xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200337Ia 4500
controlfield tag 001 001971515
006 m d s
007 cr bn|||||||||
008 081205s1992 cou b s000 0 eng d
datafield ind1 8 ind2 024
subfield code a F57-00004
Social response to the first "A" alert of the Parkfield earthquake prediction experiment
h [electronic resource] /
by Colleen Fitzpatrick and Paul W. O'Brien.
Boulder, Colo. :
b Natural Hazards Research and Applications Information Center, University of Colorado,
22 p. ;
Quick response research report ;
Includes bibliographical references (p. 22).
[Tampa, Fla. :
University of South Florida Libraries,
n Digitized from copy owned by Natural Hazards Center, University of Colorado at Boulder, in a joint project with the Louis de la Parte Florida Mental Health Institute (FMHI) Research Library's disaster mental health initiative.
x Social aspects
O'Brien, Paul W.
University of Colorado, Boulder.
Natural Hazards Research and Applications Information Center.
t Natural Hazards Center Collection
::sec IA L TO TH'E.fiRST "A" AlE.'RT OF THE FA R,l,t=H=LbARTH-G(UA'K E PREblCflONNENT Col \ee..nrltt.pOl:l:.V"'ic'k..-Pa-Lll\JJ I 0)Byo'le.-n o(Colo f'o.dO)IVl"it.ittAte.otEehC1-\JlovQ..l Sc,'e.Y'\ceNo.;tL.lYCt \ tt<.o-v-dsk es="" ea..rchd.ap-pii61="">V\Le.Y\ter/9c;Z
SOCIALRESPONSETOTHEFIRST "A"ALERTOFTHEPARKFIELDEARTHQUAKEPREDICTIONEXPERIMENTByColleenFitzpatrick,Ph.D.andPaulW.O'Brien,Ph.D.HazardsAssessmentLaboratoryandDepartmentofSociologyColoradoStateUniversityFortCollins,Colorado80523QUICKRESPONSERESEARCHREPORT #54 1992Natural Hazards This publication is Center's ongolOg Research& Report Senes. Quick Respons d edu/haz ards http://WWW.colora o.TheviewsexpressedinthisreportarethoseoftheauthorsandnotnecessarilythoseoftheNaturalHazardsCenterortheUniversityofColorado.
Social Response to the First "A" Alertofthe Parkfield Earthquake Prediction ExperimentColleen Fit'ljJatrick, Ph.D.andPaulW.O'Brien,Ph.D.Hazards Assessment LaboratoryandDepartment of Sociology Colorado State UniversityFortCollins, Colorado80523 Submitted to: The NaturalHazards ResearchandApplications Information Center UniversityofColorado Boulder, Colorado 80309 December 1992
Social Response to the First "A" Alertofthe Parkfield Earthquake Prediction ExperimentUntil just a few years ago Parkfield was but a sleepy, obscure little town tucked away in the Diablo Rangeofthe Sierra Madre mountainsofsouthern Monterey Countyincentral California. But in 1985 theresidentsofParkfield were awakened from a stateofslumber when the U.S. Geological Survey (USGS) issued an official earthquake pre diction, with time, place and magnitudeofoccurrence specifically indicated. This peaceful ranching town,ofabout 34 residents with slightly over 100 in its surrounding environs, has been stirred from a senseofsolitude and quiescent isolation many timesinthe past several years. Parkfield is now the siteofthe world's most premier earthquake prediction experiment. The 20 mile long Parkfield segmentofthe San Andreas fault is home to more earthquake prediction instruments (devices designed to detect anomalies which might provide indicators to effectively predict the occurrenceoflongand short-term earthquake events) than anywhere on earth. The long-term prediction, for a quakeofRichter magnitude 5.5 to 6.0 to occur near Parkfield by 1993, is the first earthquake prediction to be endorsed by both the National (NEPEC) and California (CEPEC) Earthquake Prediction Evaluation Councils. Includedinthe prediction is the possibility that the next Parkfield quake could be as strong as 7.0. Parkfield has a long historyofregularly occurring quake activity. Earthquakesinthe6.0orgreater range are known to have happened therein1857, 1881,1901,1922, 1934 and 1966. The "characteristic" 6.0 temblor could cause damage to windows, chimneys, walls and glasswareinParkfield, and may have slightly similar impacts in the surrounding communitiesofPaso Robles, Coalinga, San Miguel, and Avenal.Ifthe next Parkfield quake is in the 7.0 range, it could affect a geographic circumference comprisedofallorpartsofseven counties. The larger quake could have some damaging impacts and at least be feltinthe countiesofnorthern Santa Barbara, San Luis Obispo, San Benito, southern Monterey, Kings, western Kern, and Fresno. Announcementofthe Parkfield prediction drew the attentionofthe media from across the nation. Reporters flocked to Parkfield to interview its few residents, and scientists have been frequenting the area since. Initial public reaction to the prediction was formally studied shortly after it was publicly announced (Mileti and Hutton 1987). In 1988, the Southern California Earthquake Preparedness Projectofthe Governor's OfficeofEmergency Services (OES) distributed a brochure detailing the prediction. Mailed to some 122,000 householdsinthe seven-county region, the brochure included a descriptionofthe prediction, adviseonhow to get ready for earthquakes, what to do during and immediately afteranearthquake, places to call for more information, and a depictionofthe alert levels designed to portray the predicted quake's likelihoodsofoccurrence.Inaddition, over the years the public has been exposed to this unique prediction phenomenon andtoadviseonhow to become earthquake ready by a multitude1
ofsources and through a wide varietyofcommunication channels. The risk communi cation process involvedinthis prediction experiment has also been formally investigated, and analyzedina numberofways (Mileti, Fitzpatrick and Farhar 1990, Fitzpatrick 1992, Mileti, Fitzpatrick and Farhar 1992, Mileti and Fitzpatrick 1992). Since announcementofthe prediction, Parkfield has had many days in the spotlightofpublic attention. Its most recent focused attention began lateinthe eveningofOctober 19, 1992 when the USGS declared an earthquake alert for the Parkfield segmentofthe San Andreas fault.Ineffect, the alert was a short-term earthquake prediction, the first such NEPEC/CEPEC approved prediction to be issued in United States history. In response, the State OES declared an "A-level" alert, its highest typeofpublic earthquake advisory. This alert amounted to a short-term earthquake warning for residentsofthe seven-county region surrounding Parkfield. It also represents yet another chapter in the continuing sagaofthe Parkfield Earthquake Prediction Experiment. The following depicts what alerts are in the contextofthe Parkfield Earthquake Prediction Experiment and the unfoldingofevents surrounding the first Parkfield A-level alert. The purposeofresearching the alert and the methods used to investigate this phenomenon are then presented. Results from interviews and observations acquired from social research on the alert and the derived findings complete this report.The AlertAlerts,inthe Parkfield Experiment, represent various levelsoflikelihood that the Parkfield earthquake may occur. Six levelsofalert have been established for Parkfield with levels C, B, and A representing the three highest levels. The highest levelofalert is "A".At the A alert stage, the USGS predicts thereisat least a 37 percent ofthe 'characteristic' Parkfield earthquake occurring within 72 hours. The greatest likelihoodisimmediately after announcementofthe alert, the probabilities diminish during the following 72 hours.Analert stage lasts 72 hours unless a longer periodiswarranted by subsequent instrument activity at Parkfield.AnA level alert means that the highest estimatedlikelihood-37%+-hasbeen reached, not that the earth quakeiscertain to occur (Governor's OfficeofEmergency Services 1988, p. 6).A level "B" alert means that the USGS believes there to bean11% to 37% increased likelihoodofthe Parkfield earthquakeoccurring within a 72 hour period, and a level"c"represents a 2.8%to11% increased chanceofoccurrence and that instruments need to be very closely monitored for indicators which might lead to higher level alerts. Threshold criteria forissuing a short-term prediction, or A-level alert, had been established by CEPEC when theParkfield Earthquake Prediction Response Plan(Governor's OfficeofEmergency Services 1988) was developed during the initial years2
ofthe prediction project. Threshold criteria allow for a "real time review" to be foregone when indicators suggest that the "characteristic" Parkfield quake is imminent. "To avoid the repetitionof'false alarms' and because the USGS considers only the A level alert to be a short-term prediction", a level A alert is the only one which necessitates a public warning (Governor's OfficeofEmergency Services 1988, p. 6). Until this event, level A alert had never been issued. The first Parkfield "A" alert was triggered when the USGS picked up seismic readingsofa4.7Richter magnitude earthquake under Middle Mountain just northofParkfield at about 10:28 p.m. on Monday, October 19, 1992.Anearthquakeofthis approximate size had been predetermined to be a prime signal that the "characteristic" quake was imminent and that a level A alert should be issued. The alert notice also indicated thatthe"'characteristic' Parkfield earthquake (the Magnitude6)could trigger, or grow into, a larger earthquake, as large as Magnitude 7. This could result in severe shaking intensities ... and could result in casualties and damage [but that] CEPEC considers this larger event much less likely than the 'characteristic' earthquake" (California OfficeofEmergency Services 1992, p. 3).By10:52p.m.,on this same night, the USGS notified the State OES that "'there is a significant likelihood thatanearthquakeofMagnitude 6 will occur on the San Andreas Fault near Parkfield in the next 72 hours'" (California OfficeofEmergency Services 1992, p. 3). This notification complied with the pre-established procedureofcommunication between the USGS and the OES set forth in theParkfield Eanhquake Prediction Response Plan.As soon as the State OES received the alert from the USGS, it began notifying officesofemergency services in all sevenofthe potentially impacted counties. Simultaneously, State Operations Centers in Sacramento and Regional Operations CentersinRegions I (Los Alamitos), II (Pleasant Hill), and V (Fresno) were activated. Key state agencies with response roles (i.e., California Highway Patrol), organizations providing lifeline services (i.e., Diablo Canyon Nuclear Power Plant, Pacific Gas & Electric), and facilities which could threaten the public safety (e.g., dams, pipelines, toxic waste sites) were also immediately notifiedofthe alert. Then at 1:00 a.m. on October 20, 1992, about two hours after receiving the alert from the USGS, the OES issued a News Release which restated the advisory that had previously been sent to the county OESs. The State also dispatched staff to Cholame, the pre-designated site for setting up the State's mobile command post, and to USGS officesinMenlo Park, California. The principal roleofthe State OES during such an alert period is to assist local jurisdictions and to transmit any updates on the earthquake hazard to the more localized emergency response agencies, and to the press. In addition, during the warning period as well as during an eventual emergency, the State OES is prepared to provide assistancetolocal government jurisdictions by providing personnel and equipment upon requests made through mutual aid procedures.3
The A alert indicated that the Parkfield earthquake had a significantly increased likelihoodofoccurring within a72hour (three day) periodofreleaseofthe notice. However, subsequent update notices indicated that as time passed, the likelihoodofoccurrence diminished.ByThursday morning, newspaper articles and update notices indicated that the likelihoodofoccurrencehaddecreasedtoabout five percent. The A alert expired at about 10:30 p.m. Thursday, October 22, 1992, and its likelihoodofoccurrence revertedtoless than 2.8%. However, a seriesofsmall quakes were detected the following Sunday evening. This led the USGStoissue a"B"level alert which was announced to broadcast media and to the StateOESalmost simultaneously. Althoughitrequirednopublic warning, it did indicate that the USGS considered the likelihoodofoccurrence for the Parkfield quake had increased to about11%to37%during a 72-hour period. This alert expired without being upgradedtoa level A alert.ResearchPurposeThis A-level alert presented a unique opportunitytoinvestigate public and organizational responsetothe first NEPEC/CEPEC short-term earthquake prediction issued in the United States. To best take advantageofsuchanopportunity it was important that the alert situation be investigatedinreal-time. This meant that data neededtobe collected before it decayed and, therefore, during the ongoing warning process. We thus requested the Natural Hazards Research and Application Information Center at the UniversityofColoradoinBoulder to activate a pre-approved quick response granttoinvestigate the alert. The primary purposeofthis research was to examine social response to the alert. Social response was conceptualizedasinvolving two basic unitsofanalysis-localorganizations responsible for public safety and/or the provisionoflifeline services, and the general public. We soughttoanswer the following setofbroad, unoperationalized, research questions:(1)How did the public respondtothe 72-hour warning? Was it some thing that people believed and took seriously enough to engage in protective actions?Ifso, what did people do differentlyinlieuofthe warning? What kindofactions were taken?(2)How did local emergency response organizations handle the warning? What was the perceived usefulnessofthe alert fromanorganizational point-of-view? What did organizations do differently from routine operating procedures?4
(3) To what degree were channelsofcommunication effective in termsofgetting the warning out to the public? To what degree were channelsofcommunication effectiveintermsofgetting the warning disseminated to local emergency response agencies? These questions were constructedtoprovide general guidelines and parametersoffocus for this case study on perception and responsetothe alert in a real life context.MethodsThe methods used for conducting this qualitative field research were straightforward. Unstructured interviews and directed conversations were the principle toolsofdata collection. We also engagedinquasi-participant observation. This could not be considered a true participant observation study since we wereinthe area for only seven days, did not takeupresidence and it was probably obvioustolocal citizens that we were outsiders. We also closely monitored news programs and collected newspaper articlesasanother meansofacquiring as much information as possible during our short stay in the area. The alert was issuedtothe residentsofseven counties in central California. Therefore, representativesoforganizations vested with responsibility for providing emergency and lifeline support, and residents in the at-risk counties were targeted for interview. Citiesofcounty seatsofgovernment were selected as principal locationstocollect data.Itwas also important that smaller towns located throughout the region be investigated. In addition, threeofthe larger citiesintheregion-PasoRobles, Coalinga andTaft-hadalready been studiedinsocial researchonthe Parkfield Earthquake Prediction Experiment. These citieshadbeen selected for study based on the cliteria ofdistance from the predicted epicenter and experience with a damaging earthquake. Paso Robles represented a city that was relatively closetoParkfield but lacked recent experience with damaging earthquakes. Coalinga,onthe other hand, represented a city that was equally closetoParkfield buthadexperienced a damaging earthquakein1983. Taft represented a city within the areaofrisk yet considerably distant from the epicenter and without recent damaging earthquake experience.Itseemed prudent that, at a minimum, this quick research should investigate the alertinthese same cities. To do so would enableusto make a more meaningful contributiontothe growing bodyofsocial scientific knowledge on this unique and ongoing earthquake prediction experiment. We arrivedinthe area mid-dayofthe second full dayofthe three day alert period.Weleft the area seven days later or five days after the alert expired. During this timewedocumented more than27interviews and engaged in directed conversations with numerous citizens. We collected datain13towns locatedinsixofthe seven targeted5
counties (San Benito County was not visited duetodistance and lackoftime). We also visited the county seatsofFresno, Kings and San Luis Obispo. The other county seats were located considerably outsideofthe alert region and we chose to expend our resources only within the region depicted as at-risk. We also collected more than 25 articles specifically about the alert and about general earthquake readiness circulating in newspapers serving the region. Table I depicts the counties and towns wherein interviews and observations were conducted. Data were collectedbyconducting unstructured interviews with organizational spokespersons and operations personnel, with ordinary citizens, and with owners or managersofretail stores where earthquake readiness supplies were likely to be sold. Organizational interviews provided dataonboth public and private agencies that play vital rolesincommunities during disaster situations. Citizens were interviewed, inanunstructured and inobtrusive manner,toacquire data which would contributetogiving a senseofthe public's perceptionsofand response to the alert.Table1.Counties and Towns Wherein Interviews and Observations were ConductedCountyMonterey San Luis Obispo Santa Barbara Kern Fresno KingsTownMercalliRiskParkfield VII VIII San Luis Obispo VVI Paso Robles VII VIII San Miguel VII VIII Templeton VII VIII Buellton VVI Solvang-Santa Ynez VVI Taft VI VII Old River II IV Coalinga VII VIII Fresno II IV Hanford II IV Avenal VII VIII6
Questions askedoforganizations and citizens were not standardized and were used only as a meanstogain entretointerviews,tofacilitate conversations and to keep discussions focused on our research interests. The lengthofinterviews ranged from as short as a few minutestomore thananhour depending on the person being interviewed and the contextofthe interview situation. Citizen participation was elicited by engagingincasual conversationsina varietyofsituations, forexample,inrestaurants, stores, and other such public places. Organizations Organizations specifically targeted for interview were county officesofemergency services, city and county fire departments, and state highway patrol, county sheriff, and city police departments. Public utility, American Red Cross, public school district, and hospital organizations were also targeted for interview. Interviewing spokespersons and other personnel from such organizations would likely provideuswith ample information necessary to depicting organizational responsetothe alert. Questions we soughttohave answeredbyorganizations were how and when they heard about the alert,ifthey believed what the alert depicted, what they did in response,towhat extent it altered routine activities, the amount and qualityofpublic inquiries about the alert,ifresponsetofuture alerts would be alteredinany wayifthe earthquake didn't happen, the usefulness (or lack thereof)ofthe alert, and what would they have done differentlyifgiven the opportunity. Spokespersons for organizations were contacted by walking into offices, introducing our purpose, and askingtotalk about the alert. Other organization personnel were interviewedinessentially the same manner but insteadofgoingtoanoffice, we simply made contact with themintheir working environment, for example, police officers on the street, firefighters workingontheir equipment, and so forth.CitizensResidents were generally queriedonwhether they believed the predicted quake was imminent, what they were doingtoget ready, and about their general impressions concerning government issuing earthquake alerts. People who workinretail stores occupy a key positioninsociety which involves daily contact with the public. Since such people are well awareofa public's purchasing behavior, the assumption was thatifstores whichcarrypreparedness items such as bottled water and batteries experienced a marked increaseinthe purchaseofsuch items andhadtoorder additional loadstokeep the store adequately stocked, thenbyinference, at least some membersofthe public engagedina minimal degreeofreadiness response as a resultofthe alert. Therefore,7
owners and managersofhardware and grocery stores were interviewed as public in formants. Store managers/owners were interviewed by going into a store and asking to speak with whoever was in charge.Limitationsofthe ResearchGiven the qualitative natureofthis research and the limited numbers and kindsofinterviews conducted, the data collected in this effort cannot be considered to be representativeofeither organizationsorcitizensinthe region. Furthermore, findings cannot be generalized beyond the limited geographic scopeofthe alert area. Findings, therefore, can only be interpreted as meaningful to the extent that what people related tousis real and true for them as individuals or as spokespersons for organizations. The Parkfield situation is veryunique and likely very different from other prediction efforts in California. The Parkfield prediction experiment has been ongoing for more than seven years, and a great dealofinformation on the prediction itself and on earthquake readinessingeneral has been circulated throughout the regionina multitudeofways during this time. Becauseofthis, the Parkfield region must be viewed as different from any other areainCalifornia.Inaddition, this regionofintense earthquake study is sparsely populated, relativetoother regionsofCalifornia, and population density differences severely limit any reasonable comparisons. These caveats are meant to convey the generalizability and inference making limitationsofthis field research on social responsetothe first official short-term earthquake prediction to be issued in the United States. This is nottosay, however, that some findings and lessons learned as a resultofthe alert cannot be usefully appliedtoother situations.Results From Interviews and ObservationsOrganizations respondedtothe alert for different reasons than did the public. This would be expected since organizations have responsibility for public welfare that is distinct from the public itself. Differences would also be expected since organizations possess emergency plans which require a Parkfield alert response. The public does not possess Parkfield plans, per se, and havenopre-scripted scenarios to direct behavior. The following subsections on organizations and the public depict the data derived from interviews and observations on these two distinct unitsofanalysis.8
Organizations Organizationsthroughout the region heard about the alert from official organi zational sources. The Parkfield Plan worked very wellintermsofits predesignated modeofcommunication during the A alert period. The Parkfield Plan calls for theUSGSto communicate the alert to the State OES who then notifies county officesofemergency services whointurn notifies local agencies important to a potential emergency response situation. The State OES, after about a 30 minute wait, then provides a news releasetomedia agencies throughout the state. This pre-determined systemofcommunication was implemented and, for the most part, operated according to plan. Virtually all county officesofemergency services heard about the alert within a matterofa coupleofhours after the USGS issued it.Localagencies responsible for emergency response and the provisionoflife support services (i.e., county fire and sheriff departments, city fire and police departments, American Red Cross chapters, county departmentsofeducation and utility companies) also became awareofthe alert within a few short hoursofits issuance. All countiesofemergency services and city agencies with responsibilities during emergencies had response planstoserve as guidelines for what to doinvarious emergency situations. Some countieshadsections in their plans which specifically addressed the Parkfield situation, including alerts, while others had only general earthquake response sections. All organizations recognized that they were responsible, at some level, for respondingtothe alert. Without exception, organizationsresponsible for emergency services did somethinginresponse to the alert. At a minimum, police and fire departments, officesofemergency services, city mayor's/manager's offices, Red Cross chapters and utility com panies in the entire region, briefed personnel about the alert and placed them on stand-by status. The extensivenessofresponse varied considerably across organizations dependingonthe typeoforganization and its geographical location and experience with damaging earthquakes.Within organizations, perceptions about the seriousnessofthe alert and the potential for the earthquake to occur varied according to the positionofpersonnel in the hierarchyofauthority. Media responsetothe alert, while driven by factors other than those which induce responsebygovernment or private service organizations, was very active and could not be neglected.County officesofemergency services.All county officesofemergency services (DES) recognized the alertasserious enoughtoimplement appropriate response plansandtoengageinactivities outsideoftheir normal modeofoperation. County OESs werethemost activeofall organizationsinthe region. This was largely due to these agencies being the primary organizations for settingupemergency operations centers and coordinating responsetoemergency situations foranentire county. The responsibility of county OESstonotify other public service oriented agencies, such as city police and fire departments, hospitals and schools, also necessitated a great dealofactivity. In9
addition, it was the county OESstowhich the public turned for more information about the alert. During the first few hoursofthe alert, all county OESs opened up emergency operations centers (EOCs). This activity primarily involved on-call personnel arriving at predesignated locations where they began to enact a call-down procedure to notify local response jurisdictions and local mediaofthe alert. Notificationofthe alert also included reminderstoreview emergency plans andtoengage in appropriate protective actions for personnel and equipment. More than anything else, activating the EOCs for this alert simply meant waitingbyafaxmachine and other communication devices for updates from the State OES.Italso meant briefing local mediaofthe situation and making decisions on actions to be takenbyfire departments under county responsibility. After a few hoursofsuch activity, county EOCs, throughout the region, staffed-down to minimal operations. This meant that personnel went on rotating shiftssothat at least one person was in the EOC atalltimes. Within the first 24 to 36 hoursofthis 72-hour alert, most county OESs actually closed down their EOCs. The exceptiontothis wasSanLuis Obispo County whichhasa single dedicated EOC facility built with monies providedbythe nearby Diablo Canyon nuclear energy plant.Inaddition, San Luis Obispo County receives special financial support for engaging in drills and alerts becauseofthe nuclear facility. This support provides the county with the abilitytooff-set personnel expenses whenever its EOC is activated. No other county in the region has such a facility nor do other counties receive similar financial support for emergency operations. Even though mostofthe counties closed their EOCs, they remained in a quasi-alert status. Personnel were briefed daily about the alert and were reminded that they were on stand-by status and expectedtorespond immediatelyifthe earthquake occurred.Inaddition, the county OESs were kept quite busy with inquiries from the public. Most public inquiries involved questions about whatanA alert meant, what areas were affected, and what to do. Advise consistedoftelling callerstostock food and water,toknow howtoshut-off the gasifnecessary,toremove objects which might fall duringanearthquake,tolookinthe phone book at the section on earthquake prepared ness, and so on. Many county OESs also experiencedanincrease in requests for materials on earthquake readiness. Most OESs, with the exceptionofFresno County, where Coalinga is located, reported that they were abletoadequately handle public inquiries about the alert. A spokesperson for the Fresno County OES indicated that they were deluged with phone calls the day following the alert. Some callers weresoconcerned astoaskifthey neededtokeep children outofschool. This experiencewasuniquetoFresno County and must be attributedtothe ordeal that manyofthe residentsinthe area had with the devastating Coalinga earthquakein1983. Insteadofcalling county OESs, many citizens used the State's (800) earthquake hotline number which was instituted after the Landers andBigBear quakesinJune of10
thisyear. The availabilityofthis service helped to keep inquiries to county jurisdictionsatmanageable levels during this alert period. County OESs farthest away from Parkfield made staffing-down and closure decisions sooner than those closer to Parkfield. In addition, the numberofpublic phone inquiriesandrequests for materials diminished the farther one traveled from Parkfield. More than anything else, county OESs used this alert situation as a good opportunitytoengage in an unplanned and unannounced drill. When askedifthey would do anything differently, all said that, for the most part,ifanother alert was called they would do exactly what they did during this alert. The alert also provided some OESs with the opportunity to discover glitchesintheir information dissemination procedures. These glitches involved flawsincall-down procedures where some agencies had been in advertently left outofthe information loop. All offices reporting such flaws also indicated that steps were being taken to correct the problem. Furthermore, the alert induced activity to update emergency plans. While the alert was almost exclusively viewed as something useful and positive, some spokespersons expressed concern thatiftoo many A alerts are called the publicmaybegin to outright ignore them. When asked how their own organization would handle multiple A alerts, these same spokespersons indicated that they would continuetotake them seriously and to engageinactions appropriate to the situation as outlined in their emergency manuals. The fact that the Parkfield Experiment has been prolonged over so many years wasnotofconcern to County OESs. County representatives consistently reported that the drills and various other formsofassistance provided by the State has enabled all emergency response organizations to develop good working relationships with one another. Allofwhich has heiped them to become better prepared for earthquakes thantheymight have otherwise become without the extended Parkfield experience.Firedeparlments.Fire departmentsinthe central California area areoffour types----city paid, city volunteer, county, and state forestry service. Response to the alertbyfire departments throughout the region was stlikingly similar for all types except city volunteer departments. Regardlessoftype, all firefighting agencies implemented their previously developed Parkfield alert plans. But implementationofan alert plan did notmeanmuch outofthe ordinary for fire departments since they are always on an alert status. The only unusual activity recognized as necessary for response to the alert wastotake extraordinary actions to protect equipment. Response by city fire departments varied by typeofstaffing and by past experiencewithdamaging earthquakes. Not all city fire departments are staffed by paid firefighters. Some cities have only a volunteer force. Response to the alert was consistent across all cities with paid staffs. For these fire departments the immediate response was to remove all critical firefighting equipmentfrombuildings. This meant that all trucks had to be moved from within the firehousetothe driveway or onto the street. This activity was continued for the entire 72-hour11
alert period for firehouses considered structurally unsafe during earthquakes. However,insome cities, equipment was taken back inside after the first 24-hour period in areas where firehouses had been recently constructed and for those which had been earthquake retrofitted. The Coalinga Fire Department removed equipment for both the A and B alerts. Coalinga's experience with earthquakes made a definite impression on the firefightersofthis community. Upon receiptofthe alert, Coalinga's fire chief authorized the useofthe civil defense siren system. The siren was sounded for several minutes to alert residentstothe impending earthquake threat. No organizationsinany other community sounded a siren alarm as a waytowarn citizens. Volunteer fire departments respondedina uniform manner. All such departments kept equipment inside regardlessofthe conditionofthe firehouse. This decision was based on the rationale that such departments do not have people on duty 24-hours a day and that fire equipment is left unattended for lengthy periodsoftime thereby making it necessarytokeep equipment housed regardlessofthe structural conditionofthe facility or the natural hazard faced. Nevertheless, all volunteers were briefed daily and remindedofthe alert status. County and stateforestry service fire departments also engaged in response activities. Throughout the region all such fire departments, located in the expected areaofimpact, took stepstoregularly brief personnelonthe alert, protect equipment, and maintainanactive liaison with the county OESs. Removalofequipment from firehouses sparkedaninteresting debate among fire fighters. The question raised was whether or not the firefighters should sleep outside with their equipment.Inone case, firefighters did sleep outside during the first nightofthe alert becauseofconcern with theft and vandalism. But during the following nights they slept inside since their equipment had been moved toanenclosed parking lot where it was considered safe.Inother cases, however, thoseincharge made a command decision that it was okay for on-duty personneltosleep inside eveninthose buildings not consideredtobe earthquake safe. While all fire departments, at all levelsofgovernment, actively responded to the alert in ways thought appropriatetotheir organizational make-up, not all personnel perceived the alert alike. Opinions about the alert differed accordingtoposition in the organizational hierarchy. Personnelincommand and decision-making positions viewed the alert as credible and as a good opportunity to review their Parkfield response plans andtoengageinanunanticipated drill. However, "ground-level" personnel tended to view the alert as a nuisance. Even though such personnel followed the ordersoftheir commanders, as would be expected, the alert was generally considered to have provokedanunnecessary drill which required themtoexpend energy best reserved for a real emergency.Police, sheriff, and highway patrol.Policing agencies throughout the region took the alert seriously enough to daily brief personnel on all shifts about the alert and to remind themtobe mentally preparedtorespondtoanearthquake at a moments notice.12
Officers were also reminded that they could be prevailed upontoprovide mutual aid support to jurisdictions which might be more adversely impacted by the Parkfield quake. Some city police departments were also responsible for notifying local political officials about the alert. As with fire departments, police response did not require much activity above and beyond normal operating procedures. More than anything, it induced police, sheriff, and highway patrol departmentstotake stockoftheir disaster level emergency response capabilities. Since vehiclesinmost police stations are parked outside, it was not necessary to move them. Response essentially meant that officers performed additional checks on emergency generators, kept vehicles fully fueled at all times, and inspected emergency supply caches for fresh batteries and adequate amountsoffood, water and other miscellaneous supplies. Several city police departments received a few phone calls from concerned citizens. Such calls were referredtocounty OESs. Someofthese departments also used the alert asanopportunitytopublish announcementsinlocal newspapers about the availabilityofearthquake related literature at their stations. When speaking as department representatives, officers expressed that the alert was a good thingiffor no other reason than to remind them and the community that they are at risk to earthquakes.Onthe other hand, when talking to police officers on the "beat", attitudes about the alert were quite different.Aswith fire departments, ground-level personnel dutifully followed command orders. But on a personal level these officers tended to question the necessityofissuinganalert and raised general doubts as to the credibilityofscientists being abletopredict earthquakesinthe first place. However, while expressing these dubious thoughts, they also said that oneofthe first things they did in their own homes was to make sure that valuable glassware was secured, that pictures, mirrors and plants that could be thrown around were takenoffwalls, and so on. Some officers were rather amused that people should come all the way from Colorado to investigate the alert and suggested that Universal Studios might be a better place to go for an earthquake.Support service organizations.Organizations responsible for support services are very important during earthquake disasters. For example, utility companies are important for furnishing vital powertoanearthquake stricken area, it is important that hospitals be abletoprovide ongoing and extra medical services, agencies like the Red Cross are important for offering shelter and sustenance services, and schools have responsibility for ensuring the health and safetyofchildren. Even though these responsibilities are unique to each typeoforganization, they are all critically important during timesofdisaster. However, while each organization mayplayacritical role during a disaster, this did not mean that all such organizations respondedtothe alert in similar fashion nor did they similarly take the alert seriously. Observed differencesinhowthese support organizations viewed and responded to the alert liesintheir perceived rolesofresponsibility. Some organizations, such as13
