Service Assessment - The Central Texas Tornadoes of May 27, 1997

Service Assessment

The Central Texas Tornadoes of

May 27, 1997

U.S. DEPARTMENT OF COMMERCE

National Oceanic and Atmospheric Administration

National Weather Service

Silver Spring, Maryland

Cover:

Left: Ground view of the Jarrell tornado on May 27, 1997, between 3:40-3:45 p.m. Central

Daylight Time (CDT), just before it entered the Double Creek Estates subdivision. Photograph

courtesy of Scott Beckwith, Jarrell, Texas.

Right: Aerial view taken May 31, 1997, of a portion of the Double Creek Estates subdivision in

Jarrell, Texas. Note that only the home foundations remain. Photograph by L. Phan, National

Institute of Standards and Technology (NIST).

Service Assessment

The Central Texas Tornadoes of

May 27, 1997

April 1998

U.S. DEPARTMENT OF COMMERCE

William M. Daley, Secretary

National Oceanic and Atmospheric Administration

D. James Baker, Administrator

National Weather Service

John J. Kelly, Jr., Assistant Administrator

Preface

Participation on a Service Assessment survey team is both a rewarding and potentially trying

experience. Witnessing the personal tragedy caused by a disaster and assessing services provided

by the National Weather Service (NWS) are responsibilities not taken lightly. Reports on natural

disasters are very important. The recommendations, which represent the team’s views, are given

careful consideration by the Agency in its attempts to continuously improve warning services.

The magnitude of this event made it emotionally troubling for every team member. I wish to

express sincere appreciation to the Service Assessment Team members for their thorough,

professional, and unbiased review of the disaster and for their recommendations toward service

improvements.

Robert S. Winokur

Acting Assistant Administrator for

Weather Services

April 1998

iii

Foreword

This report on the series of tornadoes which struck central Texas on May 27, 1997, was prepared

by a National Oceanic and Atmospheric Administration (NOAA) Service Assessment Team

following a 4-day visit to the storm sites. The meteorological analyses presented in the report

were developed by the Team during a 4-month period following the on-site assessment. The team

has drawn upon a number of external resources, including the Storm Prediction Center (SPC),

National Environmental Satellite, Data, and Information Service (NESDIS), Operational Support

Facility (OSF), and university resources to make the interpretations of the severe weather setting.

It should be noted that the meteorological analyses and interpretations, since they were done after

the fact, make use of all available information. Some of the charts and photographs shown in this

report were not available to the staffs of the Next Generation Weather Radar (NEXRAD)

Weather Service Forecast Offices (NWSFOs) Dallas/Fort Worth and Austin/San Antonio, Texas,

prior to the occurrence of the tornadic thunderstorms.

The report is based upon information gathered by the team through a visit to the Austin/San

Antonio NWSFO in New Braunfels, Texas; aerial and ground surveys of three of the tornado

track sites; and interviews with members of other agencies and organizations and state, county

and municipal governments. The team expresses its appreciation to the county and municipal

officials who took time from urgent disaster response duties to share their impressions of the

event and interpretations of the effectiveness of NWS products and services. Similarly, the team

would like to extend its appreciation to the staffs of the Dallas/Fort Worth and Austin/San

Antonio NWSFOs for making themselves readily available and for outlining their actions before

and during the event.

This report assesses the effectiveness of NWS products and services before and during the period

of the severe thunderstorms as they moved through central Texas. Furthermore, the report

presents the team’s facts, findings and recommendations as defined by their survey.

The Service Assessment Team

Table of Contents

Page

Preface ................................................................ ii

Foreword .............................................................. iii

Service Assessment Team .................................................. v

Acronyms .............................................................. vi

Event Summary.......................................................... 1

Facts, Findings and Recommendations......................................... 15

Appendix A Chronology of Releases Related to the Central Texas Tornadoes ...... A-1

Chronology of Events at the Storm Prediction Center ............ A-1

Chronology of Events at NWSFO Fort Worth, TX .............. A-5

Chronology of Events at NWSFO Austin/San Antonio, TX........ A-9

Appendix B The Jarrell, Texas, Tornado Outbreak: A Meteorological Satellite

Perspective............................................... B-1

Appendix C Aerial Damage Survey of the Central Texas Tornadoes of

May 27, 1997 ............................................. C-1

Appendix D Persons Interviewed and/or Contacted by the Service Assessment

Team ................................................... D-1

Appendix E CDC Field Epidemiological-Health Report ....................... E-1

Appendix F Fujita Tornado Intensity Scale................................. F-1

Status of Actions......................................................... S-1

Service Assessment Team

Following a major severe weather event in which there has been extensive damage or loss of life,

a Service Assessment Team may be assigned by NOAA to evaluate the role played by the NWS,

to provide an objective appraisal of products and services, and to make findings and

recommendations for improving the service. Such a team was assembled to survey tornadic

thunderstorms which struck parts of central Texas on May 27, 1997. This document is the final

report from this team.

Team Members

Leader, James H. Henderson, Deputy Director, Aviation Weather Center, National Centers for

Environmental Prediction (NCEP), NWS, Kansas City, Missouri

Team Coordinator/Editor, Ida M. Hakkarinen, NWS Liaison and Meteorologist, NOAA Program

Coordination Office, Washington, D.C.

Assessment Facilitator, William H. Lerner, Meteorologist, Office Of Meteorology, NWS Headquarters,

Silver Spring, Maryland

Melvin R. McLaughlin, Chief, Meteorological Services Division, NWS Southern Region Headquarters,

Fort Worth, Texas

James M. Looney, Chief, Meteorological Services Division, NWS Central Region Headquarters, Kansas

City, Missouri

E. L. (Buddy) McIntyre, Chief, Transition Management Unit, NWS Southern Region Headquarters,

Fort Worth, Texas

Brian E. Peters, Warning Coordination Meteorologist (WCM), NWS NWSFO Birmingham, Alabama

Marilu Trainor, NOAA Public Affairs Specialist, NWS Western Region Headquarters,

Salt Lake City, Utah

Dr. Enrique Paz, U.S. Centers for Disease Control and Prevention (CDC), Atlanta, Georgia

Dr. Shellie Ann Kolavic, Epidemic Intelligence Service, CDC, Texas Department of Health (TDH),

Austin, Texas

David Zane, Director, Injury Prevention and Control Program, TDH, Austin, Texas

Adjunct Member

Dr. Long T. Phan, Professional Engineer, Structures Division, NIST, Gaithersburg, Maryland

(accompanied Brian Peters on the aerial surveys)

Acronyms

ARC American Red Cross

AVHRR Advanced Very High Resolution Radiometer

AWIPS Advanced Weather Interactive Processing System

CCelsius

CAPE Convective Available Potential Energy

CDC U.S. Centers for Disease Control and Prevention

CDT Central Daylight Time

CIMSS Cooperative Institute for Meteorological Satellite Studies

CIRA Cooperative Institute for Research in the Atmosphere

CR County Road

CST Central Standard Time

CSTAR Collaborative Science, Technology, and Applied Research Program

CWA County Warning Area

DLF Del Rio, Texas (DOD radar site at Laughlin Air Force Base)

DMIC Deputy Meteorologist in Charge

DOD Department of Defense

EAS Emergency Alert System

EWX Austin/San Antonio NWSFO, New Braunfels, Texas

FCC Federal Communications Commission

FWD Dallas/Fort Worth NWSFO, Fort Worth, Texas

FEMA Federal Emergency Management Agency

FFA Flash Flood Advisory

FFW Flash Flood Warning

FLW Flood Warning

ft feet

GOES Geostationary Operational Environmental Satellite

GPS Global Positioning System

GRK Granger, Texas, WSR-88D (DOD radar site at Fort Hood)

J/kg Joules per kilogram

km kilometer

LI Lifted Index

LSR Local Storm Report

m /s meters squared per second squared

mb millibar

MCS Mesoscale Convective System

MESO Mesocyclone Algorithm-Signature

MIC Meteorologist in Charge

NAWAS National Warning System

NCEP National Centers for Environmental Prediction

NESDIS National Environmental Satellite, Data, and Information Service

NEXRAD Next Generation Weather Radar

vii

NIST National Institute of Standards and Technology

nm nautical miles

NOAA National Oceanic and Atmospheric Administration

NOW Short Term Forecast (nowcast)

NWR NOAA Weather Radio

NWS National Weather Service

NWSFO NEXRAD Weather Service Forecast Office

OSF Operational Support Facility

RAMM Regional Atmospheric Modeling System

RAMSDIS RAMM Advanced Meteorological Satellite Demonstration and Interpretation

System

RVS River Statement

SEL Tornado Watch

SPC Storm Prediction Center

SPS Special Weather Statement (e.g., Hazardous Weather Outlook)

SR State Road

SVR Severe Thunderstorm Warning

SVS Severe Weather Statement

TDH Texas Department of Health

TLETS Texas Law Enforcement Telecommunication System

TOR Tornado Warning

TVS Tornado Vortex Signature

m micrometer

USWRP U.S. Weather Research Program

UTC Coordinated Universal Time

VAD Velocity Azimuth Display

WCM Warning Coordination Meteorologist

WFO Weather Forecast Office

WSR-88D Weather Surveillance Radar-1988 Doppler

ZFP Zone Forecast Product

Event Summary

Overview

During the mid-afternoon and early evening of May 27, 1997, a severe weather outbreak

produced numerous severe thunderstorms and multiple tornadoes across central Texas. The

synoptic weather situation that occurred was an atypical pattern for springtime central Texas

tornadoes. Outbreaks of Texas tornadoes are usually associated with a “classic” meteorological

pattern that has a strong mid-level (500 millibar [mb]) trough of low pressure combined with

favorable jet stream winds and integral strong low-level boundaries to form and enhance the

thunderstorms.

On this day, the 500 mb low center was located well to the north over Nebraska, and the upper-

level jet stream axis was over northern Oklahoma. There was not a well defined low-level jet. A

weak cold front extended from northeastern Texas southwestward to near Del Rio. The

numerous severe thunderstorms that occurred throughout central Texas developed in a situation

of weak wind shear and high thermodynamic instability, along the surface boundary, where the

greatest threat is typically strong winds and large hail. On this day in central Texas, there were 34

reports of severe weather—12 for large hail and 22 for tornadoes.

The F5 tornado (see Appendix F, Fujita Tornado Intensity Scale) that occurred near the

unincorporated community of Jarrell in Williamson County caused 27 of the 30 fatalities. Another

tornado death occurred in Cedar Park, which is also in Williamson County. In Travis County,

The Pedernales Valley tornado killed one person. One additional death from drowning occurred

at Shoal Creek in Austin.

Since 1950, there have been 40 tornadoes in Williamson County. On April 7, 1980, an F3

tornado killed one and injured two. A May 17, 1989, tornado (also rated an F3) killed one and

injured 28. Since 1950, 46 tornadoes have occurred in Travis County. On July 4, 1970, an F2

tornado killed one and injured four. Until May 27, 1997, there had never been an F5 or F4

tornado reported in either Williamson or Travis County.

Meteorological Situation

(Italicized sections contain detailed information which may be of more interest to technically

oriented individuals.)

In the early morning hours of May 27, a large Mesoscale Convective System (MCS) developed

over Arkansas and eastern Oklahoma. The MCS moved slowly eastward and produced gravity

waves that radiated away from the system. An analysis of the satellite data, using sophisticated

tools not available to the Austin/San Antonio NWSFO (EWX) forecasters, showed that one of

these gravity waves appears to have played a role in the initiation of the thunderstorm complex

that eventually produced the tornado activity when it intersected the cold front in the vicinity of

Waco. However, this gravity wave moved south of the storm complex before the beginning of the

tornado activity.

Meteorological conditions were conducive to widespread severe convection during the afternoon

of May 27. Area upper air soundings indicated extreme instability. Modified Convective

Available Potential Energy (CAPE) exceeded 7500 Joules per kilogram (J/kg) in central Texas.

(CAPE is a measure of the amount of energy available for convection and is directly related to

the maximum potential vertical speed within a thunderstorm updraft. Higher values for CAPE

indicate a greater potential for severe weather. Observed values in thunderstorm environments

often exceed 1000 J/kg; in extreme cases, values may exceed 5000 J/kg.)

Although the atmospheric instability was extreme, the environmental wind fields were not

indicative of those typically expected for supercell thunderstorm development. Vertical wind

shear in the atmosphere is an important ingredient for tornadic development. Storm-relative

helicity is used by NWS meteorologists to quantify this vertical wind shear. In particular, the

measure of helicity in the lowest 3 kilometers (km) of the atmosphere (3 km helicity) is used

extensively. In general, 3 km helicity values below 300 meters squared per second squared

(m /s ) correlate to weak tornadoes.

On May 27, winds were relatively light through much of the lower atmosphere, and storm-

relative helicity values from the 7 a.m. CDT NEXRAD Weather Service Office Corpus Christi

sounding were 70 m /s . (Note: All times listed are in Central Daylight Time.) A forecast

sounding from the Rapid Update Cycle numerical model, valid for 7 a.m. near the Austin, Texas,

area, had values of 117 m /s . Consequently, the convection was strongly influenced by storm-

scale processes and confined to the vicinity of the existing surface boundaries.

The thunderstorm complex that would eventually become tornadic and produce the Jarrell

tornado developed in McLennan County over the southern sections of Waco around 12:40 p.m.

in the Dallas/Fort Worth NWSFO (FWD) county warning area. The storm developed rapidly and

produced its first tornado at 1:20 p.m. near Lorena, approximately 15 miles south of Waco in

southern McLennan County. This tornado, classified as very briefly reaching category F1,

damaged several mobile homes and severely damaged one frame house 2 miles west of Lorena.

The storm continued to intensify and propagate southward and produced a second, much stronger

tornado near Moody in extreme southern McLennan County, about 22 miles south of Waco, at

1:45 p.m. The Moody tornado, classified as an F3, had a path length of 3.7 miles and a width of

150 yards. It did extensive damage to trees, destroyed a house and a barn, damaged another small

house, and tossed a car and a pickup truck several hundred feet.

A third tornado was produced by this same storm complex in Bell County near Belton at

2:35 p.m. This F3 tornado began near Morgan’s Point on the north side of Lake Belton, crossed

the lake, and came ashore near the community of Woodland, creating a path 1.4 miles long and

275 yards wide. The Belton tornado reached its maximum intensity after crossing the lake,

causing total destruction to trees and substantial damage to structures.

Early radar data of these storms from the EWX Weather Surveillance Radar-1988 Doppler

(WSR-88D) were significantly range-folded due to second trip ground clutter. An analysis of the

radar data showed that when the storm emerged from the range-folded area at 2:56 p.m., about

20 minutes prior to the development of the Jarrell tornado, there was a deep mesocyclone of

minimal strength in southern Bell County. Mesocyclonic rotation strengthened to moderate to

strong values (>40 knots at 70 nautical mile [nm] range) and a strong and deep gate-to-gate

signature was detected in southern Bell County at 3:19 p.m. Gate-to-gate velocity differences

continued to be strong until after the 3:43 p.m. radar volume scan.

The aerial survey (see Appendix C) showed that the final and most devastating tornado produced

by this particular storm complex developed near the community of Prairie Dell in the extreme

southern part of Bell County, approximately 3 miles north of the Williamson County line.

An analysis of the radar data from the Department of Defense (DOD) WSR-88D radar at

Granger, Texas (GRK), showed a detailed view of the developing circulation at 1,500 feet (ft)

elevation and above. Prior to the development of the tornadic circulation, two fine lines were

detected in the reflectivity field. The gust front on the western side of the Jarrell storm

intersected the larger scale wind shift boundary where the tornadic signature developed in an

area of strong and deep convergence. The tornadic circulation in the lowest 5,000 ft appeared

to link up with the mesocyclone rotation in the middle and upper levels. This radar ceased

operation at 3:38 p.m. and was no longer available to the EWX forecasters.

Actual time of the Jarrell tornado’s development was hard to pinpoint but is believed to be

between 3:15 p.m. and 3:20 p.m. Initially, the tornado was quite weak and small, probably being

no more than F0 or F1. The tornado moved slowly south-southwest, crossing into Williamson

County at approximately 3:25 p.m. Shortly thereafter, the tornado explosively strengthened,

becoming a multi-vortex F5 tornado before moving into the Double Creek Estates subdivision on

the west side of Jarrell at 3:40 p.m. Here, devastation was complete. The entire subdivision,

consisting of approximately 38 single-family structures and several mobile homes, was destroyed.

Three businesses, located adjacent to the subdivision, were also destroyed.

Even though the Jarrell storm produced a mesocyclone and a violent tornado, the radar

reflectivity field did not show a distinct hook echo directly preceding the tornado. A small hook

was apparent on the Granger radar data after the tornadic circulation developed. Although most

supercells contain mesocyclones on the right rear flank with respect to storm motion, the

mesocyclones on this day were located on the front flank (southwest quadrant) as they moved

slowly south-southwestward. The difference in mesocyclone location was probably due to

enhanced southwestward propagation along the boundary. Despite the fact that the reflectivity

features were not typical and could have somewhat confused a radar interpreter, the rotational

signals were clear and easy to follow.

Twenty-seven people perished in Jarrell. There were relatively few injuries reported—one serious

and less than ten minor, an inference to the small probability of survival. The Jarrell tornado had a

total path length of 7.6 miles and a maximum width of 3/4 mile. Shortly after exiting the Double

Creek Estates subdivision, the tornado dissipated and the parent tornadic storm itself weakened.

While the Jarrell tornado was evolving, a separate storm developed on the larger scale wind shift

boundary to the southwest.

On the flank of this new storm, the EWX radar detected an area of deep convergence with weak

shear around 3:25 p.m. that eventually evolved into the Cedar Park mesocyclone. The Cedar

Park circulation developed similarly to the Jarrell circulation in that it formed in an area of

strong convergence along the wind shift boundary in a large weak echo region (6 nm in

horizontal extent) well displaced from the low-level reflectivity field. The weak shear rapidly

intensified during the 3:48 p.m. radar volume scan, when a mesocyclone of moderate strength

was detected with maximum rotational velocities of 36 knots and a maximum radial velocity

difference of 70 knots at 50 nm from the radar. The moderate to strong mesocyclone rotational

velocities and strong gate-to-gate shear continued until about 4:12 p.m. and weakened

thereafter. The Cedar Park mesocyclone and gate-to-gate shears were different from their

Jarrell counterparts in that the strong signatures for Cedar Park remained aloft for much of the

time while the strong signatures for Jarrell were seen at all vertical levels.

The Cedar Park mesocyclone produced its first tornado at approximately 3:45 p.m. The tornado

began about 3.5 miles north of Cedar Park, which is also in Williamson County. This tornado

initially caused mainly F0 and F1 damage until it moved into the business district of Cedar Park at

4 p.m., where F3 damage was evident in a small area. An Albertsons Food and Drug Store was

severely damaged, and one person in the store was seriously injured as the tornado moved

through the shopping center. There were a number of other customers in the Albertsons store,

but the store manager had directed them into one of the store’s walk-in refrigeration units which

provided them safe refuge during the storm.

The Cedar Park tornado continued moving south-southwest, then gradually turned more

southwest as it moved through the Buttercup Creek subdivision on the southwest side of Cedar

Park. In the subdivision made up of well-constructed, upscale homes, damage ranged from F1 to

F3, with some homes suffering only minor roof damage while others suffered major damage. As

the tornado moved through the subdivision, 136 homes sustained damage. One person, a

69-year-old male, died from a cardiac arrest apparently as he was waiting out the storm inside his

truck in the garage.

The tornado then exited Cedar Park and dissipated about 1.1 miles from the northern shore of

Lake Travis in Travis County. As were all of the tornadoes on this day, the Cedar Park tornado

was slow-moving. In fact, some residents videotaped the approaching funnel for 10 to 15 minutes

and still had ample time to seek shelter in interior rooms of their homes before the wind began to

increase with the approach of the vortex itself. The Cedar Park tornado path dimensions were 9.2

miles in length and 250 yards in maximum width.

A second tornado developed from this storm complex near the south shore of Lake Travis around

4:30 p.m. As it rapidly achieved at least F3 intensity, it caused extensive damage to a marina on

the lake shore. Then increasing to F4 intensity, it severely damaged a reinforced building

containing a telephone switching center and completely destroying a stone house nearby, leaving

the foundation slab swept clean.

The tornado initially moved just south of west before turning southwest for a short distance, then

back to the west-southwest before moving into the Hazy Hills subdivision of the Pedernales

Valley in western Travis County. The tornado was a strong F3, with some evidence of F4

intensity, as it moved through the subdivision destroying numerous houses and mobile homes. A

25-year-old male died when he was either blown from his mobile home or from his pickup truck

as he attempted to flee the storm. The Pedernales Valley tornado damaged or destroyed

45 homes and a Southwestern Bell telecommunications switching center. It had a path length of

5.6 miles with a maximum width of 440 yards.

