directorate of Civil Works Operations, Construction and Readiness Division Readiness Branch, May 1993) p.1
17.
"USACE After Action Report for Hurricanes Andrew and Iniki" p.12
18.
"USACE After Action Report for Hurricanes Andrew and Iniki" pp.17-19
19.
"USACE After Action Report for Hurricanes Andrew and Iniki" pp.20-30
20.
"USACE After Action Report for Hurricanes Andrew and Iniki" pp. 31-35
21.
"USACE After Action Report for Hurricanes Andrew and Iniki" pp.35-84
22.
"USACE After Action Report for Hurricanes Andrew and Iniki" pp.85-89
23.
"USACE After Action Report for Hurricanes Andrew and Iniki" pp.90-95

(Updated 9/0/03 D.J. Russell)

Go to *LOOKING AT FOUR RECENT DISASTERS* - Case #4!

Return to *LOOKING AT FOUR RECENT DISASTERS - Table of Contents!

Send comments to DJ!

Back to DJ's home page!

Results from the US Army Corp of Engineers (USACE) After Action Report:

Footnotes for Case #3:

(Updated 9/0/03 D.J. Russell)

Results from the US Army Corp of Engineers
(USACE) After Action Report: