Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
VOLGEO-05104; No of Pages 18
Contents lists available at SciVerse ScienceDirect
Journal of Volcanology and Geothermal Research
j ourna l homepage: www.e lsev ie r .com/ locate / jvo lgeores
Lessons learned from the 2010 evacuations at Merapi volcano
Estuning Tyas Wulan Mei a,b,⁎, Franck Lavigne a,1, Adrien Picquout a,1, Edouard de Bélizal a,1,Daniel Brunstein a,1, Delphine Grancher a,1, Junun Sartohadi b,2, Noer Cholik c,3, Céline Vidal a,1
a Université Paris, 1 Panthéon Sorbonne, Laboratoire de Géographie Physique, UMR 8591 CNRS, France, 1, Place Aristide Briand Meudon 92195, Franceb Universitas Gadjah Mada, Fakultas Geografi, Sekip Utara, Jalan Kaliurang, Bulaksumur 55281, Indonesiac BPPTK (Balai Penyelidikan dan Pengembangan Teknologi Kegunungapian), Jalan Cendana 15, Yogyakarta 55166, Indonesia
⁎ Corresponding author at: Université Paris, 1 PanthGéographie Physique, UMR 8591 CNRS, France, 1, P92195, France. Tel.: +33 145075552.
The rapid onset and largemagnitude of the 2010 eruption ofMerapi posed significant challenges for evacuationsand resulted in a peak number of almost 400,000 Internally Displaced Persons (IDPs). A pre-existing hazardmapand an evacuation plan based on the relatively small magnitude of previous eruptions of the 20th century wereutilized by emergency officials during the initial phase of the eruption (25 October–3 November, 2010). Howev-er, when themagnitude of the eruption increased greatly on 3–5November 2010, the initial evacuation plan hadto be abandoned as danger zones were expanded rapidly and the scale and pace of the evacuation increased dra-matically. Fortunately, orders to evacuate were communicated quickly through a variety of communicationmethods and as a result many thousands of lives were saved. However, there were also problems that resultedfrom this rapid and larger-than-expected evacuation; and there were lessons learned that can improve futuremass evacuations atMerapi and other volcanoes.We analyzed the results of 1969 questionnaires and conducteda series of interviews with community leaders and emergency officials. Results were compiled for periods bothduring and after the 2010 eruption. Our results show that: (1) trust in the Indonesian government and volcanol-ogists was very high after the eruption; (2) multiple modes of communication were used to relay warnings andevacuation orders; (3) 50% to 70% of IDPs returned to the danger zone during the crisis despite evacuation or-ders; (4) preparation before the eruption was critical to the successes and included improvements to roadsand education programs, (5) public education about hazards and evacuation protocols before the eruptionwas focused in the perceived highest danger zone where it was effective yet, confusion and loss of life in otherareas demonstrated that education programs in all hazard zones are needed to prepare for larger-than-normaleruptions, and (6) improvements in registration of evacuees, in providing for livestock, and in activities andwork programs in evacuation camps (as well as government restrictions and policy changes) are also neededto prevent evacuees from returning to their homes during the crisis period.
The concept of evacuation is simple; it is to move people from athreatened area to a safer location (Southworth, 1991; Zelinksy andKosinsky, 1991). However, evacuations during a major crisis are chal-lenging; for example in the case of Hurricane Katrina in the UnitedStates in 2005, it was not possible to evacuate everyone because there
ssons learned from the 2010 e.jvolgeores.2013.03.010
were hundreds of thousands of people at risk. A mass evacuation isvery expensive, but it always costs less than the damages and recoverycosts of a disaster (Woo, 2008). Our study of the 2010 eruption and theresulting mass evacuation at Merapi volcano validates these relation-ships and provides insights into procedures and activities that couldbe used to improve the success of future evacuations.
Through the development of monitoring (Newhall et al., 2000) andrisk management of volcanoes (Blong, 2000; De la Cruz-Reyna andTilling, 2008), evacuation of population at risk has become increasinglycommon (Tobin andWhiteford, 2002). Evacuation is an effectiveway tominimize fatalities or injuries as referred to in the “road to zero victims”policy encouraged by the Indonesian government in facing natural di-sasters. However, in disaster management policy overall, evacuation isonly one phase of the activities in a series of prediction, response and re-covery (Schramm and Hansen, 1991).