One additional death from drowning occurred at Shoal Creek in Austin. This fatality, a 38-year-

old female, was likely a result of the localized flash flooding that accompanied the storm system.

NWS Products and Services

The possibility of severe thunderstorms on this day had been anticipated. The NWS’s SPC, part

of the NCEP and located in Norman, Oklahoma, had placed the area in a moderate risk of severe

thunderstorms in the Severe Weather Outlook that was issued at 1:03 a.m. on Tuesday morning,

May 27.

As the severe weather event progressed during the day, the SPC continued to monitor the

development and subsequently issued two tornado watches for the area as well as several status

updates. The SPC issued Tornado Watch #338 at 12:54 p.m. for portions of east Texas, valid

from 1:15 p.m. until 7 p.m. This included Bell, McLennan and Williamson Counties. At

3:31 p.m., the SPC issued Tornado Watch #340 for south Texas, including Williamson and Travis

Counties, valid from 3:45 p.m. until 10 p.m. (Refer to Appendix A for details of SPC actions.)

NWSFO Dallas/Fort Worth Products

The NWSFO Dallas/Fort Worth office issued a Severe Weather Outlook at 5:32 a.m. The

Outlook noted there was a moderate risk of severe thunderstorms during the afternoon and night

over much of north Texas, including Bell and McLennan Counties. Subsequent forecasts and

discussions issued in the late morning continued to mention the threat of severe weather. (Refer

to Appendix A for a chronology of NWSFO FWD actions.)

Forecasters were obviously concerned about the extreme atmospheric instability in evidence on

this day. In an updated Severe Weather Outlook for North Texas issued at 12:20 p.m., NWSFO

FWD stated:

High instabilities aloft mean that thunderstorms that do form in the slight risk

area should rapidly become severe.

Nevertheless, forecasters were still focusing on the threat of large hail and damaging straight-line

winds rather than tornadoes. This was because of a lack of strong vertical wind shear as

evidenced by early morning atmospheric soundings. In the same Severe Weather Outlook

mentioned above, the forecaster went on to state:

Strong vertical motions in these storms...due to strong CAPES [convective

available potential energy]...suggest a main threat of large hail or damaging

straight-line winds. Weak wind shear profiles suggests that the tornado potential

is low.

As convection began to develop in McLennan County shortly before 12:45 p.m., the NWSFO

FWD wasted no time in immediately issuing a severe thunderstorm warning for the county at

12:50 p.m., anticipating that the storm would quickly become severe in the extremely unstable

atmosphere. Shortly thereafter, at 12:54 p.m., the SPC issued Tornado Watch #338. This watch

was valid from 1:15 p.m. until 7 p.m. and included McLennan and Bell Counties, as well as

Williamson County further to the south.

As the NWS WSR-88D in Fort Worth began showing significant updraft rotation in the

McLennan County storm, the office quickly switched to tornado mode and issued a tornado

warning for McLennan County at 1:21 p.m., noting that the radar was detecting a tornado near

Hewitt, a suburb of Waco located approximately 14 miles north-northeast of Moody. A severe

weather statement at 1:26 p.m. noted the tornado was west of Lorena, approximately 10 miles

north-northeast of Moody. The tornado was now confirmed by spotters as well as being detected

by the WSR-88D. At 1:48 p.m., the NWSFO FWD issued a tornado warning for Bell County

stating:

National Weather Service Doppler radar detected a tornado between Eddy and

Moody. This tornado has been confirmed by storm spotters.

A follow-up severe weather statement at 2:16 p.m. noted:

Spotters and radar continue to track a tornado located in northern Bell County

near Highway 317 and Southerland Road at 2:15 p.m. If you live in Moffat or

Morgans Point Resort...take cover immediately.

Subsequent storm surveys indicated a weak tornado (probably F0 or F1) lasting about 3 minutes,

occurred at Lorena at 1:20 p.m. A brief but intense tornado (classified as F3) occurred at Moody

at 1:45 p.m., and another brief F3 tornado occurred at Lake Belton, Bell County (Morgans

Point), at 2:35 p.m.

The NWSFO FWD issued the first tornado warning at 1:21 p.m. for McLennan County, 24

minutes prior to the tornado at Moody but minus 1-minute lead time for the county (the Lorena

tornado). This was followed at 1:48 p.m. by a tornado warning for Bell County (that was

extended at 2:30 p.m.), providing a 47-minute lead time for the county (the Lake Belton tornado).

The NWSFO FWD issued a total of ten tornado warnings, five severe thunderstorm warnings,

nine severe weather statements, and eight short-term forecasts during the 4-hour period from

1 p.m. until 5 p.m. on the afternoon of May 27.

NWSFO Dallas/Fort Worth Staffing

During the period from 12 noon until 5 p.m. on May 27, the NWSFO FWD was operationally

staffed in the following manner. The Forecaster in Charge and a journeyman forecaster handled

the routine forecast operations. A meteorologist intern was in training with the Forecaster in

Charge on the synoptic (long-term) desk. As convective activity began to develop, the WCM

assumed operation and monitoring of the WSR-88D, initially the NWSFO FWD WSR-88D and

then the DOD GRK WSR-88D as the activity moved south. (The Fort Worth NWSFO is an

associated user of the DOD GRK radar.) The Meteorologist in Charge (MIC) also assumed

operational duties, assisting with the radars and issuing warnings. A second meteorologist intern

assisted with coordination activities and in warning issuance, dissemination, and confirmation of

receipt. Two hydrometeorological technicians provided public service and NOAA Weather Radio

(NWR) programming, and a summer aide assisted with the phone load and NWR programming.

A second journeyman forecaster was called in on overtime at 3 p.m. to assist with warning

operations. In addition to the “normal” operational staffing, the NWSFO FWD used seven

additional personnel to handle this severe weather event.

NWSFO Austin/San Antonio Products

Forecasters at the Austin/San Antonio office were also concerned about the very high CAPE

values present in the atmosphere on May 27. (Refer to Appendix A for a chronology of NWSFO

EWX actions.) The forecast discussion issued at 3:40 a.m. noted the unstable air mass present

over the area, and stated, “...will mention possibly severe (thunderstorms) for this afternoon.”

Subsequent public forecasts which included the counties of Williamson and Travis called for a 30

percent chance of showers and thunderstorms, some possibly severe. In addition, the NWSFO

EWX issued a Hazardous Weather Outlook at 6 a.m., calling for the possibility of severe

thunderstorms area wide during the afternoon and extending into the nighttime hours. Another

forecast discussion issued at 10:15 a.m. dropped mention of severe weather, but an updated

Hazardous Weather Outlook issued at 11:15 a.m. continued the call for possible severe

thunderstorms and isolated tornadoes and was headlined “...Severe Thunderstorm and Heavy

Rain Event Developing Across South Texas and the Hill Country this Afternoon and Tonight....”

The Outlook went on to state:

“The atmosphere is very moist and unstable across south Texas. Conditions are

becoming increasingly favorable for the development of scattered severe

thunderstorms with large hail...damaging winds and very heavy rainfall. The

most severe storms may produce isolated tornadoes.”

Following the SPC’s issuance of Tornado Watch #338 at 12:54 p.m., which included Williamson

County, the NWSFO issued updated zone forecasts at 1:09 p.m., headlining the watch in the

appropriate zones. A Short Term Forecast issued at 1:10 p.m. for Williamson and Travis

Counties was headlined with the watch information but noted no significant precipitation was

indicated on radar.

At 2:15 p.m., the staff contacted both the Georgetown Police Department and the Williamson

County Sheriff’s Department to alert them that severe thunderstorms were moving southward

through Bell County and that severe weather was very likely in the next 1 to 1 ½ hours.

At 2:48 p.m., another Short Term Forecast was issued for Williamson and Travis Counties and

mentioned “...radar indicated a severe thunderstorm centered just west of Temple moving slowly

southwestward. Through 5 PM...partly cloudy skies will prevail with scattered showers and

thunderstorms developing by late afternoon...some could possibly be severe....” This forecast

also headlined the existence of the tornado watch for Williamson County.

The forecast discussion issued at 2:50 p.m. emphasized a heavy rain threat. The discussion stated

in part:

“Complicated weather pattern setting up with weak cold front dropping through

hill country tonight then upper level disturbance near 300 mb moves across south

central Texas at or about 18Z Wednesday. Tricky to decide where and when

heaviest rain will fall because amounts will be high requiring FFA [Flash Flood

Advisory] if activity is anything more than scattered.

...Service hydro feels activity and boundary will remain mostly stationary and

drift slowly southward and get intercepted by upper level 300 mb disturbance

early Wednesday AM 12-18Z producing very heavy rain across escarpment

area....After discussion with staff...rather than issue blanket FFA will wait until

location of heavy rain becomes clearer....”

The next issuance from the NWSFO EWX was the tornado warning for Williamson County at

3:30 p.m. The text of the warning stated in part:

“At 3:25 p.m. a tornadic thunderstorm was located about 5 miles west of Jarral

[sic] moving southeast at 10 mph. This storm has had a history of producing

tornadoes and large hail. The City of Jarrel [sic] is in the path of this storm.”

This warning gave a lead time of 10 minutes for the town of Jarrell, but no lead time for the

extreme northern portion of Williamson County.

A few minutes later, at 3:35 p.m., the office issued a Short Term Forecast for Williamson and

Travis Counties (among others) stating:

“At 3:30 PM radar indicated a tornadic thunderstorm centered just north of

Jarrel [sic] with another developing rapidly north of Georgetown. Movement was

southwest at around 10 mph. Residents are warned these storms have a history of

producing baseball [size] hail and tornadoes across central Texas this afternoon.

Through 5 PM...severe thunderstorms are developing across northern portions of

south central Texas and should continue marching southwest.”

At 3:31 p.m., the SPC issued a subsequent tornado watch, #340, for south Texas which included

Williamson and Travis Counties. Tornado Watch #340 was posted at 3:45 p.m. and remained in

effect until 10 p.m.

At 4 p.m., the NWSFO EWX issued the routine afternoon package of zone forecasts for south

Texas. The forecast for Williamson and Travis Counties, in addition to the obligatory tornado

watch headline, called for “Mostly cloudy with a 50% chance of showers or thunderstorms...

some possibly severe with locally heavy rainfall.” This was followed at 4:09 p.m. with a tornado

warning for Travis County. The warning stated in part:

“At 4:05 PM a tornadic thunderstorm was located over Cedar Park. This storm

is moving south at 10 mph. This storm will move into northern portions of the

City of Austin by 4:30 PM. A tornado was spotted over Cedar Park in

Williamson county at 4 PM.”

This warning gave a lead time of 21 minutes for Travis County (the Pedernales Valley tornado).

As heavy rains accompanied the tornadic storms, a flash flood warning was issued at 4:34 p.m. for

both Williamson and Travis Counties, valid until 6:30 p.m.

The NWSFO EWX issued a total of eight tornado warnings, twenty-four severe thunderstorm

warnings, ten flash flood warnings, four severe weather statements and fourteen short-term

forecasts during the afternoon and evening of May 27. Approximately 80 percent of these

issuances took place after 5 p.m. It also should be noted that, in addition to tracking the tornadic

storm in Williamson and Travis Counties, the NWSFO EWX was working a second severe

thunderstorm complex which had developed to the west in Edwards County. After 6 p.m., more

widespread severe weather developed throughout much of the NWSFO EWX’s county warning

area (CWA).

NWSFO Austin/San Antonio Staffing

During the daytime hours on May 27 (8 a.m. to 4 p.m.), operational staffing at the NWSFO

EWX consisted of: (1) the Forecaster in Charge working the short-term forecast desk (Meso

desk) and responsible for monitoring the EWX radar; (2) a journeyman forecaster working the

synoptic (long-term) desk and responsible for monitoring the DOD WSR-88D radar at Del Rio,

Texas (DLF); (3) two hydrometeorological technicians—one operating and interpreting data from

the EWX WSR-88D and the other programming NWR; and (4) two meteorologist interns—one

handling public service duties and the other assisting the EWX and DOD DLF radar operations by

composing and issuing warnings. During this period of time, in addition to the “normal”

operational staffing, the NWSFO EWX used the MIC, the Deputy MIC (DMIC), and the WCM

to assist in working the early stages of the severe weather event.

During the late afternoon and evening hours (4 p.m. to midnight), staffing consisted of: (1)

Forecaster in Charge working the short-term forecast desk; (2) a journeyman forecaster working

the synoptic desk as well as operating and monitoring the EWX WSR-88D; (3) a second

journeyman forecaster monitoring the DOD DLF WSR-88D; (4) three meteorologist

interns—two handling public service and the phone load while the third intern assisted the DOD

DLF WSR-88D forecaster in issuing warnings. Also, two hydrometeorological technicians were

on duty—one assisting the EWX WSR-88D operator while the other handled programming the

NWR. The DMIC stayed on duty until 6:45 p.m. During the late afternoon and evening hours in

addition to the “normal” operational staffing, the NWSFO EWX used five additional personnel to

assist in working the latter stages of the event.

An analysis of the staffing and delivery of products issued by the meteorologists during the mid-

afternoon of May 27 indicated that the routine forecast packages were prepared and disseminated

on schedule. A hydrometeorological technician and a meteorologist intern operated and

interpreted the EWX WSR-88D for severe weather; neither of whom had completed formal

WSR-88D training at the OSF. Interviews also indicated the staff was somewhat perplexed with

the inconsistent behavior of the WSR-88D mesocyclone algorithm and a transitory weakening of

the storm. These facts, combined with the atypical meteorological situation, may have influenced

the warning decision process.

Dissemination

Communication of warning information, especially in Williamson and Travis Counties, was

somewhat problematic during this episode. The Emergency Alert System (EAS) was not

activated in a timely fashion in at least one case. The delay of 25-30 minutes was verified by a

representative of the Texas Broadcasters Association. According to local media representatives,

the delay in broadcasting warning information from NWFSO EWX to the general public via EAS

may have been caused by the media outlet choosing to manually activate EAS instead of allowing

automatic activation. This choice is prevalent throughout the central Texas area because

broadcasters maintain that, due to the number of warnings that can be issued, automatic activation

would cause too many interruptions of their programming. Such decisions are part of state and

local EAS plans.

The Texas Department of Public Safety Law Enforcement Telecommunication System (TLETS)

is used throughout the state to convey important information about criminal activity to agencies.

In addition to police-related messages, TLETS is used to distribute weather warning information

to law enforcement and emergency management agencies. TLETS functioned as designed

throughout the episode and was effective in communicating warning information to Williamson

and Travis Counties.

Public Response

The Service Assessment Team conducted interviews with residents of Jarrell, Cedar Park and the

Pedernales Valley to determine if the tornado warnings were received by the residents and if so,

their response to the warnings. Appendix D lists the persons contacted by the Service Assessment

Team. Concurrent with the NWS interviews, a team from the U.S. Centers for Disease Control

and Prevention and the Texas Department of Health conducted an assessment to describe

mortality and morbidity related to the tornadoes and to provide casual hypotheses for fatalities

and injuries. This report is provided in Appendix E.

Almost everyone interviewed said they were aware of the tornado warnings. Most said they first

learned of the warnings through commercial television. Others said they received telephone calls

from relatives or friends about the approaching storms. Although most were aware of the

availability of the NWR, no one mentioned NWR as a source of receiving the tornado warnings.

Due to the slow movement and high visibility of the tornadoes, most of the residents interviewed

said they watched the approach of the tornadoes prior to taking shelter. Most said they knew to

go to the center of their houses, to avoid staying in mobile homes, and to seek shelter rather than

trying to flee the tornadoes. These actions definitely saved lives. However, in the case of the F5

tornado at Jarrell, people were killed even though they took the appropriate safety measures.

Actions taken by the residents of Buttercup Creek (Cedar Park tornado) and Hazy Hills

(Pedernales Valley tornado) attest to the value of tornado preparedness efforts and adherence to

tornado safety rules. Although approximately 175 homes were damaged or destroyed in these

two subdivisions, there was only one fatality and no serious injuries. The single fatality, in the

Hazy Hills subdivision, occurred either when the occupant was blown from his mobile home or

from his pickup truck as he attempted to flee the storm.

Media stories written and aired during the aftermath highlighted people in the path of the Jarrell

tornado who fled the area and survived while many others who obviously followed prescribed

safety measures perished. Media reports such as this provide conflicting guidance to the public

about appropriate safety measures. This “mixed message” may lead to some people in the future

trying to flee the storm area rather than seeking appropriate shelter.

It appeared through interviews with store and restaurant managers that formal tornado or other

severe weather safety plans for protection of employees and customers did not exist. The

Albertsons grocery and Taco Bell managers in Cedar Park, however, instructed their employees

to move to reinforced areas of their buildings. In the case of the Albertsons store, which

sustained significant damage, these actions undoubtedly saved many from serious injury and

possibly death.

Two stories told to the Service Assessment Team illustrate how knowledge of tornado safety

rules can save lives. The manager of the Albertsons grocery store in Cedar Park went outside into

the parking lot to watch the approaching thunderstorm. While approximately 60 customers were

shopping inside, he watched for a few minutes as the tornado developed. Going back inside the

store, he instructed that an announcement be made asking everyone to gather in the center of the

store so he could lead them into the store’s cooler in the rear of the building. He also asked

customers trying to leave the store to stay inside. Although the store’s wide-span roof collapsed

into the center of the store, the customers survived in the cooler with only a few minor injuries.

One store employee who did not make it into the cooler suffered serious injury. The actions of

the store manager saved many from certain serious injury and possibly death. When asked how he

knew to keep customers from fleeing the store and to move customers and employees into the

store’s cooler, the manager replied that he “was raised in Wichita Falls and was there during the

1979 tornado.” The 1979 Wichita Falls tornado killed more than 40 people. Over half the

fatalities in Wichita Falls occurred in automobiles, many of whom were trying to flee the tornado.

In the Jarrell tornado, structures in its immediate path were totally destroyed. A few residences

on the fringe of the tornado path, while sustaining nearly total destruction, still had some walls left

standing. In one such residence, a grandmother, her daughter, daughter-in-law and two young

grandchildren survived in a centrally located bathroom. The walls of the bathroom were all that

were left standing of the home. The grandmother and daughter told the Assessment Team that

they did not know where in the house to seek shelter from the tornado, which they could see

approaching them. The daughter told the Assessment Team, “My sister-in-law saved our lives.”

The daughter-in-law, who lived just down the road, had run to the house with her child to seek

shelter from the storm. She instructed everyone to get into the central bathroom. Her relatives

commented that she had picked the correct bathroom of the two in the house. She stated “But

you always go to the center of the house.” The Assessment Team asked how she knew about

what to do and she replied that she had “learned the [tornado] safety rules as a child.” Without

this knowledge and quick action on her part, there most certainly would have been five additional

fatalities in the Jarrell tornado.

Events related to the Jarrell tornado were indescribably tragic. It is the consensus of the Service

Assessment Team that most residents in the Double Creek Estates subdivision were aware of the

impending danger and knew the proper precautionary measures to take. Nevertheless, due to the

force (F5), width, and slow rate of movement of the tornado, the protective measures taken were

to no avail for those directly in its path of destruction.

Remains of a home adjacent to the Double Creek Estates subdivision of Jarrell.

This home is located on the edge of the tornado path. Five people survived the

storm in a bathroom (center of picture). Photograph by E. L. McIntyre, NOAA/NWS.

Home destroyed in Buttercup Creek subdivision during Cedar Park tornado.

Several people took shelter in a utility closet, shown just to the left of the woman

in the picture. Photograph by NWSFO Austin/San Antonio.

Aerial view of Albertsons grocery store showing damage from the Cedar Park

tornado. Photograph by L. Phan, NIST.

Ground view of damage to Albertsons grocery store. Photograph by M. Trainor,

NOAA.

Facts, Findings and Recommendations

A. Observations

Finding 1:

The amateur radio base station at NWSFO EWX was not activated for Tornado

Watch #338. The NWSFO EWX did try to contact the amateur radio coordinator for

Tornado Watch #340 but could not raise him. They also could not contact a backup

coordinator. No one on duty at NWSFO EWX knew how to turn on the amateur

radio equipment for passive monitoring. Therefore, vital spotter information was not

received via this amateur radio link. A meteorologist intern at NWSFO EWX

indicated he did receive one report via a phone call from an amateur radio operator.

Recommendation 1:

Personnel at NWSFO EWX should be trained in how to turn on the amateur radio

equipment so that they can at least hear reports, even if the network is not activated.

NWSFO EWX should also work with the amateur radio coordinator on backup

procedures for contacting key people in impending emergencies.

Finding 2:

The Fort Hood WSR-88D (GRK) (Granger, Texas, WSR-88D) ceased to function

after the 3:38 p.m. volume scan and remained out of service for a number of hours.

This radar is operated and maintained by the U.S. Air Force. The radar was located

only 20 miles from the Jarrell tornado. An associated Principal User Processor display

system from this radar is installed in the NWSFO FWD. The radar was invaluable to

NWSFO FWD for warnings issued in McLennan and Bell Counties. If the radar had

ceased to function earlier in the afternoon, warning operations would have been

severely impacted.

It was subsequently learned that the radar had shut down because of a commercial

power failure and that its emergency power failed to come on-line because a control

switch had been left in the improper position during a previous maintenance visit.

Recommendation 2:

The NOAA and the NWS should increase efforts in working with the DOD to ensure

that DOD commanders and personnel understand that properly maintained and

operated DOD WSR-88Ds are vital to the severe weather warning programs for both

the bases and the civilian populations under the radars’ umbrellas.

Finding 3:

Satellite data provided a number of valuable pieces of information for use in severe

thunderstorm nowcasting over Texas on May 27, 1997. (Appendix B provides a

research analysis of the May 27 events using NOAA meteorological satellite data.)

The storm complex that produced the tornado activity between Moody and the

Pedernales Valley on this day had cloud top features that are often associated with

thunderstorms that have very intense updrafts. The precise location, orientation and

movement of the front along which the tornadic storm complex evolved was easily

monitored using Geostationary Operational Environmental Satellite (GOES) imagery.

Unfortunately, real-time data flow into a satellite analysis and display system that

would have allowed for a number of useful applications was not available to the

NWSFO EWX forecasters.

Recommendation 3:

The NOAA/NWS should move rapidly to ensure that all Weather Forecast Offices

(WFOs) in the NWS are equipped with the best possible satellite analysis capabilities

that are part of the Advanced Weather Interactive Processing System (AWIPS). The

NWS should also provide proper training to ensure that satellite data are used in the

warning and forecast program.

B. Forecast Guidance

Finding 4:

A coordination call was not made from the SPC to NWSFO EWX prior to Tornado

Watch #338 being issued even though the watch included a part of NWSFO EWX’s

CWA. Such a coordination call may have further heightened the severe weather

awareness of the staff at the NWSFO EWX.

Recommendation 4:

The SPC should make every effort to coordinate with any NWS office that has

warning responsibility before a watch is issued that covers a part of the CWA of that

NWS office.

Finding 5:

The tornado outbreak was a unique event in that significant tornadoes were produced

in a situation of weak shear and high instability. Although storms exhibited supercell

characteristics (mesocyclones with tornadoes, weak echo regions, deviate storm

motion), an unusual storm evolution occurred. High thermodynamic instability and

storm interactions with a pre-existing wind shift boundary resulted in highly deviate

storm motion and the rapid development of mesocyclones and tornadoes. The

mesocyclone and tornado development formed farther ahead (to the southwest) of the

low-level radar reflectivity field than is typically observed. The mesocyclones grew

from areas of deep convergence within large weak echo regions characteristic of very

strong updrafts. While the evolution may have been unexpected, easily detectable

mesocyclones developed prior to the tornadoes. However, the mesocyclone

algorithm-signature (MESO) on the EWX WSR-88D was inconsistent and missed

several significant circulations. The tornado vortex signature (TVS) algorithm did not

trigger. The team is concerned with these failures.

Recommendation 5:

The OSF should evaluate the Level-2 Archive data from the EWX radar and further

investigate these anomalies. The NEXRAD Program should continue its plans to

implement the National Severe Storm Laboratory’s experimental mesocyclone

algorithm because of its higher skill and informative outputs, including probabilistic

guidance values.

Finding 6:

This outbreak of tornadoes was perceived as an unusual event in that the tornadoes

tracked to the southwest with several tracks having a right loop at the end. The storm

environment was characterized by weak vertical wind shear and high thermodynamic

instability. Interpretation of GOES satellite loop imagery suggests the possibility of a

gravity wave moving southwest across northern Texas and the lower Mississippi

Valley into this area of extreme low-level instability. This feature may have initiated

the first supercell thunderstorm that formed.

Recommendation 6:

A thorough scientific study of these events should be undertaken by researchers and

academicians with the results made available to forecasters of the NWS. Any study

should include the U.S. Weather Research Program (USWRP), the NWS

Collaborative Science, Technology, and Applied Research Program (CSTAR) and

NOAA/NWS Cooperative Institutes.

C. Watches, Warnings and Statements

FACT C1:

The actual time of the Jarrell tornado development was hard to pinpoint, but it is

believed to be between 3:15 p.m. and 3:20 p.m. Initially, the tornado was weak and

small, probably no more than F0 or F1. It moved slowly south-southwestward crossing

into Williamson County around 3:25 p.m. Shortly thereafter, the tornado explosively

strengthened to F5 before moving into the Double Creek Estates subdivision on the west

side of Jarrell around 3:40 p.m.

FACT C2:

The SPC issued Tornado Watch #338 at 12:54 p.m. for portions of east Texas valid

from 1:15 p.m. until 7 p.m. At 3:31 p.m., the SPC issued Tornado Watch #340 for

south Texas, including Williamson County, valid from 3:45 p.m. until 10 p.m.

FACT C3:

At 2:15 p.m., the NWSFO EWX staff phoned the Georgetown Police Department and

the Williamson County Sheriff’s Department to alert them that severe thunderstorms

were moving southward through Bell County and that severe weather was likely in the

next 1 to 1 ½ hours.

FACT C4:

NWSFO EWX issued a tornado warning for Williamson County at 3:30 p.m. This gave a

lead time of at least 10 minutes for the town of Jarrell but no lead time for the northern

portion of Williamson County.

Finding 7.1:

Individuals at NWSFO EWX who were directly associated with WSR-88D

manipulation and interpretation of the data it was providing on May 27, 1997, had not

attended the intense 4-week OSF operational training course. The OSF course has

been canceled, and the NWS is planning a new method of training for those personnel

who will manipulate and interpret the WSR-88D radar.

Finding 7.2:

Prior to the development of the Jarrell tornado, a series of mesocyclones had a history

of producing tornadoes in McLennan and Bell Counties in the NWSFO FWD’s CWA.

Early data on the storms from the NWSFO EWX WSR-88D were significantly range-

folded due to second trip ground clutter. This range-folding could have been

mitigated if the NWSFO EWX radar operators had changed the pulse repetition

frequency.

Finding 7.3:

Although most supercell thunderstorms contain mesocyclones on the right rear flank

with respect to storm motion, the mesocyclones on this day were located on the front

flank (southwest quadrant) as they moved slowly south-southwestward. The

difference in mesocyclone location is probably due to enhanced southwestward

propagation along the boundary. The storms emerged from the range-folded area on

the NWSFO EWX WSR-88D about 20 minutes prior to the development of the Jarrell

tornado. Despite the fact that the reflectivity features were not typical and could have

perplexed a radar interpreter accustomed to viewing radar reflectivity signatures alone,

the rotational velocity signatures were clear and easy to follow and should not have

confused an operator trained in WSR-88D data interpretation.

Finding 7.4:

Meteorologists trained in mesoscale events recognize that fluctuations in storm

strength should not be surprising when storms form in an environment of weak shear

but high instability. At the same time, with an explosively unstable thermodynamic

environment, transient weakening in existing convection should be viewed with

extreme caution. Furthermore, this “atypical” (see Finding 5) meteorological situation

may have influenced the performance of the WSR-88D MESO.

Recommendation 7:

Use and interpretation of the WSR-88D data, especially in convective situations,

should be left to meteorologists trained in both mesoscale meteorology and the

interpretation of WSR-88D data. In addition, NWS WFOs should be appropriately

staffed with a sufficient number of such meteorologists to meet these requirements.

Further, the NWS needs to ensure that any future training programs being developed

to replace the OSF WSR-88D course should be of equal or better effectiveness than

was that course. Meteorologists must have the ability to quickly interpret reflectivity

and velocity signatures within the context of the overall mesoscale weather pattern,

and relate the radar information to the other analyses and observing tools such as a

detailed mesoscale surface analysis combined with satellite, surface observations,

profiler and radar Velocity Azimuth Display (VAD) vertical wind profiles, etc.

Accordingly, the NWS should ensure that the expertise, materials, and other

resources, including field staff time, are made available to meet such a training

requirement.

Finding 8.1:

At NWS NWSFOs, the Forecaster in Charge on duty is the “supervising operational

meteorologist” and is responsible for warning and forecast operations, including severe

weather staffing and products, if a storm event develops. This procedure was not

implemented effectively at NWSFO EWX on May 27, 1997. Responsibility for severe

weather operations was assumed by a meteorologist intern and a radar operator.

During the early stages of the severe weather episode, no single forecaster supervised,

monitored and coordinated closely the activities and decision-making processes.

Finding 8.2:

As the severe weather approached the northern part of NWSFO EWX’s CWA, short-

term forecasts were limited in number and not sufficient. This occurred as the office

transitioned from a stance of routine operations into immediate full-scale warning

operations as the severe weather moved into the area. The office issued many more

short-term forecasts, as well as a number of severe weather statements, later in the

evening to supplement warnings on a number of tornadic events that occurred further

south in the southern portions of the NWSFO EWX’s CWA.

Finding 8.3:

The product mix flowing from the office early in the severe weather event did not

seem to be coherent in terms of the warnings, short-term forecasts and severe weather

statements.

Finding 8.4:

A review of the state forecast discussion and other products issued by NWSFO EWX

indicated that the office was focusing upon the potential for a heavy rain episode,

which did occur later in the evening, instead of tornadic activity. In addition,

interviews with the staff mentioned that, early in the severe weather episode, they were

concerned with getting the routine afternoon forecast package “out on time” and

devoted considerable resources to that effort.

Recommendation 8:

Management at NWSFO EWX should establish and enforce clear lines of authority

and responsibility to be followed during severe weather operations. Further, emphasis

should be given to keeping all staff current regarding such operational procedures.

Protection of life and property is the fundamental mission of the NWS. NWS offices

need to develop a strategy for a coherent product suite encompassing warnings,

statements and short-term forecasts throughout severe weather episodes. They need

to work with customers to apprise them of priority which will be placed upon products

during such events. All NWS WFOs should have a standing policy that routine

products have a much lower priority during periods of ongoing severe weather. They

should ensure that all staffing resources be brought into a cohesive severe weather

mind set and that the focus of the office is always on the fundamental mission.

D. Internal and External Service Coordination

Finding 9:

After the issuance of warnings, considerable staff resources were invested by NWSFO

EWX to confirm receipt of warning information by emergency management officials.

In some cases, multiple phone calls were made into the same county.

Recommendation 9:

The staff at NWSFO EWX is commended for their dedication and extra effort to seek

confirmation that warning information was being received by officials who had a

critical need for it. Nevertheless, it is recommended that NWSFO EWX work with

emergency management officials to seek alternative and more efficient methods to

confirm receipt of warnings.

Finding 10:

During the severe weather events of May 27, 1997, there were occasional delays in

coordination between the NWSFO FWD and the NWSFO EWX as the tornadic

situation shifted from NWSFO FWD’s CWA to NWSFO EWX’s CWA. The

coordination mechanism of choice, commercial telephone, was hampered by busy

signals. The Texas State National Warning System (NAWAS) circuit was available to

both FWD and EWX, but NWS offices in Texas use commercial telephones for

coordination because of various logistical problems associated with NAWAS.

The Texas NAWAS links intra-state warning points and NWS offices throughout

Texas. The circuit is divided into two sub-circuits, but Texas has made the decision to

operate the circuit as a single state-level circuit. Considering the size of the state, the

circuit can be busy with law enforcement traffic, and the state discourages its use for

routine business. It was noted that NAWAS traffic was heavy on May 27, and if NWS

personnel had chosen to use the circuit for coordination purposes, they may have had

to ask others to get off the circuit (an unlikely scenario in non-emergency,

coordination-only situations).

Nationally, NAWAS was in the process of being upgraded to better accommodate the

internal/external communication and coordination requirements of the NWS and the

emergency management community. Because of Federal budget constraints, the

upgrade has been halted, and Texas NAWAS terminals have not been fully upgraded.

Recommendation 10:

The NWS should reach consensus with the Federal Emergency Management Agency

(FEMA) and the state emergency management agencies on a single communication/

coordination system that is consistent nationally and meets the internal and external

communication/coordination requirements of both the NWS and the emergency

management community. All NWS offices should be provided with needed technology to

facilitate efficient internal and external coordination.

E. Dissemination

Finding 11.1:

There was a significant delay in the activation of the EAS. For example, the official

notification to EAS from the NWR was at 3:30 p.m. The EAS was not activated until

3:56 p.m. in at least one instance.

Finding 11.2:

Broadcasters maintain that the EAS “ticker” does not identify why EAS activation is

being requested. Consequently, they have to wait for hard copy which says “EAS

activation requested.” This introduces delays in the activation procedure.

Finding 11.3:

Media outlets apparently choose manual activation of EAS instead of automatic

activation. This choice is prevalent throughout the central Texas area because

broadcasters maintain that automatic activation would cause too many interruptions of

their programming.

FACT E1:

NWS policy only requires EAS activation requests on tornado warnings and flash

flood warnings.

Recommendation 11:

The NWS has sent to the Federal Communications Commission (FCC) proposed

revisions to the EAS rules. The NWS should continue to work with the FCC to clarify

situations that need automated activation. The NWS and the FCC, in cooperation

with the National Association of Broadcasters, should agree on a limited set of critical

weather warning products which would be programmed for automatic activation of

the EAS.

Finding 12:

The TLETS is designed to carry information primarily related to law enforcement

activities. It is not designed to be a weather warning dissemination system;

nevertheless, it is used by emergency managers and law enforcement agencies to

obtain weather information. In this particular event, TLETS was effective in

disseminating warning information in a timely manner. However, since TLETS is not

a totally automated system (i.e., it requires the manual routing of products), delays in

the dissemination of time-sensitive weather information can occur.

Recommendation 12:

The NWS should continue to work with the state of Texas Division of Emergency

Management to seek ways to improve the timely transmission of weather warnings via

TLETS and should intensify efforts on both a local and state level to explore

alternative methods of communicating critical weather products to emergency

management officials.

Finding 13:

The general public interviewed by the survey team did not use NWR as a first-choice

source of weather information. They relied on commercial radio and television. The

broadcast media, in general, used NWR as a backup source of severe weather

information, but not all radio stations have NWR receivers. Some emergency

managers used NWR as a primary source of severe weather information.

Recommendation 13:

The NWS should develop additional partnership initiatives with the public and private

sector to encourage the use of NWR for the receipt of critical weather information.

Finding 14:

The NWR transmitter serving Austin and surrounding communities lost power as the

storms moved through the general Austin area, and there is no backup power for the

site.

Recommendation 14:

All the NWR transmitters should be equipped with backup power.

F. Response

Finding 15.1

After the Jarrell tornado, which moved unusually slowly, there was a widespread belief

by the general population that persons who took appropriate safety actions were

killed, while persons who took inappropriate safety actions lived. This has prompted

questions from the public as well as the media regarding the viability of tornado safety

rules the NWS promotes.

Finding 15.2:

Through a limited number of interviews with people in the Cedar Park area, the

Pedernales Valley area, and even some people in the Jarrell area, the Service

Assessment Team found that lives were definitely saved because people took

appropriate safety actions in accordance with long-standing recommendations of the

NWS, FEMA, the American Red Cross (ARC) and the emergency management

community.

Recommendation 15:

The NWS, in concert with FEMA and the ARC, should continue to emphasize the

importance of not attempting to flee an approaching tornado and reemphasize the

importance of adhering to long established tornado safety rules. (In the vast majority

of situations, people who take shelter in sturdy, well-built structures will survive most

tornadoes whereas people who try to flee in automobiles will place themselves at much

greater risk.) The Service Assessment Team also recommends that there be significant

public and industry education on the meaning of well-built tornado shelters.

G. Preparedness

Finding 16:

The majority of managers of supermarkets, restaurants and other stores that the

Service Assessment Team interviewed did not have predetermined tornado or other

severe weather safety plans to activate for protection of their employees and

customers during severe weather threats. Worthy of note, however, is that the

Albertsons and Taco Bell managers in Cedar Park instructed their employees and

customers to move to a reinforced area of their buildings as the tornado approached.

The action of the Albertsons manager undoubtedly saved many lives.

Recommendation 16:

The NWS should partner with other public agencies and the private sector to develop

plans for protection of employees and the public located in large buildings during

severe weather threats.

H. System/Infrastructure

FACT H1:

NWS offices are typically staffed for fair weather situations. During difficult

mesoscale weather events, offices frequently have to either recruit all personnel on

duty, hold people on overtime, or call in whomever they can reach from home to help

with the increased workload associated with programming the NWR, operating and

interpreting data from WSR-88Ds, performing analyses, and assisting in warning

operations.

As the NWS continues to move forward with its restructuring, the staffing at offices, such

as the NWSFOs at Fort Worth and Austin/San Antonio, although currently augmented with

transitional staffing, is projected to decrease. The NWS also plans, in a multi-phased

approach, to decentralize severe weather watches from the SPC to the field WFOs. To

mitigate the workload associated with these and other responsibilities, the NWS is

depending upon AWIPS, an automated Console Replacement System for NWR, and the

integration of other functions to allow forecasters to more easily handle duties covering the

continuum of information.

Finding 17:

The workload associated with the events of May 27, 1997, challenged the staffs of the

NWSFOs involved. At NWSFO FWD, seven additional personnel beyond the normal

shift complement were pressed into service, including the MIC and WCM who were

intimately involved in the warning process. NWSFO Austin/San Antonio augmented

their normal shift structure with five personnel, including the MIC and Deputy MIC

who were on the forecast floor and available as necessary.

Recommendation 17:

First, the NWS should ensure that all field offices develop strategies that make full use

of all office personnel qualified to assist in warning and forecast operations during

critical weather situations.

Second, the NWS should evaluate its plans regarding future staffing and service changes to

ensure that workload constraints do not compromise the agency’s safety of life and property

mission. Specifically, transitional staffing which has been placed at selected offices should

be retained until such time as new technology can demonstrate it is capable of mitigating the

workload associated with warning and forecast operations to an extent that a reduction in

staff is feasible.

Finding 18:

Over 10,000 tornadoes have occurred in the United States during the last 10 years.

During that period of time, only six of these tornadoes have been rated as an F5. Of

those six, only two tornadoes occurred under the synoptically evident southwesterly

flow. The other four occurred in atypical westerly or northwesterly flow. Such

situations are often accompanied by extreme instability. The extreme instability is not

sufficient to determine storm type and/or tornado development since these same

patterns often produce bow echoes and associated derechos instead of tornadoes.

Extreme instability could play a role in tornado intensity through stretching or some

other storm-scale effect.

Recommendation 18:

Studies should be undertaken by the research and academic community to help identify

these types of patterns which may aid in the development of these rare but catastrophic

tornadoes. Such studies should have the participation of the USWRP, CSTAR, and

NOAA/NWS Cooperative Institutes. The results of these studies should be made

available to NWS forecasters to aid in the warning and forecast process.

Finding 19:

Since NOAA and the NWS do not have a Quick Response Hazards Survey Team in

place, it usually takes 48-72 hours to assemble a team and put that team into the

centroid of the outbreak. This is especially true for the aerial part of the survey which

allows highly detailed documentation which is necessary to support findings that are

generated by the NWS modernized suite of equipment. In the case of the central

Texas tornadoes, there were not a lot of trees and local authorities were quick to

repair damage and remove debris. This hampered the exact location of the damage

paths and the times of occurrence. The uncertainty of the times or the paths of

tornadoes makes it very difficult to analyze the data for current performance or

develop strategies for better future performance.

Recommendation 19:

NOAA/NWS should compile a list of potential, quick-response team members, drawn

from appropriate expertise in the NOAA line and staff offices, other Department of

Commerce Bureaus such as NIST, and from the universities and private sector. The

team should be interdisciplinary in nature. Following a significant tornado outbreak, a

team of three to five scientists should be assembled and dispatched to the centroid of

the outbreak within 24 hours. Teams should include meteorologists, a wind engineer,

and a social scientist, an epidemiologist and a public affairs specialist. Funding to

carry out the travel costs of the team and an aerial damage-mapping survey should be

set aside yearly in a revolving budget.

Aerial view of damage from the Moody tornado. Photograph by L. Phan, NIST.

Aerial view of damage to boats at Belton Lake from the Belton tornado.

Photograph by L. Phan, NIST.

Close-up aerial view of the remains of a home in the Double Creek Estates

subdivision of Jarrell. Photograph by L. Phan, NIST.

Close-up view of a home foundation in the Double Creek Estates subdivision of

Jarrell. Note the plumbing pulled out of the concrete in the center of the picture.

Photograph by NWSFO Austin/San Antonio.

Asphalt stripped from the road by the Jarrell tornado in the Double Creek Estates

subdivision. Photograph by I. Hakkarinen, NOAA/NWS.

Remnants of a mobile home trailer in the Double Creek Estates subdivision of

Jarrell. Photograph by NWSFO Austin/San Antonio.

Damage to a telephone switching station caused by the Pedernales Valley

tornado. Photograph by M. Trainor, NOAA.