utility companies, have roles that are indirectly linkedtothe public in termsofemergency response. For example, utilities respondtoemergencies by restoring lost power by repairing and maintaining equipment not people. Other organizations, like the Red Cross and hospitals, have roles that are directly linkedtothe public. For example, providing shelter and emergency medical services which necessitates direct involvement with people. The utility company providing electric servicetothe region perceives that its role during any disaster istorestore energy as quicklyaspossible, and that providing this much needed service is dependent on the extentofdamage the utility company itself sustains duringanearthquake. But, since there isnowayofknowing aheadoftime what that damage might be, the reasoning is that there isnoneed to doanything in advance. This major utility did check to make sure that relay switchesinsubstations were all okay and that there werenocracked insulators, but this action was done in responsetothe precursor quake, notinresponse to the alert. In general, there was simply no expressed interest in the alert on the partofthe major utility company serving the Parkfield region. Drills were not conducted, emergency plans went unreviewed, vehicles were not kept completely fueled, andany needtostep outside a "businessasusual" modeofoperation was not given consideration. While the Diablo Canyon Nuclear Power Plant may not have a specifically defined emergency response roletoresidentsofthe area, it is responsible for ensuring that it remains safe during disasters and that the public is not exposedtorisk from a nuclear accidentofsome sort.Asper legal requirements, the Diablo Canyon nuclear power plant notified the Los Padres Divisionofthe PacificGasand Electric Company that fuel was not being transferred and that all nuclear fuel was safely contained. Diablo Canyon engaged in full alert procedures for the entire 72-hour period. This meant that no equipment could be sent out for repairinthe event that it would be needed, fuel transfer operations were halted, and all piecesofequipment that could possibly fall were placedinsecured positions. Organizations such as the American Red Cross and hospitals have responsibility for providing health and welfare services during a disaster period. These response roles require a direct involvement with the public. American Red Cross chapters and hospitals, serving the region, got ready to respondtoanearthquake as soon as noticeofthe alert was received. Red Cross chapters alerted volunteers that they might be neededtostaff shelters, representatives kept tabsofupdate noticesonthe alert and attended county briefing sessions, and most offices generally prepared foranonslaughtofpublic inquiries. Many chapters used the alert as an opportunitytoget readiness information featuredinlocal newspapers. Some chapters also used it as a way to encourage volunteerstoattend training classes. However, the public generally did not turntoRed Cross offices for additional information on the alert. Even though hospitals are generally considered always ready for emergencies, the alert induced reviewsofearthquake plans, additional inspections on emergency14
generators, and taking stockofemergency supply caches in hospitals throughout the region. A major hospital serving the northern San Luis Obispo area, and one which is a triage facility for disasters, reviewed its triage plans and sPeCifically placed triage teams on stand-by status. Public school districts also have responsibilities directly linked to the public.Ifan earthquake occurs during school hours, school officials are responsible for the health and safetyofchildren. Schools are also viewed as possessing responsibility for teaching children about earthquake readiness and about safe responsetoactual earthquake events. Throughout the region, schools wereeSPeCiallyactive during the alert. At a minimum, virtually all schools conducted at least one duck and cover drill with students, staff and faculty. Some schools conducted several drills during the alert period. And some even conducted evacuation drills. Many schools also took precautionary actions such as keeping drapes drawn and inspecting facilitiestomake sure that objects which could be thrown around duringanearthquake were adequately secured. While all schools conducted drills and reviewed emergency plans, the Coalinga Huron School District initiated this process well before receiving word about the alert from its Fresno County Education Department. Most residents, including school offi cials, in and around Coalinga had felt the 4.7 magnitude precursor quake. This event combined with the school districts experienceofgreat loss from the 1983 quake, induced officials to engageinearly response. School officials simply called a drill based on the quake felt the night before, they did not wait for official wordofanalert.The media.Media response turned the alert into quite the event. The roleofboth broadcast and print media organizationsinproviding informationtothe public is undeniabl y necessary. While this role is certainly taken seriously, the tendencyofthe media to sensationalize and exhaustively reportonthe alert,tothe pointofrepetition, gave it a character and life it likely would have not created on its own. Television and newspaper press agents from all pointsofthe state began arriving in the Parkfield areaearly the morning following its announcement. Mobile broadcast units from San Francisco, Bakersfield and Los Angeles were visible, not only in Parkfield, butinPaso Robles and Coalinga as well. The presenceofthemedia gave the alert a party like quality. Radio broadcasters were also very visible. One southern California radio personality went so far astobroadcast a live morning show from a cafe in Paso Roblesbyusing a cellular phone. Some radio broadcasters trivialized the alert and tended to joke about it. For example, one broadcaster would initiateanhourly update by saying "...well, we've survived one more hour." In the midstofnational election campaigns, the alert made headline news in virtually all regional newspapers for the three daysofthe alert. Some stories addressed the reactionofpeopleincities close to Parkfield, but most presented interviews with seismologists,state officials, and residentsinParkfield. While waiting for the earthquake to happen, the primary roleofUSGS and state representatives was to sit around and give interviewstothe press.15
Ina way, the press directed the alert. For example, the original plan was for the State OES to set-up its mobile command unit in Cholame, located just southofParkfield in an area less vulnerable to electronic communication impediments. But since the press converged on Parkfield, it was necessary to move the command post into Parkfield. The presenceofsuch a great numberofpeople dependent on electronic telecommunication capabilities,inthe small canyon-like environmentofParkfield, created a communications black-out. While state representatives made a special requestofPacific Belltoinstall additional telephone lines, it wasn't until the week after the alert was over that this request was met. Print and television media covered the event, for the most part, in a thorough and responsible fashion. Newspaper articles reported the specificsofthe alert, as well as more general information about what to dotoget ready for earthquakes. Almost all articles were followed with details about preparedness actions the public should taketoget ready. Television coverage also presented such information. For example, one morning news show made a quiz game about whattodotoget ready. The local tele vision station, KSBY, offered a freeEarthquake Survival Guidebooklet to those willing and able to go to the stationtopick it up. About25ofthese booklets had been given away by the final dayofthe alert. The public was kept well informedofthe parametersofthe alert,ofits scientific caveats, andofthe usefulnessofthe Parkfield prediction experiment. However, the tendency to depict the detailsofthis short-term prediction, specifically in termsofits likelihoodofoccurrence (37%orgreater chance), may have impeded people's ability to personally define the quake as potentially happeningtothem. More than anything else, it likely provided the public with the message that scientists were unsureoftheir ability to predict earthquakes. Media coverage also tended to focus obsessively on the $19 million spentbythe USGSonthe prediction experiment.The PublicThe public heard about the alert almost exclusively from media sources. As a matter-of-fact, the public could hardly escape hearing about it. Radio and television news agencies began broadcasting scripted news releases almost immediately upon re ceipt.Byearlyinthe morning following its official announcement, the alert was being detailed extensively on local, regional and statewide news programs.Italso appearedasheadline newsinvirtually all local and regional newspapers. And for the most part, it remained headline news until it expired three days later. Residentsofthe area were also hard put nottonotice the presenceofmany more outsiders than would normally be expected at this timeofyear. Many media reporters, state officials, USGS personnel, and researchers frequented the region for several days.16