There is one major difference between evacuations during volcaniceruptions and other crises: the duration of the evacuation period isnot certain (Gaudru, 2005). The uncertainty of the volcano's dangerous
vacuations atMerapi volcano, Journal of Volcanology and Geothermal
2 E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
period could lead to inevitable difficulties for the authorities, as well asthe population, and it may create frustrations among the populationduring the period of evacuation. The complex nature of a volcanic crisismay also lead to difficulties in evacuation management: (1) when theexpected/feared volcanic activity does not happen, e.g., Baker volcanoin Washington in 1975 (Frank et al., 1977); Akutan and Iliamna volca-noes in Alaska in 1996 (Lu et al., 2000; Roman et al., 2004), or atIwate volcano in Japan in 1998 (Aizawa et al., 2009); (2) when thereare gaps in communication, for example during the1985 eruption ofNevado del Ruiz volcano in Colombiawheremore than20,000 residentsof Armero city were killed as a result of a failure in communications(Hall, 1990; Voight, 1990; Fielding, 2002); and (3) when there is resis-tance from people to leave the area at risk, such as seen during theMount St. Helens eruption in 1980 (Green et al., 1981; UN, 1985).
At least 9.6% of the world population lives in areas that may beaffected by volcanic activity andmost such areas are located in develop-ing countries (Tilling, 2008). In pyroclastic density current (PDC) proneareas, evacuation is considered to be the most effective means to mini-mize the risk and loss of life. However, successful evacuations dependon several interrelated factors: (1) effective crisis management; (2) un-certainty in the occurrence and consequent difficulties in the predictionof eruptions (Tilling, 2008); (3) difficulties in physical aspects of evacu-ation, e.g., topographic constraints or insularity (Gaudru, 2005); (4) so-cioeconomic conditions, e.g., evacuation refusal due to economic factors(Hewitt, 1983; Gaillard and Le Mason, 2007; Mei and Lavigne, 2012);and (5) problems related to socio-cultural perception of the volcano(Carlino et al., 2008).
In order to implement a successful evacuation, five fundamentalquestions need to be considered: who, when, where, how and howlong. The 2010 Merapi evacuation led to displacement of almost400,000 people living within 20 km from the summit for one and ahalf months. This evacuationwasmuch larger and longer than anticipat-ed by contingency planners. In this paper, we analyze this mass evacua-tion chronologically and we discuss the challenges and constraints itpresented for disaster management, as well as lessons learned that canhelp guide future evacuation policy for such large volcanic evacuations.
2. Merapi volcano
Merapi volcano is one of the most active volcanoes worldwide,with more than 70 eruptions since 1548 (Voight et al., 2000b). Thevolcano is a basaltic–andesite stratovolcano (Andreastuti et al.,2000), located 30 km north of the inner city of Yogyakarta (Fig. 1).
2.1. The paroxysmal eruption of Merapi in 2010
Merapi volcano is known for its frequent small to moderate erup-tions and dome-collapse pyroclastic flows (Surono et al., 2012). From1913 to 2001, the activity of the Merapi volcano was largely confinedto the western and south-western flanks (Camus et al., 2000; Newhallet al., 2000). However, in June 2006, for the first time since 1942, pyro-clastic flows shifted to the Gendol valley towards the south (Thouret etal., 2010; Charbonnier and Gertisser, 2012; Ratdomopurbo et al., in thisissue). In 2010, anunusually largemagnitude explosive eruption causedover 350 fatalities and generated ripple effects in the social and eco-nomic sectors of this densely populated region of central Java. Thelargemagnitude of the 2010 eruption validates the prospective analysismade byNewhall and others ten years before (Newhall et al., 2000) thatthe regular effusive and dome-collapse eruption cycle at Merapi couldbe broken by a larger explosive eruption in the coming decades(Charbonnier and Gertisser, 2008).
2.2. When hazards meet elements at risk
Since the fourteenth century, 61 eruptions of Merapi have killedover 7000 people (Lavigne et al., 2000). Notable among these was
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010Research (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
the eruption in 1672, which killed 3000 people, and the highly explo-sive eruption in 1872, which killed 200 people (Hartmann, 1934).More recently, the eruptive events in 1930–1931 and 1954 killed1400 and 54 victims respectively (Thouret et al., 2000). In 1961, a12-km long pyroclastic flow destroyed more than eight villagesalong the Batang River, killing six persons and in November 1994, apyroclastic flow reached 7 km distance down the Boyong River(Wilson et al., 2007). Due to the unpredictable nature of the 1994dome's collapse, the lack of short-term precursors (Voight et al.,2000a) and the presence of a hill which triggered a decoupling ofthe dilute ash-surge from the basal valley-confined flow, 69 peoplewere killed by pyroclastic flows and 6000 people were evacuated dur-ing and after the 1994 event (Abdurachman et al., 2000). And in 2006,an avulsion of a pyroclastic flow killed two people near the Gendolriver channel at Kaliadem village. More than 22,000 people wereevacuated in 2006 (UNOCHA, 2006) and the Kaliadem touristic arealocated on the upper Gendol river was buried by a pyroclastic deposit(Charbonnier and Gertisser, 2008).
The slopes of Merapi have a high population density, of 935–1901inhabitants/km2 (BPS, 2008). More than 50,000 people live in “dangerzone III” (Kawasan Rawan Bencana III in Indonesian, or KRB III) (BPS,2010). Rapid population growth of over 3% annually took place from1976 until 1995 in this most dangerous zone (Thouret et al., 2000).This rate of growth exceeded the national population growth ratefor the same time interval. This increase is due to environmental fac-tors (soil fertility and water availability) and to socioeconomic factors(development of roads, electricity, telecommunications, fresh watersystems, and touristic areas).
The people living on the flanks of the Merapi volcano are highlyvulnerable for several reasons:
(1) Socio-economic factors of vulnerabilityThe majority of Javanese people living on the slopes of a volcanoconsider that the link between volcanic environment and theirsocial life to be very strong (Lavigne et al., 2008). The volcano isa sacred place, which offers natural resources that can be usedby people for their daily needs. Natural resources are crucial forthe livelihood of local communities because the majority of peo-ple living on Merapi's flanks are farmers, stock farmers, and vol-canic sand miners.
(2) Socio-cultural factors of vulnerabilityTraditional beliefs are strongly influenced by the local com-munity as documented in the Philippines (Gaillard and LeMason, 2007), in Indonesia (Schlehe, 1997, 2007; Texier,2007; Texier et al., 2009) and in Mexico (Gavilanes-Ruiz etal., 2009). These studies also point out that the perception ofthe volcano can affect decision-making in facing volcanic di-saster. At Merapi volcano during the 2006 eruption, MbahMarijan, the volcano's gatekeeper, gained notoriety by stayingin his village despite Merapi's imminent threat and repeatedrequests to follow the evacuation order made by the local gov-ernment (Texier et al., 2009). During the 2010 eruption 34people remained with Marijan in their village of Kinahrejo(Fig. 2); they all perished.
2.3. Managing volcanic disasters at Merapi
There are five levels of government administration in Indonesia; na-tional (pusat), provinces (propinsi), districts (kabupaten/kota), subdis-tricts (kecamatan) and municipalities (kelurahan/desa). A municipalityencompasses several villages (dusun). The volcanic crisis managementis organized in each district and is based on the recommendations ofthe Center for Volcanology and Geological Hazard Mitigation (CVGHM),an institution in charge of assessing and monitoring volcanic activity inIndonesia.
evacuations atMerapi volcano, Journal of Volcanology and Geothermal
Fig. 1.MapofMerapi volcano and surrounding areas showingmajor government districts and subdistricts, Yogyakarta city,major rivers and danger zones (KRB I, II, and III) as revised afterthe 2010 eruption. Inset maps show the location of Java within the Indonesia region of the western Pacific, and the location of the Merapi area within Java.
3E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
An updated hazard map for Merapi displaying three danger zones(KRB) ranked from III (high) to I (low) (Fig. 2) was constructed,printed and made available online by CVGHM in 2002 (Hadisantonoet al., 2002). The KRB III encompasses areas located close to thehazard source, and in the case of Merapi, frequently affected by pyro-clastic flows, lava flows, rock falls and ejected rock fragments. Theboundaries of KRB III are based on study of the impacts of eruptionsthat occurred throughout the 20th century and take into accountthe distribution of mainly small-volume dome-collapse pyroclasticflows as well as morphological changes in the summit area that influ-ence the directions of dome collapse. The danger zones also take intoaccount geologic structure and stability of the summit and the loca-tions of the most recent eruptive activity. The KRB II zone may alsobe affected by pyroclastic flows, lahars, volcanic ash fall, volcanicbombs and other ejected rocks, although the frequency and severityof impacts are considered to be lower than in KRB III. Zone KRB I is po-tentially affected only by lahar and floods during typical Merapi erup-tions (Sayudi et al., 2010). An updated (post-2006, pre-2010) Merapihazard map was intended for situations in which (1) the eruption oc-curs in the central area of the summit dome complex, (2) any associ-ated ash column is vertical, and (3) there has been no drasticmorphological change in the volcano's summit. This map wasintended to assist volcanologists in describing the pattern of past
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010 eResearch (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
eruptions and to estimate the areas potentially affected by varioushazards (Suryo and Clarke, 1985). The updated Merapi hazard map(Fig. 2) was used by local authorities as an input for contingency plan-ning that took place in 2009 in each district surrounding the volcano.
Beside the hazard map, CVGHM also provides four warning levelsof volcanic activity (Table 1). For each warning level recommenda-tions are given for what people living on the volcano slopes are sup-posed to do (e.g. no need for concern, stand by for technical directionissued by CVGHM, prepare to evacuate, evacuate). There are five ob-servatory posts at Merapi: Kaliurang, Babadan, Ngepos, Jrakah andSelo (Fig. 2). Information about the condition and the morphologyof the volcano is reported from each post to CVGHM's Volcano Inves-tigation and Technology Development Office (Balai Penyelidikan danPengembangan Teknologi Kegunungapian, or BPPTK) and to the MerapiVolcano Obsevatory (MVO, a section of BPPTK) in Yogyakarta andthen transmitted to CVGHM. The same information is also reportedto local governments. The National Disaster Management Agency(Badan Nasional Penanggulangan Bencana, or BNPB) and the local au-thorities at district level are in charge of disseminating the alert levelto the public following the established communication protocolshown in Fig. 3. Evacuation orders are given by BNPB and local gov-ernments at district level. However, if the danger is imminent, theBPPTK can use sirens to inform people directly to evacuate.
vacuations atMerapi volcano, Journal of Volcanology and Geothermal
Fig. 2. Merapi hazard maps. a, Before the 2010 eruption, based on Hadisantono et al. (2002) with revisions based on the 2006 eruption. b, Following the 2010 eruption, based onHadisantono et al. (2002) and Sayudi et al. (2010).
4 E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
The Regional Disaster Management Agency (Badan PenanggulanganBencana Daerah, or BPBD) is the coordinator of crisis management. In2009, the district governments associated with BPPTK, along with
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010Research (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
UNICEF and several other institutions created a contingency plan inorder to deal with future Merapi crises. At that time a scenario forMerapi's eruption was prepared for each district (Table 2) and the
evacuations atMerapi volcano, Journal of Volcanology and Geothermal
Table 1Warning levels for volcanic eruptions in Indonesia (Sayudi et al., 2010).
Level Name Criteria Implications
I Aktif normal (normally active) Monitoring of visual, seismicity and other volcanicevents does not indicate change.
People living in hazard zonesIII, II and I are able to perform daily activities.
II Waspada (on guard) Increasing activity of seismicity and other volcanicevents, and visual changes around the crater.
People living in hazard zones II and I are permitted toperform daily activities, but people living in hazardzone III should be more aware and attentive to thetechnical direction issued by CVGHM.
III Siaga (prepared) Intense increase in seismic activity,and obviouschanges of visual observation in the crater.
People must be prepared to evacuate. The evacuation isorganized by the local authority at district levelfollowing the technical advice issued by CVGHM.
IV Awas (beware) The eruption is about to begin. People must be evacuated to safer places.
Note: Italics indicate Indonesian language equivalents of words in parentheses.
5E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
roles of each institution involved were defined (Anonymous, 2009a,b,c). As also shown in Fig. 4, coordination of evacuations is handled byone ormore institutions: the Indonesian Department of Transportation,the Department of Public Works, and the Army. Official transport suchas busses, trucks or cars are provided by the Indonesian Departmentof Transportation, Department of Health, Social Department, Army, Po-lice, by the Sub-district Government, and/or by Non-Government Orga-nizations (NGOs) and by a number of other organizations andvolunteers. In addition, private vehicles are used (Fig. 3).
3. Material and methods
This study is based on five methodological approaches (Table 3):
(1) Interviews with key informantsWe conducted semi-structured interviewswith officials from thefour districts surrounding Merapi and with two municipalleaders and an official from the Department of Public Works
Fig. 3. Diagram showing the organization and responsibilities of national, regional, and location of Merapi volcano. Terms in italics are Indonesian language equivalents.