Truck in the Hazy Hills subdivision that was blown up a hill by the Pedernales

Valley tornado. Photograph by I. Hakkarinen, NOAA/NWS.

A-1

Appendix A

Chronology of Releases

Related to the Central Texas Tornadoes

Chronology of Events at the Storm Prediction Center

From 1:00 am CDT through 11:00 pm CDT - Tuesday, May 27, 1997

103 AM DY1 CONVECTIVE OUTLOOK...MODERATE RISK OF SEVERE THUNDERSTORMS

FOR HAIL AND DAMAGING WINDS INCLUDING TRAVIS AND WILLIAMSON

COUNTIES...

1016 AM DY1 CONVECTIVE OUTLOOK...MODERATE RISK OF SEVERE THUNDERSTORMS.

NCLUDING TRAVIS AND WILLIAMSON COUNTIES...CONDS WILL BE FVRBL

FOR THE DVLPMT OF SCTD INTENSE STMS WITH VRY LRG HAIL..LCLY DMGG

WINDS AND PSBLY ISOLD BRIEF TORNADOES.

1254 PM AWW QUICKIE NOTIFICATION OF TORNADO WATCH 338

1254 PM SEL8 TORNADO WATCH #338 ISSUED FOR EAST TEXAS INCLUDING BELL,

MCCLENNAN AND WILLIAMSON COUNTIES. DISCUSSION MENTIONED VERY

LARGE HAIL/LOCALLY DAMAGING WINDS AND ISOLATED TORNADOES

POSSIBLE THIS AFTERNOON IN VERY UNSTABLE AIR ASS/CAPE TO 5000 J/KG/

OVER REGION. ANY TORNADOES WILL LIKELY BE CONFINED TO BOUNDARY

INTERSECTIONS GIVEN COMPARATIVELY WEAK VERTICAL SHEAR.

306 PM MCD STATUS OF TORNADIC STORM NEAR TPL...VERTICAL WIND PROFILES ARE

QUITE WEAK...WITH LATEST KEWX DATA SHOWING VARIABLE WINDS AT 5-

10 KNOTS THROUGH 700 MB. DEGREE OF INSTABILITY CERTAINLY

SUPPORTS LARGE HAIL AND WIND DAMAGE...BUT IT IS UNCLEAR HOW

TORNADIC SUPERCELL NEAR TPL WILL EVOLVE AS IT ENCOUNTERS

APPARENTLY WEAKER MID LEVEL FLOW OVER THE NEXT FEW HOURS. DUE

TO PROPAGATION OF THIS STORM AND EXPECTED RAPID DEVELOPMENT

OF NEW CONVECTION TOWARD DRT...A NEW TORNADO WATCH WILL BE

REQUIRED SOUTH AND WEST OF #338 SHORTLY.

331 PM AWW QUICKIE NOTIFICATION OF TORNADO WATCH 340

331 PM SEL0 TORNADO WATCH 340 ISSUED FOR SOUTH TEXAS INCLUDING WILLIAMSON

AND TRAVIS COUNTIES. DISCUSSION MENTIONED INTENSE

THUNDERSTORMS WITH VERY LARGE HAIL...HIGH WINDS AND ISOLATED

TORNADOES POSSIBLE THIS EVENING WHERE MODERATE LOW LEVEL

INFLOW OF VERY MOIST AIR CONTINUES BENEATH AXIS OF STEEP LOW TO

MID LEVEL LAPSE RATES AHEAD OF SOUTHEAST MOVING COLD FRONT.

DEEP LAYER VERTICAL SHEAR LIMITED...BUT TORNADOES STILL POSSIBLE

IN VICINITY OF BOUNDARY INTERSECTIONS.

A-2

339 PM WWA STATUS OF WW#338...ISOLD SVR TSTMS ARE OCCURRING OVR PARTS OF ERN

TX ALG/AHD OF WEAK SFC CDFNT. SUPERCELL NOW OVR BELL CNTY TX NR

TPL CONTS TO DVLP SLOWLY SWD ALG FRONT. THIS STORM HAS ALREADY

PRODUCED SVRL TORNADOES WITH RPTS OF SGFNT DMG. ANOTHER

SUPERCELL APPRS TO HAVE DVLPD OVR NAVARRO CNTY. TORNADO

THREAT SHUD REMAIN FOCUSED ALG BNDRY OVR

NAVARRO/HILL/MCLENNAN/BELL/WILLIAMSON/TRAVIS CNTYS NEXT 1-2

HOURS WHERE STG CONVERGENCE IS OCCURRING. WW 340 HAS JUST BEEN

ISSUED S OF CURRENT WATCH.

457 PM WWA STATUS OF WW#338...AN EXTRMLY UNSTBL AMS RMNS IN PLACE OVR MUCH

OF ERN TX/WRN LA. SFC OBS SHOW TEMPS IN THE UPR 80S AND LWR 90S

AND DWPNTS IN THE LOW TO MID 70S...YIELDING SFC BASED CAPE VALUES

AOA 4500 J/KG. THE STGST CELLS WERE CURRENTLY LOCATED IN

PONOLA...LIMESTONE...AND WILLIAMSON CNTYS. A TORNADO WAS RPRTD

WITHIN THE PAST HR IN CEDAR PARK IN NRN WILLIAMSON CNTY. THIS

PARTICULAR CELL CONTS TO MOV/PROPAGATE TO THE S AND WILL LIKELY

MOVE INTO WW 340 WITHIN THE NEXT HR OR SO. ALTHOUGH WNDS ARE

WK ALF...THERE IS STG VEERING IN THE LOW TO MID LVLS. GIVEN THE

DEGREE OF INSTABILITY...THE WND FIELDS SUG THE PTNL FOR FURTHER

TORNADIC DVLPMNT WL CONT FOR THE NXT SVRL HRS.

A-3

SPC Tornado and Severe Thunderstorm Reports

Partial Preliminary List For Texas

For 0600 Central Standard Time Tue May 27 1997

Thru 0600 CST Wed May 28 1997

EVENT LOCATION REMARKS

DATE/TIME

LAT/LON

.....TORNADO REPORTS...................TORNADO REPORTS..............TORNADO REPORTS......

1 TORN 3 W LORENA TX (16 S ACT) 27/1224

FTW/LSR 3138 9727

2 TORN 2 W LORENA TX (16 S ACT) 27/1254

SEVERAL MOBILE HOMES DAMAGED AND ONE HOUSE

DESTROYED FTW/LSR 3138 9725

3 TORN 1 E MOODY TX (11 NNE TPL) 27/1259

BUILDINGS DAMAGED FTW/LSR 3131 9733

4 TORN BELTON TX (5 SSW TPL) 27/1331

FTW/LSR 3106 9746

5 TORN 3 N BELTON TX (3 SW TPL) 27/1340

FTW/LSR 3111 9746

6 TORN 3 W BELTON TX (8 SW TPL) 27/1354

TORNADO BETWEEN NOLANVILLE AND BELTON

FTW/LSR 3106 9752

7 TORN BLOOMING GROVE TX (44 NE ACT) 27/1405

FTW/LSR 3210 9671

8 TORN 2 W HUBBARD TX (27 ENE ACT) 27/1450

FTW/LSR 3184 9683

9 TORN 6 N GEORGETOWN TX 27/1455

ALONG I-35 SAT/LSR 3073 9766

10 TORN LEANDER TX (20 NNW AUS) 27/1500

SAT/LSR 3058 9784

11 TORN JARRELL TX (25 SSW TPL) 27/1515

*** 30 DEAD *** SAT/LSR 3081 9759

12 TORN AUSTIN TX (2 WSW AUS) 27/1515

AUSTIN PD REPORTED TORNADO ON GROUND

NEAR 2222 AND ANDERSON MILL

SAT/LSR 3028 9774

13 TORN CEDAR PARK TX (13 NNW AUS) 27/1515

*** 2 DEAD *** REPORT OF GROCERY STORE

COLLAPSE IN CEDAR PARK. REPORTS

SAT/LSR 3049 9781

14 TORN AUSTIN TX (2 WSW AUS) 27/1550

TORNADO ON GROUND NEAR LAKEWAY AND 620

SAT/LSR 3028 9774

15 TORN LEANDER TX (20 NNW AUS) 27/1600

SAT/LSR 3058 9784

A-4

16 TORN KENDALIA TX (28 N SAT) 27/1745

SHERIFF REPORTS TORNADO IN SW BLANCO

COUNTY MOVING TOWARDS KENDALI

SAT/LSR 2996 9851

17 TORN BLANCO COUNTY (40 W AUS) 27/1745

TX SAT/LSR 3026 9838

SHERIFF REPORTS TORNADO IN SW BLANCO COUNTY

MOVING TOWARDS KENDALI

18 TORN BLANCO TX (38 N SAT) 27/1808

TORNADO ON GROUND SW OF BLANCO

SAT/LSR 3009 9841

19 TORN UTOPIA TX (27 NW HDO) 27/1900

SHERIFF REPORTED TORNADO ON GROUND

SAT/LSR 2961 9953

20 TORN MOUNTAIN HOME TX

(34 SE JCT) 27/1903

TORNADO REPORTED ALONG I-10 IN KERR

COUNTY SAT/LSR 3016 9936

21 TORN SISTERDALE TX (33 NNW SAT) 27/1930

HAM REPORTED POSSIBLE TORNADO ON GROUND

SAT/LSR 2996 9873

.............LRG HAIL/STRONG WIND RPTS..................LRG HAIL/STRONG WIND RPTS......

22 A275 BRUCEVILLE TX (14 NE TPL) 27/1305

FTW/LSR 3131 9723

23 A275 EDDY TX (13 NE TPL) 27/1306

FTW/LSR 3130 9725

24 A175 TROY TX (6 ENE TPL) 27/1325

FTW/LSR 3120 9730

25 A125 TATUM TX (25 ESE GGG) 27/1330

NWS COOP OBSERVER ON RUSK-PANOLA

COUNTY LINE REPORTED HAIL ON BOTH

SHV/LSR 3219 9433

26 A125 TATUM TX (17 S GGG) 27/1330

NWS COOP OBSERVOR ON RUSK-PANOLA

COUNTY LINE REPORTED HAIL ON BOTH

SHV/LSR 3213 9476

27 G52 ROUND ROCK TX (12 N AUS) 27/1545

SAT/LSR 3049 9768

28 A175 MEXIA TX (42 E ACT) 27/1553

FTW/LSR 3168 9650

29 G62 AUSTIN MUELLER MUNI ARPT TX

(0 E AUS) 27/1620

KAUS METAR PK WIND 02062/2220

SPC/LSR 3030 9769

30 G52 ROUND ROCK TX (12 N AUS) 27/1645

SAT/LSR 3049 9768

31 A175 DRIFTWOOD TX (23 WSW AUS) 27/1650

REPORTED BY SHERIFFS DEPT SAT/LSR 3011 9803

A-5

Chronology of Events at NWSFO Fort Worth, TX

From 5:00 am CDT through 6:00 pm CDT - Tuesday, May 27, 1997

Time Product Description

532 AM SPS SEVERE OUTLOOK FOR NORTH TEXAS...MODERATE RISK OF

SEVERE THUNDERSTORMS THIS AFTERNOON AND TONIGHT

GENERALLY SOUTHEAST AND EAST OF A KILLEEN, TO MERIDIAN

TO TEXARKANA LINE.

1100 AM SFD NORTH TEXAS FORECAST DISCUSSION. MORNING SOUNDINGS

SHOWING A STRONG CAPPING INVERSION OVER NORTH

TEXAS..HOWEVER, ABOVE THE CAP..A STEEP LAPSE LEADS TO

VERY STRONG INSTABILITY IF THE CAP IS BROKEN.

1130 AM ZFP BELL AND MCLENNAN COUNTIES. PARTLY CLOUDY WITH A 30

PERCENT CHANCE OF THUNDERSTORMS..SOME SEVERE AND

WITH LOCALLY HEAVY RAIN.

1250 PM SVR SEVERE THUNDERSTORM WARNING FOR MCLENNAN COUNTY.

A RAPIDLY DEVELOPING THUNDERSTORM LOCATED NEAR

WOODWAY CURRENTLY PRODUCING PEA TO MARBLE (SIZE)

HAIL...CONDITIONS ARE FAVORABLE FOR MUCH LARGER HAIL

BY 1 PM.

0.75 INCH HAIL REPORTED NEAR WOODWAY AT 110 PM.

1255 PM NOW FOR BELL AND MCLENNAN COUNTIES..SCATTERED

THUNDERSTORMS NOW IN PROGRESS WILL CONTINUE TO

DEVELOP AND WILL BE CAPABLE OF PRODUCING DAMAGING

WINDS AND LARGE HAIL.

121 PM TOR FOR MCLENNAN COUNTY. NWS DOPPLER RADAR INDICATES

TORNADO NEAR HEWITT.

126 PM SVS NWS RADAR INDICATES A TORNADO 3 MILES WEST OF LORENA

(MCLENNAN COUNTY). STORM SPOTTERS CONFIRM THIS

TORNADO..PEOPLE IN LORENA TAKE COVER!

SUBSEQUENT INFORMATION INDICATED A TORNADO, PROBABLY F1 AND

LASTING ABOUT 3 MINUTES, OCCURRED NEAR LORENA AT 120 PM. SEVERAL

MOBILE HOMES DAMAGED, ONE HOUSE DESTROYED.

132 PM SVR SEVERE THUNDERSTORM WARNING FOR FALLS COUNTY (JUST

SE OF MCLENNAN COUNTY).

142 PM TOR TORNADO WARNING FOR FALLS COUNTY. NWS DOPPLER RADAR

DETECTED A DEVELOPING TORNADO NEAR EDDY.

A-6

AT 145 PM, A TORNADO, CLASSIFIED AS F3, DID EXTENSIVE DAMAGE NEAR

MOODY, TX, IN MCLENNAN COUNTY. A TORNADO WARNING HAD BEEN

ISSUED AT 121 PM AND REMAINED IN EFFECT UNTIL 200 PM.

148 PM TOR TORNADO WARNING FOR BELL COUNTY. NWS DOPPLER RADAR

DETECTED A TORNADO BETWEEN EDDY AND MOODY (IN

MCLENNAN COUNTY). CONFIRMED BY SPOTTERS.

157 PM SVS TORNADIC THUNDERSTORM LOCATED FROM JUST SOUTH OF

MOODY TO THE TOWN OF NORTH PARIRIE. TORNADO

REPORTED BY SPOTTERS AND NOW ENTERING BELL COUNTY.

200 PM NOW SEVERE STORMS WILL MOVE SOUTH TO SOUTH SOUTHWEST.

TORNADO WATCH IN EFFECT.

REPORTS RECEIVED AT 205 PM AND 206 PM RESPECTIVELY OF 2.75 INCH

HAIL IN MCLENNAN AND FALLS COUNTIES.

200 PM SVR SEVERE THUNDERSTORM WARNING FOR HILL AND NAVARRO

COUNTIES. (THUNDERSTORMS PROMPTING THIS WARNING WERE

SEPARATE AND TO THE NORTHEAST OF THE TORNADIC STORM

IN MCLENNAN COUNTY)

216 PM SVS SPOTTERS AND RADAR CONTINUE TO TRACK TORNADO IN

NORTHERN BELL COUNTY NEAR HIGHWAY 317 AND

SOUTHERLAND ROAD AT 215 PM. IF YOU LIVE IN MORGANS

POINT RESORT..TAKE COVER IMMEDIATELY. STORM ALSO

PRODUCING VERY LARGE HAIL.

230 PM TOR EXTENSION OF TORNADO WARNING FOR BELL COUNTY. NWS

DOPPLER RADAR DETECTED A TORNADIC THUNDERSTORM

BETWEEN TROY AND BELTON. STORM MOVING SOUTH. RADAR

CONTINUE TO INDICATE CIRCULATION AND TORNADOES ARE

PROBABLE.

SUBSEQUENT REPORTS INDICATE A TORNADO CLASSIFIED AS F3 STRUCK

THE MOGANS POINT AREA OF LAKE BELTON AT 235 PM DOING EXTENSIVE

PROPERTY DAMAGE. TORNADO WARNING HAD BEEN IN EFFECT SINCE 148

PM WITH EXTENSION AT 230 PM.

237 PM SVS DOPPLER RADAR DETECTED TWO TORNADIC THUNDERSTORMS.

ONE CIRCULATION CENTER WAS NEAR BELTON AND THE OTHER

ONE WAS 5 MILES NORTH OF BELTON. STORMS MOVING SOUTH.

PEOPLE IN AND NEAR BELTON SHOULD TAKE COVER NOW!!!

300 PM TOR TORNADO WARNING FOR NAVARRO COUNTY. NWS DOPPLER

RADAR DETECTED A DEVELOPING TORNADO NEAR CORSICANA.

(AGAIN, THIS IS A SEPARATE STORM TO THE NORTHEAST OF THE

BELL COUNTY STORM)

A TORNADO WAS REPORTED NEAR BLOOMING GROVE IN NAVARRO

COUNTY AROUND 305 PM. NO SIGNIFICANT DAMAGE REPORTED.

A-7

305 PM SVS TORNADO WARNING IN EFFECT. NWS DOPPLER RADAR DETECTED A VERY

DANGEROUS STORM NEAR BELTON. THIS STORM HAS A HISTORY OF

PRODUCING TORNADOES WITH NUMEROUS REPORTS OF SUBSTANTIAL

DAMAGE BETWEEN KILLEEN AND TEMPLE...TAKE COVER AS THIS STORM

APPROACHES. IT WILL CONTAIN BASEBALL OR LARGER SIZE HAIL AND

PRODUCE OCCASIONAL TORNADOES.

309 PM SVS UPDATE ON TORNADO WARNING FOR NAVARRO COUNTY.

315 PM TOR TORNADO WARNING FOR BELL COUNTY EXTENDED UNTIL 4 PM.

NWS DOPPLER RADAR DETECTED AND STORM SPOTTERS

CONFIRMED A TORNADO JUST NORTH OF SALADO. STORM

MOVING SOUTH AT 15 MPH.

322 PM NOW SEVERE THUNDERSTORMS OVER NAVARRO AND BELL

COUNTIES WERE MOVING SLOWLY SOUTH. TORNADOES HAVE

BEEN REPORTED WITH THE STORM IN BELL COUNTY..AND THE

STORM IN NAVARRO COUNTY HAS THE POTENTIAL TO PRODUCE

A TORNADO. TAKE COVER AS THEY APPROACH.

328 PM SVR SEVERE THUNDERSTORM WARNING FOR HILL COUNTY. THIS

WAS REDEVELOPMENT TO THE NORTH OF THE TORNADIC

STORM IN BELL COUNTY.

336 PM TOR TORNADO WARNING FOR HILL AND LIMESTONE COUNTIES.

TORNADO REPORTED NEAR HUBBARD IN HILL COUNTY AT 350 PM. NO

SIGNIFICANT DAMAGE REPORTED.

343 PM TOR TORNADO WARNING FOR FREESTONE COUNTY...EXTENDED FOR

NAVARRO COUNTY. DOPPLER RADAR INDICATED A TORNADIC

STORM NEAR CORSICANA IN NAVARRO COUNTY. LARGE HAIL

ALSO LIKELY.

345 PM ZFP ROUTINE ZONE [FORECAST] ISSUANCE.

352 PM SVS UPDATE ON TORNADO WARNINGS FOR HILL, NAVARRO,

FREESTONE AND LIMESTONE COUNTIES.

421 PM SVS FURTHER UPDATE ON WARNINGS FOR HILL, NAVARRO,

FREESTONE AND LIMESTONE COUNTIES.

428 PM TOR EXTENSION OF TORNADO WARNING FOR LIMESTONE COUNTY.

430 PM NOW SHORT RANGE FORECAST ISSUED FOR ENTIRE COUNTY

WARNING AREA (SEGMENTED). SPECIFIC DETAILS PROVIDED ON

SEVERE STORMS IN COUNTIES AFFECTED.

458 PM SVS UPDATE ON TORNADO WARNING FOR LIMESTONE COUNTY.

A-8

505 PM SVR SEVERE THUNDERSTORM WARNING FOR FREESTONE COUNTY.

STORM WILL AFFECT TEAGUE...FREESTONE...KIRVIN..AND DONIE.

EXPECT LARGE HAIL.

0.75 INCH HAIL REPORTED NEAR MEXIA IN FREESTONE COUNTY AT 520 PM.

514 PM TOR TORNADO WARNING FOR LIMESTONE AND MCLENNAN

COUNTIES. NWS DOPPLER RADAR INDICATED A TORNADIC

THUNDERSTORM IN LIMESTONE COUNTY NEAR

GROESBECK...MOVING SOUTH. NO REPORTS OF TORNADOES

BUT RADAR SHOWS ROTATION...TORNADOES POSSIBLE.

525 PM NOW UPDATE ON POSSIBLE TORNADIC STORM NOW IN EASTERN

MCLENNAN COUNTY.

A-9

Chronology of Events at NWSFO Austin/San Antonio, TX

From 11:00 am through 11:00 pm CDT - Tuesday, May 27, 1997

Time Product Description

1115 AM SPS HAZARDOUS WEATHER OUTLOOK...SEVERE THUNDERSTORM AND

HEAVY RAIN EVENT DEVELOPING ACROSS SOUTH TEXAS AND THE HILL

COUNTRY THIS AFTERNOON AND TONIGHT...

1254 PM SEL TORNADO WATCH FOR PORTIONS OF EAST TEXAS (INCLUDES

WILLIAMSON COUNTY OF SOUTH CENTRAL TEXAS). VALID FROM 115 PM

CDT UNTIL 700 PM CDT.

109 PM ZFP UPDATED ZONE FORECAST FOR WILLIAMSON COUNTY AND CITY OF

GEORGETOWN WITH TORNADO WATCH UNTIL 7 PM.

110 PM NOW UPDATED SHORT TERM FORECAST FOR WILLIAMSON COUNTY AND CITY

OF GEORGETOWN WITH TORNADO WATCH UNTIL 7 PM. CURRENTLY NO

THUNDERSTORMS...BUT STORMS EXPECTED TO DEVELOP BY LATE

AFTERNOON.

215 PM PHONE CALL TO GEORGETOWN POLICE DEPARTMENT AND WILLIAMSON COUNTY

SHERIFFS DEPARTMENT ALERTING THAT SEVERE THUNDERSTORMS

WERE LOCATED IN BELL COUNTY...MOVING SOUTHWARD AND THAT

SEVERE WEATHER WAS VERY LIKELY IN THE NEXT HOUR TO HOUR AND

A HALF.

248 PM NOW UPDATED SHORT TERM FORECAST FOR WILLIAMSON COUNTY AND CITY

OF GEORGETOWN WITH TORNADO WATCH UNTIL 7 PM. SEVERE

THUNDERSTORM AT 245 WEST OF TEMPLE MOVING SLOWLY

SOUTHWESTWARD.

300 PM PHONE CALL TO AUSTIN EOC [EMERGENCY OPERATIONS CENTER] ALERTING THAT

TORNADIC THUNDERSTORMS IN BELTON WERE MOVING SOUTHWARD.

SEVERE WEATHER WAS HIGHLY POSSIBLE AROUND 400 PM.

330 PM TOR TORNADO WARNING VALID UNTIL 430 PM CDT FOR WILLIAMSON

COUNTY. STORM 5 MILES WEST OF ’JARRAL’ MOVING SOUTHEAST AT 10

MPH. CITY OF ’JARREL’ IS IN THE PATH OF THIS STORM.

AT APPROXIMATELY 340 PM CDT..AN F5 TORNADO STRUCK

JARRELL...KILLING 27. IT WAS ½ MILE IN WIDTH AND HAD A PATH

LENGTH ESTIMATED AT 5 TO 6 MILE TRACK.

334 PM NOW UPDATED SHORT TERM FORECAST FOR SEVERAL COUNTIES...

INCLUDING WILLIAMSON COUNTY AND CITY OF GEORGETOWN WITH

TORNADO WATCH UNTIL 7 PM. TORNADIC THUNDERSTORM JUST

NORTH OF 'JARREL' WITH ANOTHER DEVELOPING RAPIDLY NORTH OF

GEORGETOWN.

A-10

334 PM SVR SEVERE THUNDERSTORM WARNING VALID TIL 430 PM CDT FOR

EDWARDS COUNTY.

3/4 INCH HAIL REPORTED AT 415 PM BY POLICE DEPARTMENT AT

CARTA VALLEY. GOLFBALL-SIZED HAIL 26 MILES SSW OF

ROCKSPRINGS AT 425 PM.

331 PM SEL TORNADO WATCH FOR MUCH OF CENTRAL AND SOUTH CENTRAL TEXAS

UNTIL 10 PM CDT.

358 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 500 PM FOR REAL

COUNTY. STORM LOCATED ABOUT 15 MILES WEST OF LEAKY. HAIL

REPORTED NEAR BARKSDALE ON THE RIVER AT 410 PM.

400 PM ZFP ZONE FORECASTS FOR SOUTH TEXAS...INCLUDING WILLIAMSON

COUNTY AND CITY OF GEORGETOWN WITH TORNADO WATCH UNTIL 7

PM.

409 PM TOR TORNADO WARNING VALID UNTIL 515 PM CDT FOR TRAVIS

COUNTY. TORNADIC THUNDERSTORM OVER CEDAR PART AT 405 PM IS

MOVING SOUTH AT 10 MPH AND INTO THE NORTHERN PORTIONS OF THE

CITY OF AUSTIN.

AT 415 PM...A TORNADO IS REPORTED AT 2222 AND ANDERSON MILL

IN NORTHWEST TRAVIS COUNTY. AT 450 PM..AN F4 TORNADO

STRIKES NEAR LAKEWAY AND 620 IN THE PEDERNALES VALLEY

AREA...1 DEAD IN THIS TORNADO.

426 PM FFW FLASH FLOOD WARNING VALID UNTIL 630 PM FOR EDWARDS AND REAL

COUNTIES.

AT 500 PM...FLASH FLOODING IS REPORTED ALONG HIGHWAY 377

NEAR CARTA VALLEY.

431 PM NOW UPDATED SHORT TERM FORECAST FOR SEVERAL COUNTIES...

INCLUDING WILLIAMSON AND TRAVIS COUNTIES. TORNADO WATCHES

AND WARNINGS HIGHLIGHTED. TORNADO INDICATED OVER

NORTHWEST TRAVIS COUNTY NEAR FM2222 AND ANDERSON

MILL...MOVING SOUTH AT 20 MPH.

432 PM FFW FLASH FLOOD WARNING VALID UNTIL 630 PM CDT FOR TRAVIS AND

WILLIAMSON COUNTIES.

1 DROWNING IN SHOAL CREEK IN AUSTIN... TIME NOT AVAILABLE.

437 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 545 PM CDT FOR

WILLIAMSON COUNTY. HIGH WINDS AND HAIL TO BE EXPECTED IN THE

ROUND ROCK...GEORGETOWN...GRANDER AND TAYLOR AREAS.

AT 445 PM CDT...60 MPH WINDS REPORTED IN ROUND ROCK

449 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 545 PM CDT FOR

EDWARDS COUNTY. LARGE HAIL POSSIBLE BETWEEN BARKSDALE AND

CARTA VALLEY.

A-11

458 PM TOR TORNADO WARNING VALID UNTIL 600 PM CDT FOR HAYS COUNTY.

INCLUDED A REPORT OF A TORNADO AT LAKEWAY AND HIGHWAY 620

IN SOUTH TRAVIS COUNTY...MOVING TOWARD DRIPPING

SPRINGS...HAYS...BUDA...KYLE...WIMBERLY AND SAN MARCOS.

AT 538 PM CDT...TREES AND POWER LINES REPORTED BLOWN DOWN

IN THE KYLE AREA.

500 PM NOW UPDATED SHORT TERM FORECAST FOR ALL COUNTIES... HIGHLIGHTING

TORNADO WATCH. RADAR UPDATES ON STORMS IN BOTH WESTERN

AND CENTRAL COUNTIES AND THEIR MOVEMENTS.

506 PM SPS HAZARDOUS WEATHER OUTLOOK...TORNADO WATCH IN EFFECT

THROUGH 10 PM CDT. SYSTEM WILL EVOLVE INTO A HEAVY RAIN

EVENT LATER THIS EVENING.

519 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 615 PM CDT FOR

TRAVIS COUNTY- MOST DANGEROUS STORMS JUST EAST OF

PFLUGERVILLE AND BEECAVE. FUNNEL CLOUDS WERE BEING

REPORTED IN SOUTH AUSTIN.

AT 524 PM...WIND GUSTS TO 64 MPH RECORDED AT ROBERT

MUELLER AIRPORT IN DOWNTOWN AUSTIN.

524 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 630 PM CDT FOR

BLANCO - SEVERE THUNDERSTORM JUST SOUTHEAST OF JOHNSON CITY

NEAR HIGHWAY 290.

UNCONFIRMED REPORT BY PUBLIC OF A TORNADO NEAR BLANCO

AT 530 PM CDT.

531 PM NOW UPDATED SHORT TERM FORECAST FOR ALL COUNTIES... HIGHLIGHTING

TORNADO WATCH. RADAR UPDATES ON STORMS IN BOTH WESTERN

AND CENTRAL COUNTIES AND THEIR MOVEMENTS.

537 PM SVR/FFW SEVERE THUNDERSTORM AND FLASH FLOOD WARNING VALID UNTIL

715 PM CDT FOR UVALDE COUNTY. SEVERE THUNDERSTORM MOVING

SOUTH ALONG THE NUECES RIVER TOWARD MONTEL.

552 PM TOR TORNADO WARNING VALID UNTIL 700 PM CDT FOR KENDALL COUNTY.

SHERIFFS DEPARTMENT IN BLANCO OBSERVED A TORNADO

APPROACHING KENDALIA.

558 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 8 PM CDT FOR

BASTROP AND CALDWELL COUNTIES. LEADING EDGE OF SEVERE

THUNDERSTORMS FROM ELROY IN BASTROP COUNTY TO NEAR SAN

MARCOS

AT 635 PM... NICKEL-SIZED HAIL IS REPORTED IN ELGIN IN BASTROP

COUNTY.

A-12

601 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 700 PM CDT FOR HAYS

COUNTY. GOLFBALL SIZED HAIL HAD BEEN REPORTED IN DRIFTWOOD

WITH WIDESPREAD NICKEL SIZED HAIL.

AT 645 PM CDT...3/4 INCH HAIL IS REPORTED AT SAN MARCOS.

606 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 715 PM CDT FOR VAL

VERDE COUNTY. RADAR INDICATED A SEVERE THUNDERSTORM WITH

LARGE HAIL NEAR LOMA ALTA AND JUST NORTHEAST OF COMSTOCK.

AT 615 PM CDT...GOLFBALL-SIZED HAIL IS OBSERVED AT COMSTOCK.

614 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 715 PM CDT FOR

COMAL COUNTY. A SEVERE THUNDERSTORM WAS ENTERING COMAL

COUNTY NEAR FISHER AND WILL MOVE OVER CANYON LAKE IN THE

NEXT 15 MINUTES.

AT 645 PM CDT...DIME-SIZED HAIL IS REPORTED FROM CANYON

LAKE.

621 PM NOW UPDATED SHORT TERM FORECAST FOR ALL COUNTIES... HIGHLIGHTING

TORNADO WATCH. RADAR UPDATES ON STORMS IN BOTH WESTERN

AND CENTRAL COUNTIES AND THEIR MOVEMENTS.

624 PM TOR TORNADO WARNING VALID UNTIL 700 PM CDT FOR NORTHERN UVALDE

COUNTY. THE TORNADIC THUNDERSTORM WAS WEST OF UTOPIA.

AT 635 PM CDT...GOLFBALL SIZED HAIL IS OBSERVED AT SABINAL.

AT 645 PM CDT...THE PUBLIC REPORTS A FUNNEL ALONG FM1050. AT

655 PM...A REPORT OF A POSSIBLE TORNADO BETWEEN UTOPIA AND

SABINAL IS PASSED BY THE SHERIFF TO A HAM OPERATOR.

628 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 730 PM CDT FOR

EDWARDS COUNTY. THE THUNDERSTORM WAS NEAR HIGHWAY 377...15

MILES NORTHEAST OF ROCKSPRINGS. RADAR HAS INDICATED HAIL OF

UP TO 2 INCHES WITH THIS STORM.

629 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 930 PM CDT FOR

GUADALUPE COUNTY. THE STORM WAS MOVING SOUTHWARD INTO

THE NORTHERN PART OF THE COUNTY.

AT 640 PM CDT...REPORT IF SEVERE WINDS BLOWING A PLANE OVER

ON ITS TOP AT THE SEGUIN AIRPORT

629 PM SEL TORNADO WATCH FOR MUCH OF NORTH CENTRAL AND EAST TEXAS

UNTIL 1 AM CDT WEDNESDAY MORNING.

656 PM TOR TORNADO WARNING VALID UNTIL 8 PM CDT FOR KENDALL COUNTY.

TORNADIC THUNDERSTORM INDICATED JUST SOUTH OF SISTERDALE

MOVING TOWARD THE SOUTH AT 15 MPH. THE STORM IS MOVING

ALONG FM1367 TOWARDS BOERNE.

AT 730 PM CDT...HAMS REPORT A TORNADO AT SISTERDALE. NO

DAMAGE INDICATED AT THAT TIME.

A-13

659 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 830 PM CDT FOR

BEXAR COUNTY. SEVERE THUNDERSTORMS WITH HIGH WINDS MOVING

INTO NORTHERN BEXAR COUNTY EXTENDING FROM RANDOLPH AFB TO

BULVERDE AND BOERNE.

AT 803 PM CDT...A WIND GUST TO 122 MPH REPORTED AT KELLY AFB

IN SOUTHWEST SAN ANTONIO. DAMAGE WAS WIDESPREAD IN THE

SAN ANTONIO AREA...WITH POWER OUT TO NEARLY 100,000

PERSONS.

700 PM SVS SEVERE WEATHER STATEMENT SUMMARIZING THE TORNADO WATCH

VALID UNTIL 1 AM CDT...AND THE WARNINGS FOR

KENDALL...UVALDE...HAYS...COMAL...VAL VERDE...EDWARDS...

GUADALUPE...CALDWELL AND BASTROP COUNTIES.

705 PM RVS RIVER STATEMENT FOR BRUSHY CREEK IN WILLIAMSON COUNTY.

HEAVY RAIN BETWEEN 2 AND 4 INCHES HAS BEEN RECEIVED AND

BRUSHY CREEK WILL BE OUT OF BANKS FROM THE HEADWATERS

ABOVE ROUND ROCK TO DOWNSTREAM ABOVE THORNDALE THIS

EVENING.

706 PM ZFP ZONE FORECASTS FOR MANY SOUTH CENTRAL TEXAS COUNTIES...

INCLUDING WILLIAMSON COUNTY AND CITY OF GEORGETOWN WITH

TORNADO WATCH UNTIL 1 AM.

710 PM NOW UPDATED SHORT TERM FORECAST FOR ALL COUNTIES... HIGHLIGHTING

BOTH TORNADO WATCHES. RADAR UPDATES ON STORMS IN BOTH

WESTERN AND CENTRAL COUNTIES AND THEIR MOVEMENTS.

714 PM SVS SEVERE WEATHER STATEMENT INCLUDING TORNADO WATCH AND

WARNINGS FOR KENDALL...UVALDE...HAYS...COMAL...VAL VERDE...

EDWARDS... GUADALUPE...CALDWELL AND BASTROP COUNTIES.

710 PM RVS RIVER STATEMENT FOR RED BLUFF AND SYCAMORE CREEKS IN

EDWARDS AND VAL VERDE COUNTIES. RADAR INDICATES AN AREA OF

4 TO 5 INCHES OF RAIN IN THE PAST 3 HOURS IN EDWARDS COUNTY.

722 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 9 PM CDT FOR

EDWARDS...KERR AND VAL VERDE COUNTIES. AT 720 PM A SEVERE

THUNDERSTORM WAS BETWEEN DEL RIO AND LOMA ALTA IN VAL

VERDE COUNTY...MOVING EAST AT 10 MPH. M OER SEVERE

THUNDERSTORMS WERE NEAR HWY 277 IN NORTHWEST EDWARDS

COUNTY AND NEAR HWY 83 WEST OF MOUNTAIN HOME IN KERR

COUNTY.

AT 830 PM CDT...3/4 INCH HAIL REPORTED IN THE CITY OF DEL RIO

WITH .88 INCH HAIL AT THE DEL RIO AIRPORT. AT 836 PM CDT...THE

DEL RIO AIRPORT RECORDS A WIND GUST TO 61 MPH.

727 PM FFW FLASH FLOOD WARNING VALID UNTIL 1030 PM CDT FOR BLANCO AND

KENDALL COUNTIES. HEAVY RAINS OVER NORTHERN KENDALL AND

SOUTHERN BLANCO COUNTY ESTIMATED AT 2 TO 3 INCHES WITH

ISOLATED 3 TO 5 INCHES. EXTRA CAUTION URGED AT NIGHT.

A-14

REPORTS WERE LATER RECEIVED OF MOST ROADS IN SOUTHERN

BLANCO AND NORTHERN KENDALL COUNTIES CLOSED DUE TO

FLASH FLOODING BETWEEN 830 PM CDT AND 10 PM CDT.

734 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 830 PM CDT FOR

GONZALES COUNTY. SEVERE THUNDERSTORMS EXTENDING FROM

SMILEY NORTH TO BELMONT...WITH ANOTHER STRONG STORM NEAR

FLATONIA.

AT 735 PM CDT A FUNNEL WAS OBSERVED NEAR WRIGHTSBORO.

ALSO...AT 735 PM CDT...TREES WERE REPORTED BLOWN DOWN

BETWEEN LEESVILLE AND CHEAPSIDE. AT 815 PM CDT...1.5 INCH

HAIL FELL JUST NORTH OF THE TOWN OF NIXON.

738 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 915 PM CDT FOR FRIO

COUNTY. SEVERE THUNDERSTORM WITH LARGE HAIL NORTHWEST OF

PEARSALL NEAR HIGHWAY 140...MOVING SOUTHEAST AT 20 MPH.

AT 830 PM CDT...GOLFBALL-SIZED HAIL FELL IN PEARSALL WITH

QUARTER-SIZED HAIL AT 855 PM CDT REPORTED FROM DILLEY BY

HAM SPOTTERS.

739 PM SVS SEVERE WEATHER STATEMENT INCLUDING TORNADO WATCH.

WARNINGS CANCELED FOR CALDWELL AND BASTROP.

751 PM NOW UPDATED SHORT TERM FORECAST FOR ALL COUNTIES... HIGHLIGHTING

LINES OF THUNDERSTORMS IN TWO DIFFERENT ALIGNMENTS. RADAR

UPDATES ON STORMS IN ALL COUNTIES.

752 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 9 PM CDT FOR

MEDINA COUNTY. SEVERE THUNDERSTORM WITH LARGE HAIL

BETWEEN YANCEY AND DEVINE MOVING EAST AT 20 MPH. THIS STORM

WILL AFFECT DEVINE.

AT 830 PM CDT...3/4 INCH HAIL REPORTED IN DEVINE.

800 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 9 PM CDT FOR BEXAR

COUNTY. LINE OF THUNDERSTORMS MOVING THROUGH BEXAR

COUNTY WITH POSSIBLE LARGE HAIL...DAMAGING WINDS AND

LIGHTNING. RAINFALL RATES OF 1 TO 2 INCHES PER HOUR.

AGAIN...AS NOTED BEFORE..AT 803 PM CDT..A WIND GUST TO 122

MPH WAS RECORDED AT KELLY AFB IN SAN ANTONIO...ALONG WITH

WIDESPREAD DAMAGE AND POWER OUTAGES TO 100,000 PERSONS.

801 PM TOR TORNADO WARNING VALID UNTIL 9 PM CDT FOR GONZALES COUNTY.

RADAR INDICATED A TORNADO NEAR THE GONZALES

AIRPORT...MOVING SOUTHEAST AT 15 MPH.

AT 810 PM CDT...FUNNEL CLOUDS WERE SIGHTED NORTH OF THE

CITY OF GONZALES.

A-15

809 PM SPS SPECIAL WEATHER STATEMENT INDICATING THAT KIMBLE...

MASON...LLANO...BURNET AND WILLIAMSON COUNTIES HAVE BEEN

CLEARED FROM THE TORNADO WATCH.

803 PM NOW UPDATED SHORT TERM FORECAST FOR THE COUNTIES THAT HAD BEEN

CLEARED FROM THE WATCH.

813 PM NOW UPDATED SHORT TERM FORECAST FOR BURNET...LLANO AND

WILLIAMSON COUNTIES. STORMS HAVE MOVED SOUTHWARD AND NO

LONGER THREATEN THESE COUNTIES.

813 PM FFW FLASH FLOOD WARNING VALID UNTIL 1015 PM CDT FOR BANDERA AND

VAL VERDE COUNTIES. HEAVY RAIN FROM DEL RIO TO LOMA ALTA

HAVE PRODUCED BETWEEN 3 AND 5 INCHES BY RADAR ESTIMATION

SINCE 5 PM. ADDITIONAL RAINFALL IS EXPECTED THROUGH 9 PM.

BETWEEN 830 PM CDT AND 9 PM CDT... FLASH FLOODING WAS

REPORTED IN VAL VERDE COUNTY... ACROSS CITY STREETS IN DEL

RIO..AND ALONG HIGHWAY 90 FROM DEL RIO TO COMSTOCK.