Atthe very least, this event provided the local motel and restaurant industry with a mini boon to their cash registers. Important questions to answer from this alert were, how would the public perceiveanA alert and what would citizens doinresponse? For example, it might be anticipated that the public would ignore the alert or, at the other extreme, take the alertsoseriously that it would engage in panic like behavior. Another concern was thatifthe quake, which the alert predicted, didn't happen, as was the case, would the public perceive itasan alarmistor"cry wolf' situation? Neither the extreme reaction nor the "cry wolf' perception was observed. Instead, the public perceived the alert as a sincere and honest effort on the partofscientiststopredictanimminent earthquake and to share concerns about potential danger with the public at-large. There seemedtobe a general knowledge that earthquake prediction isanimprecise science which implicitly meant that certaintyinthe prediction coming true was impossible. The public likewise took the middle groundintermsofresponsetothe alert. While the alert was not ignored and was taken seriously enough to warrant some protective measures, the public did not engageinextreme behaviors. Hearing about the alert,inallofits detail, did not necessarily mean that the public perceived the alert as defininganimminent earthquake threat. Nor did it mean that the public engagedinsubstantial amountsofreadiness behaviors. This is not to say that the public did not take the alert seriously enough to engage in some readiness behaviorsina rational way. But, membersofthe public who felt the precursor quake were much more likelytoengageinreadiness activities than those who did not feel it. For the most part, the public perceived the alert as real and credible, but people did not allow their perceptionsofthe alerttodisrupt daily routines. And the extentofpublic response varied by proximitytoParkfield and experience with damaging earthquakes. Another factor variably affected responsetothealert-feelingthe 4.7 precursor quakeonMonday night. For example, peopleinPaso Robles and Coalinga took the alert more seriously and engagedinmore readiness behaviors than did people in A venal, Hanford,SanLuis Obispo, and Taft. The residentsofCoalinga and Paso Robles also more frequently reported having felt the precursor quake. The occurrenceofthe precursor quake which was strong enoughtobe felt not only acted as a precursor for callinganalert, it also appearstohave acted as a precursor to readiness activity. Merchantsinthe area seemtoimplicitly understand this phenomenon. For example, retailers related that they could predict purchasing behavior based on earthquake activity. When peoplefeelanearthquake they tend to buy more water, batteries, and other readiness supplies. But what really increases buying isifone earthquake is followed-upbyanother. The follow-up quake acts as a reinforcementofconcern raised by the first quake. Coalinga was by far the most active and alertofall communities visited. This wasnodoubt due, in large part,toCoalinga's experience with a devastating earthquake in 1983. Virtually everyone interviewed in Coalinga broughtupthe 1983 quake. In17
addition, nearly everyone felt the 4.7 precursor quake which ledtothe alert. Coalinga's experience with a damaging earthquake, which has become a significant event in its cultural history, combined with feeling the precursortomake the alert a very salient issue. Overall, the public engagedinreadiness activities which could be forthrightly accomplished and which were easily affordable. Purchasing bottled water and fresh batterieswere the most noted actionsinwhich the public engaged. Some people reported removing pictures and mirrors from walls and placing breakable items such as glassware and crystal in safer areas. However, behavior such as removing pictures and mirrors more demonstrates that the public may not take its general earthquake vulnerability very seriously in the first place.Ifsuch actions had been accomplished on a more permanent basis prior to the alert, it would not have been necessary to do so during the alert. Coalinga is probably the best prepared and most vigilantofall communities in the Parkfield region. Ironically, it is also likely the most able community to structurally withstand earthquakes. However, this abilitytosurvive future quakes came at a dear price when the majorityofstructures were severely damagedinthe 1983 earthquake. But it is that event which ledtothe rebuildingofCoalinga inanearthquake sound fashion. The earthquake experienceof1983 madeanindelible impression on the people ofCoalinga. The alert also made residentsofCoalinga more "edgy", there was a senseoftension in the air and stores ran outofwater, batteries, and other such supplies.FindingsThe 4.7 magnitude precursor quake, which ledtoissuing the first A-level alertofthe Parkfield Earthquake Prediction Experiment, was Mother Nature's wayofcommuni cating to her inhabitants that they liveinearthquake country. And the official A alert provided a direct means for the USGS and StateOBStocommunicate Mother Nature's warning to other organizations and to the public. More than anything else, the alert served as a reminder,tothe public andtoorganizations serving the public, that they occupy earthquake territory. More importantly, it provided an opportunity for organizations and residents alike to take stockoftheir earthquake readiness capabilities.AnEarlhquake Culture?Impressions derived from this case study suggest that mostofcentral California possesses, at best, only a cursory earthquake culture. People throughout the central California region are aware, at one level or another, that they live in earthquake country. However, keeping this awareness at such a level so that appropriate readiness becomes18