818 PM TOR TORNADO WARNING VALID UNTIL 915 PM CDT FOR FRIO COUNTY BASED

ON A FUNNEL REPORTED BY THE SHERIFF IN PEARSALL. DEVELOPING

TORNADO BETWEEN PEARSALL AND DILLEY. SHERIFF HAS SPOTTED A

FUNNEL NEAR PEARSALL. THE STORMS ARE MOVING SOUTHEAST AT 20

MPH.

AT 820 PM CDT...THE SHERIFF REPORTED NICKLE-SIZED HAIL AS

WELL AS A TORNADO JUST NORTHEAST OF MOORE.

819 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 915 PM CDT FOR

ATASCOSA COUNTY. THUNDERSTORMS ENTERING NORTHERN

ATASCOSA COUNTY...MOVING EAST AT 20 MPH. WILL MOVE THROUGH

POTEET...JOURDANTON...AND PLEASANTON DURING THE NEXT HOUR.

AT 830 PM CDT...TREES WERE BLOWN OVER IN PLEASANTON...WITH

SOME BLOWN ONTO CARS CAUSING DAMAGE.

821 PM ZFP ZONE FORECAST UPDATE FOR SOUTH TEXAS COUNTIES GENERALLY

ALONG AND SOUTH OF A LINE FROM LAREDO TO VICTORIA.

823 PM SEL TORNADO WATCH FOR MUCH OF SOUTH CENTRAL AND SOUTHEAST

TEXAS UNTIL 3 AM CDT WEDNESDAY MORNING.

A-16

829 PM FFW FLASH FLOOD WARNING VALID UNTIL 1130 PM CDT FOR GONZALES

COUNTY. THUNDERSTORMS WITH HEAVY RAIN INDICATED OVER MUCH

OF GONZALES COUNTY. 2 TO 3 INCHES OF RAIN HAVE FALLEN OVER

THE PAST HOUR WITH AN ADDITIONAL 1 OR 2 INCHES POSSIBLE.

FLASH FLOODING WAS REPORTED WITH FM532 AND 1116 CLOSED

DUE TO HIGH WATER.

840 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 945 PM CDT FOR

DEWITT...KARNES...WILSON AND LAVACA COUNTIES. RADAR

INDICATED SEVERE THUNDERSTORMS FROM JOURDANTON TO KARNES

CITY TO CUERO AND HALLETTSVILLE...MOVING TOWARD THE

SOUTHEAST AT 20 MPH.

AT 925 PM CDT...TREES AND POWER LINES WERE REPORTED BLOWN

DOWN IN DEWITT COUNTY BY THE SHERIFF’S DEPARTMENT. LARGE

TREES WERE UPROOTED AND SPLIT BY HIGH WINDS IN THE LA

VERNIA AREA IN WILSON COUNTY NEAR 9 PM CDT.

851 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 10 PM CDT FOR

SOUTHERN VAL VERDE COUNTY. SHOWERS AND THUNDERSTORMS

CONTINUED TO DEVELOP OVER THE SOUTHERN PART OF THE COUNTY.

THE STORMS WERE MOVING SOUTHEAST AT 15 MPH.

AT 9 PM CDT...GOLFBALL-SIZED HAIL WAS REPORTED 6 MILES EAST

OF COMSTOCK.

855 PM ZFP ZONE FORECAST UPDATE FOR ALL SOUTH TEXAS COUNTIES TO

INCLUDE THE NEW WATCH VALID UNTIL 3 AM CDT WEDNESDAY

MORNING.

855 PM FLW FLOOD WARNING FOR RED BLUFF AND SYCAMORE CREEKS IN EDWARDS

AND VAL VERDE COUNTIES. CONTINUED RAINFALL OVER THE

HEADWATERS HAS BROUGHT RAINFALL AMOUNTS UP TO NEARLY SIX

INCHES IN MANY AREAS. RED BLUFF CREEK WILL BE WELL OUT OF

BANKS AND FLOODING SECONDARY AND PRIMARY ROADS. SYCAMORE

CREEK WILL ALSO BE WELL OUT OF BANKS AND FLOODING THE FLOOD

PLAIN AND ROADS.

858 PM FFW FLASH FLOOD WARNING VALID UNTIL 11 PM CDT FOR FRIO COUNTY.

RAINFALL OF NEARLY 2 INCHES PER HOUR ACROSS THE COUNTY...WITH

ADDITIONAL 1 TO 2 INCHES EXPECTED. THE STORMS WERE LOCATED

FROM MOORE TO PEARSALL TO DIVOT AND SOUTHWEST THROUGH THE

COUNTY.

NUMEROUS ROADS WERE CLOSED FROM 9 PM UNTIL 11 PM CDT

ACROSS FRIO COUNTY.

A-17

905 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 1030 PM CDT FOR

DIMMIT COUNTY. SEVERE THUNDERSTORM WITH VERY LARGE HAIL

BETWEEN BIG WELLS AND CARRIZO SPRING...MOVING EAST AT 10 MPH.

AT 915 PM CDT...3/4 INCH HAIL WAS REPORTED NEAR BRUNDAGE.

914 PM NOW UPDATED SHORT TERM FORECAST FOR COUNTY WARNING AREA WITH

LATEST RADAR LOCATIONS AND MOVEMENTS.

933 PM LSR LOCAL STORM REPORT SUMMARIZING EVENING SEVERE WEATHER.

939 PM FFW FLASH FLOOD WARNING VALID UNTIL 1245 AM CDT WEDNESDAY

MORNING FOR KARNES COUNTY. DOPPLER ESTIMATES OF 2 TO 3

INCHES OF RAIN IN THE NORTHERN PART OF THE COUNTY FROM

GILLETT TO HOBSON SINCE 8 PM. 1 OR 2 MORE INCHES POSSIBLE.

956 PM FFW FLASH FLOOD WARNING VALID UNTIL 1 AM WEDNESDAY MORNING CDT

FOR FAYETTE COUNTY.

958 PM FFW ZONE FORECAST UPDATE FOR SOUTHEAST TEXAS COUNTIES.

1000 PM NOW UPDATED SHORT TERM FORECAST FOR COUNTY WARNING AREA WITH

LATEST RADAR LOCATIONS AND MOVEMENTS.

1003 PM SVR SEVERE THUNDERSTORM WARNING VALID UNTIL 1145 PM CDT FOR VAL

VERDE COUNTY. SEVERE THUNDERSTORM NEAR LANGTRY. THIS

STORM INDICATED WITH LARGE HAIL...MOVING SOUTHEAST AT 10 MPH.

1015 PM SEL TORNADO WATCH FOR MUCH OF SOUTH CENTRAL TEXAS UNTIL 4 AM

CDT WEDNESDAY MORNING.

1035 PM ZFP ZONE FORECAST UPDATE FOR COUNTIES OF DEEP SOUTH TEXAS.

1051 PM ZFP ZONE FORECAST UPDATE FOR THE HILL COUNTRY AND SOUTH

CENTRAL TEXAS COUNTIES.

1052 PM SVS SEVERE WEATHER STATEMENT SUMMARIZING TORNADO WATCH AND

FLASH FLOOD AND SEVERE THUNDERSTORM WARNINGS VALID.

1054 PM NOW UPDATED SHORT TERM FORECAST FOR COUNTY WARNING AREA WITH

LATEST RADAR LOCATIONS AND MOVEMENTS.

1100 PM RVS RIVER STATEMENT FOR THE BLANCO...SAN MARCOS AND GUADALUPE

RIVERS. FORECASTS OF RIVER STAGES AND COMMENTS ON OVERBANK

AND BANKFULL RISES.

1145 PM LSR LOCAL STORM REPORT SUMMARIZING EVENING SEVERE WEATHER.

B-1

Appendix B

The Jarrell, Texas, Tornado Outbreak:

A Meteorological Satellite Perspective

James F. W. Purdom and Brian C. Motta

1,2 2

NOAA/National Environmental Satellite, Data and Information Service

Office of Research and Applications

Regional and Mesoscale Meteorology Team

and

Cooperative Institute for Research in the Atmosphere (CIRA)

Colorado State University

Fort Collins, Colorado

Authors’ note: This paper has been abridged in order to fit within the Service

Assessment report. For the complete text, contact either of the two authors at:

<James.Fw.Purdom@noaa.gov> or <motta@terra.cira.colostate.edu>.

Service Assessment Team Note: Much of the data and many of the techniques described

within this research report were not available in real-time to the NWSFO FWD and EWX

forecasters on May 27, 1997.

Introduction

This note describes thunderstorm development and evolution that may be observed using NOAA

meteorological satellite data over the Texas area on May 27, 1997, the day of the Jarrell tornadic

storm event. A number of interesting features were clearly discernable using satellite imagery.

Those features are discussed in this report along with image examples when appropriate.

Large-scale features, such as cumulus cloudiness along the frontal zone on which the major

tornado complex formed and evolved, are easily detected using GOES imagery from early in the

day until late afternoon when the low cloudiness becomes obscured by anvil cirrus. Products

derived from GOES sounder data reveal the very unstable air on which the tornadic storm

complex fed. On the cloud scale, satellite-based anvil-top characteristics that are often associated

with severe thunderstorms are apparent in satellite imagery. When the imagery is viewed in a

storm-relative mode and tornadic tracks are added to the sequence, it is evident that the major

B-2

tornadic activity was associated with a thunderstorm complex that evolved to the south along the

frontal zone.

Data Sources

GOES-8 and GOES-9 data were acquired by CIRA’s direct receive station where they were

processed for analysis on a Regional Atmospheric Modeling System (RAMM) Advanced

Meteorological Satellite Demonstration and Interpretation System (RAMSDIS) workstation.

Both GOES satellites were in a routine mode of operation during the entire episode. Imagery was

available at 15-minute intervals with the exception of once every 3 hours when a 30-minute full

disk image was taken. Unfortunately, the satellite was not placed in a 7 ½-minute severe storm

observing mode during the event, nor had a special 1-minute interval imaging day been requested

for research purposes. GOES-8 sounder products were obtained from the NESDIS Forecast

Product Development Team’s Home page, accessible via Internet. The Advanced Very High

Resolution Radiometer (AVHRR) imagery from the polar-orbiting NOAA-14 satellite 1957

Coordinated Universal Time (UTC) ascending node pass was obtained from the National Climatic

Data Center. [Note: Subtract 5 hours from UTC to get CDT.] A complete set of surface and

upper air observations were obtained from the Cooperative Institute for Meteorological Satellite

Studies (CIMSS), CIRA’s sister Institute at the University of Wisconsin. Tornado track

information was extracted from a preliminary NOAA report that was based on aerial damage

surveys. Times for the tornadoes were based on estimates from members of the NOAA Service

Assessment Team. Track information is very accurate, to within a few hundred feet, while time

information may be off by several minutes, depending on the portion of the track with which time

is associated.

B-3

Figure 1. Eight kilometer resolution 6.7 m image from GOES-9 displayed at the

4 km scale. The image was taken at 1200 UTC on May 27, 1997. The 200 mb winds and

streamlines are plotted on the image

Pre-Storm: Synoptic and Regional Scale Analysis

The 6.7 micrometer ( m) channel on GOES senses emission from upper-level water vapor and

high cloud; because of this, 6.7 m imagery is very useful for revealing upper-level flow patterns

and for analysis of mesoscale features within the larger scale flow. As might be expected,

animated 6.7 m channel imagery very nicely revealed upper-level flow at the 300-200 mb level

(Figure 1).

When Figure 1 is viewed as part of an animated image sequence, the circulation around the upper-

level low in Nebraska and the location and movement of the upper-level trough and jet stream are

all well defined. Both 6.7 and 10.7 m imagery show the development of a large MCS over

Arkansas and eastern Oklahoma during the night (Figure 2).

B-4

Figure 2. Four kilometer resolution 10.7 m image from GOES-8, taken at 0915 UTC

on May 27, 1997. Temperature (in Celsius [C]) as it relates to the gray scale is shown at

the bottom of the image. The cloud shield of the large mesoscale convective cloud system,

referred to in the text, covers most of Arkansas.

This MCS moved slowly eastward and produced gravity waves that radiated away from the

system. One of those gravity waves appears to have played a role in the initiation of the

thunderstorm complex that eventually produced the tornado activity; however, that gravity wave

moved south of the storm complex before the beginning of tornado activity. In advance of the

upper trough, both 6.7 and 10.7 m imagery showed numerous patches of cirrus move across

Mexico and into Texas during the nighttime and into the early afternoon.

(Figure 3 and its discussion are not part of this abbreviated report.)

B-5

Figure 4a. One kilometer resolution visible image from GOES-8, taken at 1645 UTC on

May 27, 1997. Surface observations are plotted on the image

From early in the day, combining surface observations with GOES visible imagery allowed precise

positioning of the surface front—evident as a northeast- to southwest-oriented line of cumulus

coincident with the surface low pressure trough and cyclonic wind shift in Figure 4a. Details in

the front’s orientation and movement were easily monitored by animating GOES visible imagery

on RAMSDIS until anvil cirrus, associated with thunderstorms along the front, obscured the

satellite’s view. In Figure 4a, notice the well-defined line of organized convection that extends

from northeast to southwest along the frontal convergence zone, from A-B-C-D. Initiation of

deeper cumulus that eventually develop into the main tornado-producing thunderstorm complex is

underway at location B. At B, a broad area of cumulus intersects the frontal cloud band, from B-

E-F. The cloudiness from B-E-F is associated with a gravity wave that emanated from the MCS

shown in Figure 2.

B-6

Figure 4b. One kilometer resolution visible image from GOES-9, taken at 1715 UTC on

May 27, 1997

Figure 4b, from GOES-9 and taken 30 minutes after the image in Figure 4a, shows that deeper

convection has initiated at B and that the gravity wave has moved to the south of the convective

genesis area. Notice the difference in viewing perspective between GOES-9 and GOES-8.

Because there is very little deep convection or thick cirrus, low-level cloud details are seen

equally well in both images.

B-7

Figure 5. GOES sounder derived LI values ( C) for 1500 UTC on May 27, 1997,

plotted over the 1615 UTC NOAA GOES-8 visible satellite image. The gray scale

represents ranges of LI values. LI’s greater than 0 indicate stable atmospheric conditions;

less than 0, unstable. LI values less than -10 C indicate extremely unstable conditions.

Experimental GOES-8 satellite soundings are generated by the NESDIS Forecast Products

Development Team on an hourly basis, using GOES-8 sounder data, surface observations and

information from the Eta numerical forecast model. These experimental products are available via

Internet, from which Figure 5 was extracted. The Lifted Index (LI) values shown in Figure 5

highlight strong instability along the frontal boundary, with values of -11 C to -13 C in the

o o

Jarrell, Texas, region. Sounder products from later in the day were consistent with what is shown

in Figure 5; very unstable air was available to fuel intense storms in the vicinity of the frontal

zone. It should be noted that sounder calibration problems during this time may have resulted in

boundary layer moisture values that were too high, and thus the extreme LI values. However, the

relative values and gradients of instability shown in Figure 5 indicate the potential of GOES

sounder products for severe storm nowcasting. It is important to realize that soundings derived

from GOES sounder data are only made for clear or partly cloudy conditions.

B-8

Figure 6a. One kilometer resolution visible image from GOES-8 displayed at a

0.25 km scale. The image is centered at 31 17’54" North and 97 20’07" West, the

beginning point of the Moody, Texas, tornado track. The time for the scan line at picture

center is 18:49:10 UTC. Both beginning and ending points of ground (*) and cloud top (+)

locations of tornado track are shown. Note the adjustment that must be made due to parallax

to associate features at the ground to those at cloud top.

Storm Scale Analysis

For the storm scale analysis, all GOES figures are centered at the beginning location of the

tornado track that corresponds to a particular image and image times are for that center scan line.

When viewing GOES imagery, the observer must keep in mind that three dimensional features are

being observed over a curved Earth, while being displayed on a two dimensional flat surface. This

means that when locating features with respect to a surface-based grid, features at the surface

will be accurately located while features above the surface will be mislocated. The mislocation

will be radially away from the satellite subpoint and will increase as cloud height increases. This

parallax shift is illustrated in Figure 6a, where the ground tornado track is located between

asterisks, and the track with respect to cloud top is located between plus marks. (Note: For the

sake of brevity, only GOES-8 imagery is used. Refer to the author’s original paper for the

comparison of GOES-8 and GOES-9 imagery.) For all infrared imagery, the gray scale bar

versus temperature (Celsius) relationship at the bottom of Figure 6b is valid.

B-9

Figure 6b. Four kilometer resolution GOES-8 infrared image matching the visible image in Figure 6a.

Minimum cloud top temperature in the overshooting top area is -66.7 C. Beginning and end points of the

tornado track (+) are mapped to a cloud top height of 15 km.

The visible images reveal plumes of veil-like cirrus cloud extending downwind from the

thunderstorm’s overshooting top. Above-anvil cirrus plumes are often observed with very intense

and long-lived thunderstorms. In most of the corresponding infrared images, an enhanced V

signature (a V-shaped region of colder cloud extending downwind from the overshooting top)

was evident; such signatures are common with very intense thunderstorms. Strong overshooting

top activity was evident with the main tornado-producing thunderstorm area throughout its life.

The Moody, Belton and Jarrell, Texas, tornadoes (Figures 6, 7 and 8) were from the same

thunderstorm complex. When GOES imagery is animated in a storm-relative mode with tornado

locations superimposed, it is evident that the tornadoes occurred in almost exactly the same

location—on the west side of the storm and associated with the coldest overshooting top. During

the Moody to Jarrell portion of the storm’s life, the overshooting top region of the storm

developed southwestward (from 31 at 10 meters per second). This motion is a result of the

storm developing south along the northeast to southwest oriented front which is moving gradually

B-10

Figure 7a. One kilometer resolution visible image from GOES-8 displayed at a

0.25 km scale. The image is centered at 31 08’55" North and 97 28’17" West, the

beginning point of the Belton, Texas, tornado track. The time for the scan line across picture

center is 19:34:16 UTC. Both beginning and ending points of ground (*) and cloud top (+)

locations of the tornado track are shown.

eastward. At the time of the Jarrell tornado, a new cell was rapidly developing along the front to

the south of Jarrell. The overshooting top associated with the new storm is evident at the

southwest end of the anvil in Figure 8. (Figure 9 and its discussion are not part of this

abbreviated report.) During the next 1 ½ hours, that new storm produced the Pedernales Valley

tornado. Storm-relative animation revealed that at 2100 UTC the storm changed velocity,

moving faster and in a more westerly direction. That change in storm motion coincides with a

change in orientation of the frontal boundary, to a more east-to-west orientation, as is evident at

point C in Figure 4. This change in direction helps explain the near east-to-west tornado track at

Pedernales Valley. As with the earlier activity, those tornadoes were located in close proximity to

the overshooting/cold cloud top area.

B-11

Figure 7b. Four kilometer resolution GOES-8 infrared image matching the visible image

in Figure 7a. Minimum cloud top temperature in the overshooting top area is

-70.2 C. Beginning and end points of the tornado track (+) are mapped to a cloud top height

of 15 km.

B-12

Figure 8a. One kilometer resolution visible image from GOES-8 displayed at a

0.25 km scale. The image is centered at 30 53’54" North and 97 35’12" West, the

beginning point of the Jarrell, Texas, tornado track. The time for the scan line at picture

center is 20:50:45 UTC. Both beginning and ending points of ground (*) and cloud top (+)

locations of the tornado track are shown. Note the shadow immediately downwind of the

overshooting top (almost traced out by a line between the southwestern most * and +

marks); this overshooting top is associated with the Jarrell storm’s updraft at cloud top.

B-13

Figure 8b. Four kilometer resolution GOES-8 infrared image matching the visible image

in Figure 8a. The minimum cloud top temperature is -72 C and corresponds to the bright

overshooting top area in the visible image in Figure 8a. This cloud top area was

6 degrees warmer 20 minutes prior to this image. Beginning and end points of the tornado

track (+) are mapped to a cloud top height of 15 km.

B-14

Figure 10a. One kilometer resolution visible image from GOES-8 displayed at a

0.25 km scale. The image is centered at 30 23’21" North and 98 00’45" West, the

beginning point of the Pedernales Valley tornado track. The time for the scan line at picture

center is 22:04:20 UTC. Both beginning and ending points of ground (*) and cloud top (+)

locations of the tornado track are shown.

B-15

Figure 10b. Four kilometer resolution GOES-8 infrared image matching the visible

image in Figure 10a. Minimum cloud top temperature in the overshooting top area is

-75 C. Beginning and end points of the tornado track (+) are mapped to a cloud top height

of 15 km.

B-17

Figure 12. Minimum overshooting cloud top temperature for the tornadic storm complex

versus UTC for GOES-8.