a partofthe cultural fabricofthis society is yet another matter. Perhaps it is to be eXPeCted that when a public constantly lives under the threatofnatural events, which are largely unpredictable and without season, thatitwould become complacent and fatalistic. Such an understanding, however, must not allowusto dismiss the need for an earthquake threatened society to develop ways to sew readiness behaviors into its cultural fabric. This alert provided a good opportunity to sew one more threadofearthquake salience into the social fabricofcentral California. While information about the Parkfield Experiment, beginning in 1985, induced a measurable increase in public readiness behaviors (Mileti, Fitzpatrick and Farhar 1990), this A alert revealed that there is much more to be accomplished before the Parkfield region can be considered truly ready for a damaging earthquake. There is ample evidence to suggest that central California has not yet developed an earthquake culture, eSPeCially ifsuch a culture entails a routine senseofpermanent readiness behaviors.Forexample, when people said that they had taken pictures off the walls over their children's bedsinresponse to the alert, by inference they indicated that they have not yet developedanearthquake culture.Ifthey had, such behaviors would not have been necessary since boltingofpictures to walls would be routineinanearthquake culture.What the Alerl MeantWhat the alert meant for emergency response agencies was to "hurryupand wait." The initial decision facing these organizations was what to do immediately and then whether to maintain that decision. News organizations, as would be expected, were the principle mediumofcommuni cation for getting wordofthe alert to the public. Response by media was immediate and thorough, and the presenceofjournalists could not go unnoticed. As with the initial Parkfield prediction announcementin1985, this alert became quite the media event. In some ways media response was the most visibleofall organizational response and in part actually determined response by other organizations.Localand state agencies all did somethinginresponse to the alert. The standard "company-line" was that all alerts must be taken seriously and plans implemented. However, such statements may be due more to recognition that plans exist and must therefore be followed than due to perceiving and believing the actual risk an alert is intended to convey. The degreeoflocal organizational response to the alert varied depending on available resources, proximity to Parkfield, and past experience with damaging earthquakes.Ifnothing else, the alert caused organizations to take their emergency response manuals offofshelves, dust them off, review them, and update them at a time other than when fiscal bureaucratic procedures call for it to be done.19
What the alert meant for citizens essentially amounted to a reminder that they are at continued risk to earthquake activity. Beyond such a reminder, the alert, and the media coverage surrounding it, brought to the attentionofcitizens the concerted efforts on the partofgovernment, at all levels, to address the myriad problems facing earth quake-prone California. But perhaps more than anything else,itreminded all involved that Mother Nature can be a tease at times and that successful short-term prediction of earthquakes continues to resideinthe future.UsefulnessofPlanningThe widespread implementationofearthquake plans by organizations demonstrates the usefulnessofdeveloping emergency response plans. The useofdeveloped plans facilitates organizational response and constrains tendencies by individuals to rely upon personal perceptionsofriskindeciding what to do. This usefulness is apparent when looking at individual versus organizational perceptionsofthe alert. Personnel in all organizations did somethinginresponse, regardlessofthe predisposition by individuals in those organizations to dismiss the alert as meaningless. Even though many operations personnelmayhave personally viewed the alertasa nuisance, the organization's response plans took precedence, and were unquestionably implemented.Imporlant Influential FactorsTwo factors, more than any others, appear to influence the developmentofpercep tions about short-term predictions which then lead to readiness behaviors-feeling precursor earthquake activity and prior experience with damaging quakes. Precursor quakes, large enough to be felt but notsolargeasto cause damage, would likely enhance taking short-term earthquake predictions seriously enough to engage in rational readiness activities.Incommunities with damaging earthquake experience, the precursor quake provided an especially good reminder to be earthquake ready while simultaneously acting as reinforcement for the short-term prediction. Effectively communicating warnings about increased earthquake risks is undeniably important.Animportant phaseinthe processofcommunication is the forming and shapingofrisk perceptions. And the mannerinwhich the risk is depicted can significantly influence how perceptions are developed (Fitzpatrick and Mileti 1991; Mileti and Fitzpatrick 1991). Social researchonthe Parkfield Experiment (Mileti, Fitzpatrick and Farhar 1992) revealed that people prefertodeal with earthquake riskina dichotomous fashion. They attach meaning about future earthquakes as either something germane to them or as something that does not apply to them.Inother words, they have20
noneed to attach probabilityofoccurrence definitions in as refined a manner as do scientists. As in developing perceptionsofrisk in the long-term Parkfield prediction, in this short-term situation, membersofthe public simply decided whether they wereorwere not at increased risk to a damaging earthquake. And this decision was primarily basedonpast experience with damaging earthquakes, feeling the precursor quake, and the influenceoforganizations such as schools. People were more likely to use the probability statements issued in the A alert more as a way to define the scientists' sense of inaccuracy and uncertainty than as a way to define their own personal risk. Offering probability statements may very well have been confusing to the public and may have helped some people decide that they were not at risk when indeed they were. Perhaps it would be best to simply state to people that they are at increased risk rather than providing them with unnecessary excessive probability data. In the future, therefore, it might be more wise to adhere to the "KISS" ("keep it simple stupid") philosophy when attempting to effectively communicate to the public about increased vulnerability to a specific earthquake event.21
ReferencesCalifornia OfficeofEmergency Services. 1992. "Parkfield 'LevelA'Alert." Situation report. Sacramento, CA:20 October. Governor's OfficeofEmergency Services. 1988. Parlifield Eanhquake Prediction Response Plan.Sacramento, CA: StateofCalifornia OfficeofEmergency Services. Fitzpatrick, Colleen. 1992.Constructing Lines-oj-Conduct: Theoretical Explanations with Empirical Evidence.Unpublished Ph.D. dissertation. Fort Collins, CO: DepartmentofSociology, Colorado State University. Fitzpatrick, Colleen and DennisS.Mileti. 1991. "Motivating Public Evacuation."International JournalojMass Emergencies and Disasters9(2):137-152.Mileti, DennisS.and Colleen Fitzpatrick. 1991. "CommunicationofPublic Risk: The Theory and its Application."Sociological Practice Review1(3):20-28.Mileti, DennisS.and Colleen Fitzpatrick. 1992. "The Causal SequenceofRisk Communicationinthe Parkfield Earthquake Prediction Experiment."Risk Analysis12(3):393-400. Mileti, Dennis S., Colleen Fitzpatrick and Barbara C. Farhar. 1990.Risk Communication and Public Response to the Parlifield Eanhquake Prediction Experiment.Final reporttothe National Science Foundation. Fort Collins, CO: Hazards Assessment Laboratory, Colorado State University. Mileti, Dennis S., Colleen Fitzpatrick and BarbaraC.Farhar. 1992. "Fostering Public Preparations for Natural Hazards: Lessons From the Parkfield Earthquake Prediction."Environment34(3): 16-20,36-39. Mileti, DennisS.and JaniceR.Hutton. 1987. "Initial Public Response to the 5 April 1985 Parkfield Earthquake Prediction." Boulder, CO: the Natural Hazards Research Applications and Information Center, UniversityofColorado. Southern California Earthquake Preparedness Project. 1988. "The Parkfield Earthquake Prediction." Mailed brochure preparedbythe Governor's OfficeofEmergency Servicesincooperation with the California DepartmentofConservation, DivisionofMines and Geology and the United States Geological Survey. Pasadena, CA: Southern California Earthquake Preparedness Project.22