The relationship between overshooting tops, cloud top temperature and storm updraft intensity is

fairly well understood. It is instructive to inspect plots of minimum cloud top temperature for the

tornadic storm complex as a function of time for GOES-8 during the outbreak period. This is

done in Figure 12. It is interesting that the cloud top temperatures closest to the Moody, Belton

and Jarrell tornado times are all 4 to 5 C colder than a smooth curve fitting minimum cloud top

temperatures at non-tornado times. This apparent relationship is not as clear for the Cedar Park

or the Pedernales Valley storms.

B-16

Figure 11. GOES visible upper left; AVHRR visible upper right; GOES infrared lower

left; AVHRR lower right. The GOES imagery was taken near 2004 UTC while

the AVHRR imagery was taken approximately 1 minute earlier.

Between the times of the Belton and Jarrell tornadoes, NOAA-14 provided AVHRR imagery over

the tornado outbreak area. While GOES and AVHRR visible channel imagery is of comparable

resolution, AVHRR’s 10.7 m infrared channel provides 1-km resolution imagery versus the 4

km resolution available from GOES. The four panel image in Figure 11 may be used to compare

visible and 10.7 m imagery from GOES with that from AVHRR. The GOES and AVHRR

visible images have been enhanced to show detail at cloud top, while the same color enhancement

is used for the infrared images. Notice how well the enhanced GOES and AVHRR visible

imagery define cloud top features—the veil-like cirrus plume above the anvil can be seen

extending downwind from well-defined overshooting tops. In the corresponding infrared images,

those same features are less well defined in GOES imagery than in AVHRR imagery. Cloud top

temperature structure is better defined in the AVHRR imagery, with AVHRR minimum

temperatures in the overshooting top region approximately 10 C colder than in the GOES

imagery. In the AVHRR imagery, relatively warmer anvil downwind from the overshooting top

area is clearly surrounded on both downwind sides of the anvil by colder cirrus—the “enhanced-

V” signature. This comparison of AVHRR and GOES points to the potential for use of AVHRR

imagery as a companion to GOES imagery for convective nowcasting purposes.

B-18

Conclusions

Satellite data provided a number of valuable pieces of information for use in severe thunderstorm

nowcasting over Texas on May 27, 1997. The storm complex that produced the tornado activity

between Moody and Pedernales Valley on this day had features at cloud top that are often

associated with thunderstorms that have very intense updrafts: (a) overshooting tops in the visible

imagery; (b) veil-like cirrus plumes above the anvil extending downstream from the overshooting

top area; (c) cold top areas associated with the overshooting region that were colder than the

ambient anvil temperature; and (d) enhanced “V” shaped cold regions with warm wakes extending

downwind from the cold top area. It should be noted that cold cloud top features similar to these

were noted in other intense thunderstorm complexes that produced intense weather over Texas on

this day. The precise location, orientation and movement of the front along which the tornadic

storm complex evolved was easily monitored using GOES imagery.

Real-time data flow into a satellite analysis system with capability similar to RAMSDIS would

have allowed for a number of useful applications. Storm-relative animation would have revealed

that tornadic activity was confined to one major thunderstorm complex, highlighting the threat to

any who lay in the path of the storm and its intersection with the front. Imagery could have been

enhanced, revealing cloud top characteristics commonly associated with severe thunderstorm

activity.

As shown in Figure 12, there appears to be a relationship between storm intensity, as revealed by

cloud top temperature, and tornado activity. However, it is not possible to verify this with the 15-

minute imagery, because storm dynamics with intense tornadic storms such as these are very

strong, and explosive overshooting at cloud top may last for only a few minutes. It is indeed

unfortunate that the satellite was not placed in a 7 ½-minute severe storm observing mode during

the event, nor had a special 1-minute interval imaging day been requested for research purposes.

The AVHRR imagery should be included as a part of routine real-time satellite imagery provided

to field users; it also gives an enticing glimpse to what might be observed with higher resolution

infrared imagery from geostationary satellites.

C-1

Appendix C

Aerial Damage Survey of the

Central Texas Tornadoes

of May 27, 1997

Brian E. Peters

Warning Coordination Meteorologist

National Weather Service, Birmingham, Alabama

Introduction

During the afternoon of May 27, 1997, severe thunderstorms spawned a series of tornadoes over

portions of Central Texas from just south of Waco to west of Austin. As part of the NOAA

Service Assessment Team investigating the storm event, an aerial damage survey was conducted.

The intent of the aerial survey was to provide information to pinpoint the beginning and ending

points of the tornado tracks, as well as provide information on the Fujita Tornado Intensity Scale

(F-scale) along the track.

The aerial survey was conducted on Friday and Saturday, May 30-31, using a fixed-wing aircraft

flying at approximately 3,000 feet and a helicopter flying between 500 and 1,000 feet. Additional

aerial surveying was done on Monday, June 2, using a helicopter arranged by the Travis County

Emergency Management Agency and the Fire Marshal’s office. A total of about

9 hours of air time was logged in the flights. Some of that air time was spent in getting to and

between the various tornado tracks. Additional assistance was obtained by using a Global

Positioning System (GPS) and mapping software to accurately plot damage locations. The

GPS/software combination claims an accuracy of ±300 feet.

THE MOODY TORNADO

Maximum F-scale: F3

Path Length = 3.7 miles Path Width = 150 yds

Path Length by County: McLennan [3.2 miles]; Bell [0.5 miles]

Beg: 31 17.91’/97 20.16’ End: 31 14.91’/97 21.48’

The Moody tornado began on the southside Farm to Market Road (FM) 107, 1.6 miles east-

southeast of Moody in open country. The tornado traveled south-southwest moving across

Dowell Road where two structures, a house and a barn, were destroyed. Another structure, a

small house, was damaged just north of this location. A pickup truck and a car were each tossed

several hundred feet; the pickup truck to the left of the tornado path and the vehicle along the

path. The tornado continued south-southwest from Dowell Road, crossing the McLennan/Bell

County line covering open country with trees down in numerous locations.

C-2

Ground Track of the Jarrell Tornado

Maximum F-scale assignment was based upon the damage that occurred at only one location,

where the house and barn were destroyed. There was an absence of other structures to make an

effective assessment.

THE BELTON TORNADO

Maximum F-scale: F3

Path Length = 1.4 miles Path Width = 275 yds

County: Bell

Beg: 31 08.92’/97 28.29’ End: 31 08.01’/97 29.19’

The tornado began 6.5 miles north of Belton in an area called Morgan’s Point on the northside of

Belton Lake. The tornado moved from land along a slough with tree damage on the west side of

the slough. A number of structures also on the west side of the slough sustained damage which

appeared to be minor. Continuing south-southwest, the tornado crossed a bend in Belton Lake,

moving ashore just northeast of a community called Woodland. As the tornado moved ashore,

destruction to trees was nearly total, with substantial damage to at least six structures.

The tornado ended abruptly only 1/3 of a mile after coming ashore at a point 5.6 miles north-

northwest of Belton.

THE JARRELL TORNADO

Maximum F-scale: F5

Path Length = 7.6 miles Path Width = 3/4 miles

County: Bell [2.4 miles]; Williamson [5.2 miles]

Beg: 30 53.90’/97 35.20’ End: 30 49.18’/97 40.12’

The damage path associated with the Jarrell tornado actually begins in Bell County at a point

about 0.8 miles northwest of the Prairie Dell exit from Interstate 35, near mile marker 280. The

C-3

tornado tracked south-southwestward across open country. Damage was primarily to trees, with

only a couple of structures damaged.

The tornado remained primarily in open country as it crossed the Bell/Williamson County line

before taking a more southwestward turn that took it to the northwestern edge of Jarrell. The

tornado crossed the county line very close to where Williamson Road ends and County Road

(CR) 304 (Bell County) begins. The tornado crossed CR 308, CR 305, and then CR 307. Where

the tornado crossed each of these county roads, approximately 525 feet of asphalt was ripped off

each of the roadways. This particular destruction was believed to be very close to the centerline

of the tornado circulation.

Observations recounted by eyewitnesses and revealed in interviews made by other members of the

Service Assessment Team indicated that the damage path may not have been made strictly by one

tornado. A number of eyewitnesses reported seeing several small, rope-like funnels before the

character of the tornado changed drastically into the killer tornado.

As the tornado crossed the intersection of CR 305 and 307, a business on the corner was

destroyed. The tornado moved into the Double Creek Estates subdivision at this point with total

destruction. F5 destruction begins shortly after the tornado moved into Williamson County and

continued until very close to the end of the damage path.

Moving through the Double Creek Estates subdivision and the surrounding area, the tornado

widened to the maximum width of three-quarters of a mile. From the air, the ground appearance

changed abruptly in the vicinity of CR 308 and continued until very near the end of the path. No

definitive circulation patterns or suction spots were evident, but there was the noted obvious

change in the appearance of the ground.

In the Double Creek area, approximately 40 structures were totally destroyed. One of the most

striking signs in approaching this area was the distinct lack of debris of any size. Closer

inspection showed lots of little debris, but no sign of large items. At least half a dozen cars were

identified from the air lying in the open areas, most of them flattened and encrusted with mud and

grass. Later, a ground survey revealed that most of the debris that was left in the area was

extremely small, indicating the power of the tornadic winds. All 27 deaths associated with the

Jarrell tornado occurred in the Double Creek area.

After passing through the Double Creek area, the tornado moved across CR 309 and into a

heavily wooded area of cedar trees. The total destruction of the tornado ends abruptly shortly

after entering the wooded area. However, a small swath of tree damage on the north side of the

main damage path suggested the possibility of a multiple vortex pattern. No other evidence of

multiple vorticies was observed.

C-4

Ground Track of the Cedar Park Tornado

THE CEDAR PARK TORNADO

Maximum F-scale: F3

Path Length = 9.2 miles Path Width = 250 yds

County: Williamson [5.6 miles]/Travis [3.6 miles]

Beg: 30 33.39’/97 49.24’ End: 30 27.30’/97 54.11’

The Cedar Park tornado began about 3.5 miles north of Cedar Park, at a location 0.6 miles south

of CR 178 and 1.4 miles east of the intersection of US 183 and CR 178. The initial damage was

to trees, however, the ground survey revealed damage nearby to a church and a trucking

company. The aerial survey did not reflect this damage as being in line with the damage path. It

is quite possible this damage was caused by strong wind near the tornado. The beginning point

was in a relatively open area, with damage primarily to a few trees and minor shingle damage to

one house.

The tornado moved south-southwestward, skirting a residential area before it crossed CR 180

immediately east of US 183. It was at this intersection that an Albertsons grocery store was

severely damaged. The tornado crossed US 183, causing additional damage to a number of

businesses. One business on the west side of US 183 lost nearly the entire roof. Most damage to

other businesses was believed to be minor.

A historic train located on the north side of CR 180, just to the east of US 183, was in the direct

path of the tornado. While the engine remained on the track, a coal tender converted to hold

diesel fuel and weighing approximately 65,000 pounds, including the 1,000 gallons of diesel fuel,

was flipped over and thrown a short distance.

After crossing US 183, the tornado moved across Marquis Lane and North Park Circle through

an area with widely scattered housing and a relative abundance of trees. Most damage to

structures in this area was minor.

C-5

From North Park Circle, the tornado moved into the northwestern portion of Buttercup Creek, a

subdivision of well constructed homes. Damage to homes was irregular, with one house losing a

roof but the house next door losing only shingles. Damage ranged from F0 to F2. At this point,

the tornado track was taking a gentle right turn and became more southwesterly. The tornado

moved into a wooded area crossing into Travis County before ending 1.1 miles from Lake Travis.

Damage in the wooded area was irregular, ranging from near total destruction of all trees to

sections with about 10 percent of the trees down.

C-6

Ground Track of the Pedernales Valley Tornado

THE PEDERNALES VALLEY TORNADO

Maximum F-scale: F4

Path Length = 5.6 miles Path Width = 440 yds

County: Travis

Beg: 30 23.35’/98 00.75’ End: 30 22.39’/98 06.25’

The Pedernales Valley tornado began on the shore of Lake Travis, destroying trees and a floating

marina where nearly all of the watercraft were destroyed. While numerous trees were twisted and

uprooted in this area, several structures sustained only what appeared to be minor damage that

would be no more than F0.

The tornado was initially on a heading of 265 degrees as it moved into rough terrain. A number

of structures sustained varying damage until the tornado reached Bee Creek Road. At that

location, a Southwest Bell building housing telephone switching equipment was destroyed. The

building was well-constructed and was one of several buildings which indicated at F4 rating for

this tornado. Bee Creek Road takes a bend close to the telephone building, and across the street a

house was destroyed with walls knocked down.

Approximately 2.2 miles from the lake, the tornado path takes a distinct dog-leg turn toward the

southwest. The point at which this turn occurred also corresponds with a knoll. Trees and

buildings at the top of the knoll were destroyed.

After the turn, the tornado assumed a heading of 250 degrees and crossed a major power

distribution line. One steel tower was destroyed, bringing all lines to the ground. The tornado

remained on the 250 degree heading, moving through the area described as the Hazy Hills

subdivision. Numerous houses and several mobile homes were totally destroyed. Several houses

survived but sustained major damage that made them totally uninhabitable.

C-7

The only death associated with this tornado occurred here, when one man was killed. He lived in

a mobile home that was demolished and his vehicle was tossed several hundred feet. Service

Assessment Team members were unable to learn whether he was in the mobile home during the

storm or had left it to drive away.

The tornado continued west-southwest, moving across State Road (SR) 71. A number of well-

built homes in the Hazy Hills subdivision were heavily damaged or destroyed. Crossing SR 71,

the tornado moved into another subdivision with widely-separated houses in the Lick Creek

valley, a steep walled creek that feeds into the Pedernales River. One stone-walled house located

just north of Pedernales Drive and west of SR 71 was completely deroofed. Other structures in

this subdivision sustained roof damage in the F2 range. After following the terrain into the cree,

the tornado climbed another rise in the land before ending shortly after passing the crest of the

small hill. As the tornado ended, damage was minimal to trees.

D-1

Appendix D

Persons Interviewed and/or Contacted

by the Service Assessment Team

Name Date Affiliation

John Sneed 05/29/97 Williamson County Sheriff’s Department

Shawn Newsom 05/29/97 Williamson County Sheriff’s Department

Mike Faught 05/29/97 Georgetown Fire Department

Mark Moss 05/29/97 Chief, Georgetown Fire Department

Kevin McEleney 05/29/97 Jarrell Volunteer Fire Department

Larry Fore 05/30/97 Albertsons grocery store manager

Tim Travis 05/30/97 Cedar Park Police Department

Paul Christ 05/30/97 Austin Police Department

Pat Pingston 05/30/97 Post Net store owner

Nick Olarti 05/30/97 Taco Bell store manager

Bob Russell 05/30/97 Cedar Park fire department

Mrs. & Mrs. Jim McCarty 05/30/97 Buttercup Creek residents

Ken Van Rens 05/30/97 Chief, Pedernales Valley Volunteer

Fire Department

Steve Poulson 05/30/97 Pedernales Valley Volunteer Fire

Department

David & Ann Butler 05/30/97 Hazy Hills residents

Scott Beckwith 06/01/97 Jarrell resident

Cliff Tschoerner, Jr. 06/01/97 Jarrell Volunteer Fire Department

Joyce Wilson 06/01/97 Double Creek Estates subdivision resident

LaDonna Peterson 06/01/97 Double Creek Estates subdivision resident

Jaunitta Peterson 06/01/97 Double Creek Estates subdivision resident

Bonnie Hammett 06/01/97 Double Creek Estates subdivision resident

Ken Adams 06/01/97 Double Creek Estates subdivision resident

Steve Collier 06/02/97 Austin Office of Emergency Management

Eric Carter 06/02/97 Austin Office of Emergency Management

Joe Hidrogo 06/02/97 Austin Office of Emergency Management

Scott Swearengen 06/02/97 Austin Office of Emergency Management

Clay Shell 06/02/97 Georgetown Fire Department

Bob Rose 06/02/97 Lower Colorado River Authority

Geniva Simpson 06/02/97 Williamson County Sheriff’s Office

Richard Gasten 06/02/97 Travis County

Brad Beauchamd 06/02/97 Travis County

Pete Baldwin 06/02/97 Travis County Emergency Management

Chris Causen 06/02/97 Travis County Emergency Services

Bobby Dewsman 06/02/97 State Office of Emergency Management

D-2

Jo Moss 06/02/97 State Office of Emergency Management

Linda Moorer 06/02/97 State Office of Emergency Management

Steve Vaughan 06/02/97 State Office of Emergency Management

Ken Lott 06/02/97 Austin American Statesman newspaper

Clara Herrera 06/02/97 Austin American Statesman newspaper

Larry Kolvoor 06/02/97 Austin American Statesman newspaper

Dave Martis 06/02/97 KVET/KASE radio

Michael Schmid 06/02/97 Texas Association of Broadcasters

David Gaiewski 06/02/97 Time Warner Communications

George Warmingham 06/02/97 Austin Cablevision

Kenneth Pulliam 06/02/97 KDFW-TV

Cliff Morrison 06/02/97 KEYE-TV

Andrew Sowder 06/02/97 KTBC-TV

Kimberly Flemming 06/02/97 KTBC-TV

Mike Wenglar 06/02/97 KVUE-TV

Keith Harris 06/02/97 KVUE-TV

Mark Murray 06/02/97 KVUE-TV

Curtis Nichols 06/02/97 KVUE-TV

Brent Annear 06/02/97 KVUE-TV

Troy Kimmel 06/02/97 KTBC-TV

R. J. DeSilva 06/02/97 KXAN-TV

Dale Fuller 06/03/97 Cedar Park Emergency Management

Rodney Baden 06/13/97 McClennan County Emergency

Management

E-1

Appendix E

CDC Field Epidemiological-Health Report

Dr. Enrique Paz, Epidemic Intelligence Service

U.S. Centers for Disease Control and Prevention (CDC)

Dr. Shellie Kolavic, Epidemic Intelligence Service, CDC

and

Texas Department of Health (TDH)

David Zane, MS, Director, Injury Prevention and Control Program, TDH

Background

On May 28, 1997, the CDC received an invitation from the NWS to participate in a NOAA

Service Assessment Team in central Texas, where several tornadoes occurred on May 27. A

Texas-based CDC epidemiologist and a TDH injury epidemiologist joined a scientist from the

CDC to take part in the assessment. The CDC assessment was conducted from Friday, May 29,

to Monday, June 2, 1997.

Objective

The purpose of the CDC assessment was to describe mortality and morbidity related to the

tornadoes in the impacted population and to provide causal hypotheses for fatalities and injuries.

Assessment Methodology

Data-gathering techniques consisted of the following methods:

1. On-site observation

Time: 1-3 days

Resources: Automobiles, Texas Department of Transportation area maps

Indicators: Direct observations

Discussions with local leaders and health workers

Number of deaths

Number of injuries

E-2

2. Quick field surveys

Time: 2-3 days

Resources: National Weather Service

Texas Department of Health

Local hospitals

Newspapers and other media

Local law enforcement officials

Emergency Response Services

Williamson County Health Department

American Red Cross

Indicators: Rapid surveys, number of deaths, number of injuries

Results

Preliminary mortality and morbidity data was obtained from various sources, including the

American Red Cross, local hospitals, newspapers, television, and anecdotal information in the

field. This information was used to establish a base line for further surveillance and

epidemiological studies regarding the health impact and possible causal factors of tornado-related

deaths and injuries in the affected counties. Extensive damage, most notably in the Double Creek

Estates subdivision, hampered the assessment. No random sample or cluster sampling techniques

were conducted and no questionnaires were administered due to the gravity of the events, the

short period of our survey, and the lack of persons in the immediate area during our visits.

Fatalities and Injuries

A total of 30 deaths occurred: 27 (90 percent) in Jarrell, 1 in Pedernales Valley, 1 in Cedar Park,

and 1 in Austin. The 27 people died from multiple trauma; the other deaths were caused by heart

attack, suffocation, and drowning. The mean age of the decedents was 27 years and the range

was 5-69 years. The mean age for male decedents was 24 years; for females, the mean age was

33 years. These deaths directly affected 13 family units.

In Jarrell, 27 deaths occurred in the Double Creek Estates subdivision that had an estimated

population of 132. The mean age of the decedents was 26 years and the range was 5-51 years.

Of those who died, 14 were male (52 percent) and 13 were female (48 percent). Fourteen (52

percent) school-aged persons died. Ten families experienced two or more deaths; one family of

five perished. The deaths in Jarrell represented 7 percent of that town’s 1993 population.

In Cedar Park, one death (a male, age 69) occurred from cardiac arrest, probably due to storm-

related stress. In Austin, one person (a female, age 38) drowned in a storm-swollen creek. In

Pedernales Valley, one mobile home resident (a male, age 25) died. It is unclear where the

E-3

decedents were located during the storms. None of the decedents lived in residences with

basements, and most of the dead were recovered various distances from their homes. Several

Double Creek Estate residents survived in bathtubs, while others successfully sheltered in a

neighbor’s storm cellar. This was apparently the only underground shelter in this subdivision.

Local hospitals reported that a total of 33 persons (13 from Jarrell, 15 from Cedar Park and

5 from Pedernales Valley) were treated for tornado-related injuries. One of the 13 injured from

Jarrell died in the emergency room. Overall injuries, many multiple, included lacerations

(55 percent), contusions (46 percent), abrasions (30 percent), strain/sprain/muscle spasm

(18 percent), penetrating wound (0.03 percent), fracture (0.06 percent), and closed head injury

(0.03 percent). The median age of the injured persons was 38 years with the range 1-75 years.

Twenty-six injured persons were treated and released, one injured person died in the emergency

room (as noted above), and five persons were admitted. The median length of hospital stay was

11.5 days and the range was 1-31 days. Four persons were discharged home and one to an

inpatient rehabilitation facility.

E-4

Table 1. Mortality Data by County and Affected Area

County Town or Place Age Sex

Williamson Jarrell

51 F

46 M

46 F

45 F

41 M

40 F

40 M

40 F

38 F

36 F

35 F

21 M

17 F

17 M

16 M

15 M

14 F

14 M

13 F

11 M

10 F

Cedar Park

69 M

Travis Shoal Creek

Austin

38 F

Pedernales Valley

25 M

Sources of information: American Red Cross/CDC surveillance system and Texas Department of Health.

E-5

Table 2. Preliminary Descriptive Analysis

Williamson Co. Travis Co.

Total

Jarrell Cedar Austin Pedernales

Park Valley

Fatalities 30 27 (90%) 1 1 1

Males 16 14 (52%) 1 0 1

Females 14 13 (48%) 0 1 0

Mean age 27 26 69 38 25

Min age 5 5

Max age 69 51

Mean age males 24 21

Mean age females 33 32

Deaths under 14 14

18 years old

Affected families 13 10

Families with 9 9

more than 1 death

Conclusions

The majority of reported deaths occurred in Jarrell, in the Double Creek Estates subdivision. In

Cedar Park, there was one reported death, and 15 injured persons were treated at local hospitals.

In Austin, one drowning occurred. In the Pedernales Valley area, there was one reported death,

and 5 persons were treated at local hospitals.

Limitations of the Survey

This survey has some limitations because quantitative data were unobtainable and access was

limited by the devastating impact of the event. The data collection provided partial information

and may be limited by selection bias. A comprehensive survey that included a questionnaire was

not administered due to the short field visit.

E-6

Recommendations

Encourage the recovery efforts of the affected population.

Continue collecting data from the American Red Cross-CDC surveillance system, Texas

Department of Health, hospitals, and other data sources related to fatalities and injuries for the

final report.

Follow up with data sources regarding risk factors and causal factors of the mortality and

injuries due to the tornadoes in different areas, particularly in Jarrell.

Continue improving warning systems and develop future technology to prevent fatalities and

injuries.

Continue effort in prevention, disaster mitigation and education in local communities.

F-1

Appendix F

Fujita Tornado Intensity Scale

The Fujita Tornado Intensity Scale is a scale of wind damage intensity which wind speeds are

inferred from an analysis of wind damage.

Category Definition and Effect

F0 Gale tornado (40-72 mph): Light damage. Some damage to chimneys; break

branches off trees; push over shallow-rooted trees; damage sign boards.

F1 Moderate tornado (73-112 mph): Moderate damage. The lower limit is the

beginning of hurricane wind speed; peel surface off roofs; mobile homes pushed off

foundations or overturned; moving autos pushed off the roads.

F2 Significant tornado (113-157 mph): Considerable damage. Roofs torn off frame

houses; mobile homes demolished; boxcars pushed over; large trees snapped or

uprooted; light-object missiles generated.

F3 Severe tornado (158-206 mph): Severe damage. Roofs and some walls torn off

well-constructed houses; trains overturned; most trees in forest uprooted; heavy

cars lifted off ground and thrown.

F4 Devastating tornado (207-260 mph): Devastating damage. Well-constructed

houses leveled; structure with weak foundation blown off some distance; cars

thrown and large missiles generated.

F5 Incredible tornado (261-318 mph): Incredible damage. Strong frame houses lifted

off foundations and carried considerable distance to disintegrate; automobile-sized

missiles fly through the air in excess of 100 yards; trees debarked; steel-reinforced

structures badly damaged; incredible phenomena will occur.

S-1

Status of Actions

April 1998

Recommendation 1:

Personnel at NWSFO EWX should be trained in how to turn on the amateur radio equipment so

that they can at least hear reports, even if the network is not activated. NWSFO EWX should

also work with the amateur radio coordinator on backup procedures for contacting key people in

impending emergencies.

Status:

Personnel at NWSFO EWX were given instructions during the late summer of 1997 on how to

turn on and monitor amateur radio equipment in the event no amateur radio operator is on duty to

operate the equipment.

Recommendation 2:

The NOAA and the NWS should increase efforts in working with the DOD to ensure that DOD

commanders and personnel understand that properly maintained and operated DOD WSR-88Ds

are vital to the severe weather warning programs for both the bases and the civilian populations

under the radars’ umbrellas.

Status:

Since the early days of the deployment of the tri-agency WSR-88D network, there has been an

ongoing concern at the NEXRAD Program Management Committee of the radar maintenance

program capability. The GAO brought this factor to closer scrutiny in a 1995 report titled

"Weather Forecasting--Radar Availability Not Being Met" (AIMD-95-132). Following the

publication of this report, senior NWS and USAF senior management met to discuss these issues

and agreed that the USAF would cooperate in any way possible to support the NWS mission.

Lines of communication were established between DOD and NWS technical personnel for quick

problem resolution.

In recent months, DOD availability has matched the tri-agency standard of 96 percent, and there is

no systematic problem with the WSR-88D maintenance program.

Recommendation 3:

The NOAA/NWS should move rapidly to ensure that all Weather Forecast Offices (WFOs) in the

NWS are equipped with the best possible satellite analysis capabilities that are part of the

Advanced Weather Interactive Processing System (AWIPS). The NWS should also provide

proper training to ensure that satellite data are used in the warning and forecast program.

Status:

As of the end of March 1998, AWIPS is being used in operations at 20 WFOs and 8 River

Forecast Centers. The next group of AWIPS sites will be deployed this summer at 15 WFOs and

3 RFCs. Authorization for nationwide deployment of the remaining systems is pending from the

S-2

Secretary of Commerce. The offices that currently have AWIPS are accessing the high-resolution

data from geostationary satellites. With AWIPS, these satellite data can be integrated with radar

data, observational data, and numerical model data in a way never before available to operations.

The NWS will complete the initial phase of training on using the new GOES satellite by June

1998 via the following methods: (1) Provision of an in-depth, 2-week course on mesoscale

satellite meteorology at COMET; (2) Completion of a series of three computer-based learning

modules; (3) Establishment and maintenance of a Satellite Meteorology Web page.

Recommendation 4:

The SPC should make every effort to coordinate with any NWS office that has warning

responsibility before a watch is issued that covers a part of the CWA of that NWS office.

Status:

The SPC concurs that it did not coordinate with the Austin-San Antonio NWSFO prior to issuing

the Tornado Watch in advance of the Jarrell tornado.

The SPC is working with the Office of Meteorology, NWS Headquarters, to instigate the

teleconferencing call lists needed to ensure such oversights cannot happen in the future. The

NWS plans to have these lists in place nationwide by June 1, 1998.

Recommendation 5:

The OSF should evaluate the Level-2 Archive data from the EWX radar and further investigate

these anomalies. The NEXRAD Program should continue its plans to implement the National

Severe Storm Laboratory’s experimental mesocyclone algorithm because of its higher skill and

informative outputs, including probabilistic guidance values.

Status:

On June 18, 1997, Don Burgess, Chief, Operations Branch, Operational Support Facility,

delivered a report to the Service Assessment Team. Conclusions from that report were included

in the Event Summary.

The NEXRAD Technical Advisory Committee is charged with evaluating experimental algorithms

to review proposed scientific changes and additions. The tri-agency System Recommendation and

Evaluation Committee approves and prioritizes proposed changes to the WSR-88D baseline.

Recommendation 6:

A thorough scientific study of these events should be undertaken by researchers and academicians

with the results made available to forecasters of the NWS. Any study should include the U.S.

Weather Research Program (USWRP), the NWS Collaborative Science, Technology, and Applied

Research Program (CSTAR) and NOAA/NWS Cooperative Institutes.

Status:

S-3

Scientific research on the Jarrell, Texas, tornado is being undertaken by the university community,

NOAA Cooperative Institutes, and NWS personnel, often in close collaboration. During the

CSTAR Workshop held October 7-10, 1997, a special session on the Jarrell tornado was

presented to illustrate science and service issues that must be addressed in order for forecasts and

warnings of extreme weather events to be improved in the future. Presentations were given by

researchers from the NWS and NOAA Cooperative Institutes. The workshop was attended by more

than 60 representatives from the universities, the NWS, other NOAA research laboratories, and the

USWRP.

The full workshop report can be found on the Internet at:

<http://www.nws.noaa.gov/om/cstar97.htm>.

Recommendation 7:

Use and interpretation of the WSR-88D data, especially in convective situations, should be left to

meteorologists trained in both mesoscale meteorology and the interpretation of WSR-88D data.

In addition, NWS WFOs should be appropriately staffed with a sufficient number of such

meteorologists to meet these requirements. Further, the NWS needs to ensure that any future

training programs being developed to replace the OSF WSR-88D course should be of equal or

better effectiveness than was that course. Meteorologists must have the ability to quickly

interpret reflectivity and velocity signatures within the context of the overall mesoscale weather

pattern, and relate the radar information to the other analyses and observing tools such as a

detailed mesoscale surface analysis combined with satellite, surface observations, profiler and

radar Velocity Azimuth Display (VAD) vertical wind profiles, etc. Accordingly, the NWS should

ensure that the expertise, materials, and other resources, including field staff time, are made

available to meet such a training requirement.

Status:

The NWS Southern Region policy regarding the operation of the WSR-88D and the interpretation

of radar data was restated to all NWSFOs in a memo dated July 9, 1997.

The Operations Training Branch of the OSF is developing a set of on-site training modules to

replace the in-residence WSR-88D course. The Distance Learning (DL) Operations Course

consists of printed materials, CD-ROMs, Web-based materials and Teletraining. The same final

exam will be used in the new DL version to ensure the course is equally effective. All Southern

Region meteorologist interns are currently in training using the modules which have been

released.

Recommendation 8:

Management at NWSFO EWX should establish and enforce clear lines of authority and

responsibility to be followed during severe weather operations. Further, emphasis should be given

to keeping all staff current regarding such operational procedures.

Protection of life and property is the fundamental mission of the NWS. NWS offices need to

develop a strategy for a coherent product suite encompassing warnings, statements and short-

term forecasts throughout severe weather episodes. They need to work with customers to apprise

them of priority which will be placed upon products during such events. All NWS WFOs should

S-4

have a standing policy that routine products have a much lower priority during periods of ongoing

severe weather. They should ensure that all staffing resources be brought into a cohesive severe

weather mind set and that the focus of the office is always on the fundamental mission.

Status:

Management at NWSFO EWX has instructed the staff regarding lines of authority, i.e, that the

Forecaster in Charge on shift has the responsibility to assign tasks during significant weather

events. Also, existing policy regarding priority of duties has been reemphasized to the staff and

such instructions are included in the Station Duty Manual.

Recommendation 9:

The staff at NWSFO EWX is commended for their dedication and extra effort to seek

confirmation that warning information was being received by officials who had a critical need for

it. Nevertheless, it is recommended that NWSFO EWX work with emergency management

officials to seek alternative and more efficient methods to confirm receipt of warnings.

Status:

NWSFO EWX efforts are ongoing to promote the Emergency Managers Weather Information

Network (EMWIN) technology for the dissemination of NWS warnings as well as the NOAA

Weather Radio SAME (Specific Area Message Encoding) technology. EMWIN is a personal

computer-based system using inexpensive, user-friendly software that provides warning and

forecast products, graphics, and limited satellite imagery in near real-time. A SAME-capable

receiver allows the user to program the weather radio to receive warnings for the areas of the

user’s choosing.

Recommendation 10:

The NWS should reach consensus with the Federal Emergency Management Agency (FEMA) and

the state emergency management agencies on a single communication/coordination system that is

consistent nationally and meets the internal and external communication/coordination

requirements of both the NWS and the emergency management community. All NWS offices

should be provided with needed technology to facilitate efficient internal and external

coordination.

Status:

The Office of Meteorology and FEMA’s Information Technologies Directorate have examined a

number of communication technologies over the past year. Currently, neither FEMA nor the

NWS have the funding necessary to complete the NAWAS Upgrade started in 1996.

In response to the continuing requirement for a National coordination/communication system, the

NWS Office of Meteorology (OM) and the Office of Systems Operations (OSO) have identified

FTS2000 as an interim system that can meet the basic internal and external coordination

requirements of NWS field offices and National Centers. The goal is to have the interim FTS2000

system operational by the end of June 1998.

Recommendation 11:

S-5

The NWS has sent to the Federal Communications Commission (FCC) proposed revisions to the

EAS rules. The NWS should continue to work with the FCC to clarify situations that need

automated activation. The NWS and the FCC, in cooperation with the National Association of

Broadcasters, should agree on a limited set of critical weather warning products which would be

programmed for automatic activation of the EAS.

Status:

The NWS works very closely with the FCC on all matters relating to the EAS. The FCC is

expected to adopt NWS-generated proposals in the fall of 1998 that include the optional use of

many more non-weather related and a few more weather-related EAS event codes.

The EAS media facilities, however, are not required by Federal law (FCC Rules, Part 11) to

disseminate any messages other than Presidentially-declared disaster information. If EAS facilities

choose to broadcast local emergency information, whether from the NWS or other authorities,

they have the option to either automatically activate or delay for later manual activation.

Furthermore, many EAS facilities want to use their own broadcasters rather than use NWR

broadcasts directly. These actions are decided and documented in local EAS Plans across the

country on a case-by-case basis. The NWS participates in these local plans via the efforts of the

WCM and designated staff.

Recommendation 12:

The NWS should continue to work with the state of Texas Division of Emergency Management

to seek ways to improve the timely transmission of weather warnings via TLETS and should

intensify efforts on both a local and state level to explore alternative methods of communicating

critical weather products to emergency management officials.

Status:

Working relationships with the Texas Division of Emergency Management have been greatly

expanded during the past 10 months. Changing the TLETS to a fully automated system is a state

funding issued beyond the control of the NWS. The NWS continues to promote the Emergency

Managers Weather Information Network and NOAA Weather Radio with SAME capabilities.

Recommendation 13:

The NWS should develop additional partnership initiatives with the public and private sector to

encourage the use of NWR for the receipt of critical weather information.

Status:

With the FCC’s adoption of NWR’s digital protocols for the new EAS, the NWR now becomes

the NWS’s primary input into this nationwide warning delivery service. Much interest and

visibility has already been generated by NWR input to the 14,000 some participating radio and TV

stations across the country. Starting in 1999, cable TV will join the EAS and can optionally

broadcast, for the first time, NWS warnings via NWR.

S-6

Recommendation 14:

All the NWR transmitters should be equipped with backup power.

Status:

A large minority of NWR transmitters do have backup power. Many of these typically are

systems co-located or associated with commercial broadcast facilities. Others with backup power

are found within large government or commercial buildings.

There is no backup power for the majority of transmitter sites because it is cost-prohibitive to

install generators at more than 400 locations, provide personnel with special knowledge on

generator maintenance and logistical support, and comply with EPA regulations on fuel storage

and containment.

Recommendation 15:

The NWS, in concert with FEMA and the ARC, should continue to emphasize the importance of

not attempting to flee an approaching tornado and reemphasize the importance of adhering to

long established tornado safety rules. (In the vast majority of situations, people who take shelter

in sturdy, well-built structures will survive most tornadoes whereas people who try to flee in

automobiles will place themselves at much greater risk.) The Service Assessment Team also

recommends that there be significant public and industry education on the meaning of well-built

tornado shelters.

Status:

The NWS remains a strong player in the National Disaster Education Coalition and will continue

to work within the coalition to effect this recommendation. The coalition consists of

representatives from FEMA, the U.S. Geological Survey, NWS, and The American Red Cross.

In meetings after the Jarrell tornado, the coalition agreed that tornado safety measures should not

be modified. It agreed to continue to emphasize the standard tornado safety measures well before

the recommendations from Jarrell were published realizing those safety measures still applied.

The current 12-page, tri-agency brochure “Tornadoes...Nature’s Most Violent Storms” explicitly

explains the long established tornado safety rules. Other Red Cross pamphlets also fully detail

these same consistent safety messages. The NWS will ensure that any future tornado brochures

will strengthen this message. The coalition will continue to support this recommendation, not

only in printed material but in all other venues.

The NWS has worked with local media in the central Texas area (primarily through efforts of the

EWX WCM) to reemphasize the importance of existing tornado safety rules. We have also

communicated the importance of reemphasizing this message to all our WCMs in the Southern

Region through routine coordination calls.

S-7

Recommendation 16:

The NWS should partner with other public agencies and the private sector to develop plans for

protection of employees and the public located in large buildings during severe weather threats.

Status:

The process for developing a network of public shelters has been in place for many years. As part

of the emergency management’s all hazard response, each state is required to develop a shelter

network to meet the evacuation and sheltering needs of its citizens. Many states have legislated

that counties and cities identify shelter space within public meeting places, such as court houses

and shopping malls. In addition, local Warning Coordination Meteorologists are constantly

working with emergency managers to educate the public on how to protect themselves. For the

past 5 years, the NWS has partnered with FEMA and the American Red Cross to produce and

distribute natural hazard awareness materials necessary to heighten the public’s awareness.

Recommendation 17:

First, the NWS should ensure that all field offices develop strategies that make full use of all office

personnel qualified to assist in warning and forecast operations during critical weather situations.

Second, the NWS should evaluate its plans regarding future staffing and service changes to ensure

that workload constraints do not compromise the agency’s safety of life and property mission.

Specifically, transitional staffing which has been placed at selected offices should be retained until

such time as new technology can demonstrate it is capable of mitigating the workload associated

with warning and forecast operations to an extent that a reduction in staff is feasible.

Status:

See Recommendation 8.

On February 26, 1998, NWS senior managers met to discuss AWIPS capabilities and field office

staffing requirements. They agreed to maintain transition staffing at NWS offices with multiple

associated WSR-88Ds and at offices with heavy NWR programming loads until the software

builds of AWIPS provide the necessary functionality to assist effectively and efficiently in the

delivery of these services. The Southern Region has ensured that staffing levels at offices meet

these guidelines.

Recommendation 18:

Studies should be undertaken by the research and academic community to help identify these

types of patterns which may aid in the development of these rare but catastrophic tornadoes.

Such studies should have the participation of the USWRP, CSTAR, and NOAA/NWS

Cooperative Institutes. The results of these studies should be made available to NWS forecasters

to aid in the warning and forecast process.

Status:

See Recommendation 6.

Recommendation 19:

S-8

NOAA/NWS should compile a list of potential, quick-response team members, drawn from

appropriate expertise in the NOAA line and staff offices, other Department of Commerce Bureaus

such as NIST, and from the universities and private sector. The team should be interdisciplinary

in nature. Following a significant tornado outbreak, a team of three to five scientists should be

assembled and dispatched to the centroid of the outbreak within 24 hours. Teams should include

meteorologists, a wind engineer, and a social scientist, an epidemiologist and a public affairs

specialist. Funding to carry out the travel costs of the team and an aerial damage-mapping survey

should be set aside yearly in a revolving budget.

Status:

NOAA line offices led by the NWS have partnered with FEMA, the U.S. Geological Survey, and

the Department of Agriculture in the development of a Post-Storm Data Acquisition (PSDA)

capability. Funding for the PSDA is shared by the participating agency partners. The NWS has

allocated an annual funding of $300K to meet post storm event assessment needs. The oversight

responsibility and activation of the PSDA resides with the Office of the Federal Coordinator for

Meteorological Services (OFCM). The OFCM, with assistance of the participating agencies has

drafted an plan for timely and coordinated post storm data collection.

To meet the aerial support requirement for damage-mapping, the OFCM, with the assistance of

the Office of Meteorology, has secured a Memorandum of Understanding between the OFCM and

the Civil Air Patrol. The data collected and the post storm reports will be made available to

agencies who are represented at the OFCM and to the remainder of the research and operations

communities through the OFCM Web page.

Currently, the NWS and OFCM attempt to field and deploy rapid service assessment teams within

24 hours of the event. The greatest limitation to date has been the small pool of human resources

available for these teams. The Office of Meteorology has proposed that the NWS and OFCM

support the development of a quick response training program to meet the demands being placed

on this small cadre of individuals. The goal is to identify, train and support individuals in the

deployment of quick response teams on a regional basis.

The NWS already have several names from volunteers and from people suggested by other

elements of the agency. OM will continue to add to it as the opportunity arises.