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010 eResearch (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
(Table 4). Interviews were conducted both before (May–August2010) and after the 2010 eruption (December 2010–March2011). This methodwas used to determine the various strategiesand measures used by the institutional actors in the region. Theinterviews allowed us to analyze problems and challenges inevacuationmanagement presented by the 2010Merapi crisis. In-terviewees were chosen for their functions within their host in-stitutions and based on their personal experiences during the2010 volcanic response. They were also selected to be represen-tatives of the institutions we surveyed. Interviews were done inIndonesian and Javanese languages. The interview processallowed the interviewer to repeat the questions, accept answersthat were not expected and then verify them, even though thistook additional time (Houtkoop-Steenstra, 2000).
(2) Questionnaires to Internally Displaced Persons (IDPs)During the crisis, questionnaires were distributed to IDPs in thecamps in order to analyze the community response to evacua-tion. The questionnaire-based survey as a tool for understanding
l authorities during volcanic evacuations in Indonesia as utilized during the 2010 erup-
vacuations atMerapi volcano, Journal of Volcanology and Geothermal
Table 2Eruption scenario from the 2009 contingency plan (Anonymous, 2009a,b,c).
District Eruption scenario Threatenedpeople
Expectedduration forevacuation
Sleman 12 km pyroclastic flows in the valley ofGendol and 7–8 km pyroclastic flows in thevalleys of Kuning, Boyong, Bedog and Krasak.
12,660 4 months
Magelang 15 km pyroclastic flows in the valley ofBebeng, 7–8 pyroclastic flows in the valleyof Putih and Blongkeng, and 7 kmpyroclastic flows in the valley of Pabelan.
32,987 2 months
Klaten 10 km pyroclastic flow to the west andsouthwest of the summit.
4420 2 months
6 E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
population behavior during a volcanic crisis has been used inmany studies (Gaillard and Le Mason, 2007; Texier, 2007;Lavigne et al., 2008; De Bélizal et al., 2011a). Respondentswere randomly chosen within countless IDPs. They includedmen and women aged between 7 and 95 years old. We collectedasmany respondents as possible such that conclusions about thetarget population would be statistically significant (Fowler andCosenza, 2008). We carried out two surveys: the first surveywas undertaken among 387 respondents in 5 IDP camps on 3and 4 November 2010, during the period of initial explosiveeruptions and dome growth but before the main eruption on 5November. The second survey was conducted between 8 and27 November 2010 during the waning phases of the eruption(see Surono et al. (2012) for details of the eruption chronology).The second survey included a total of 1582 respondents within28 IDP camps. Our surveys include a total of 1969 respondentsin 33 IDP camps. Six important questions were asked to inter-viewees and addressed: time to evacuate, mode of transportused, type of evacuation (voluntary or forced), return to villageduring evacuation period, time and duration of return to village,and reason for return to village. Two additional questions wereasked during the second survey, which addressed awareness ofthe danger of Merapi after the major explosion and the level oftrust in the government and in the volcanologists.
(3) Survey of road conditionsThis survey consisted of assessing the composition and the qual-ity of roads utilized in evacuations (unpaved, stone, cement, orasphalt). To retrieve road data, we took geographical referencepoints using GPS and we categorized the roads according totype and quality (Fig. 4).
(4) Secondary data collectionEruption monitoring data from September to December 2010were obtained from BPPTK. During the crisis period, monitoringdata and recommendations were released by BPPTK every sixhours.We analyzed the visual and seismic data as well as the con-clusions and recommendations from BPPTK to public and localgovernment with respect to their affect on evacuation policy.For our analysis of displaced person locations and populations, wedefine an IDP camp as the location where IDPs live during the cri-sis. Daily data on IDP numbers, from 26 October until 9 December2010were obtained fromBPBD in Sleman,Magelang, Boyolali andKlaten Districts. In addition, clippings from local and nationalnewspapers concerning the evacuation were collected from thebeginning of the eruption until February 2011.
(5) Questionnaire-based survey at village level: SurdaduIn order to study the behavior of people during the 2010 volcaniccrisis as well as the effects of the eruption on the broader
Fig. 4. Quality of Merapi roads. a, Types of road material and quality: A1, asphalt road withsurface deformation and repaired potholes; A3 asphalt road with severe cracking, surfacecement road; C2, stone road with good surface condition; C3, stone road with poor surface cfrom a stone to asphalt road using GPS-derived coordinates.
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010Research (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
community, we performed another questionnaire-based surveyduring the period from January to August 2011. This enquirywas called Surdadu — SURvei DAmpak DUsun or survey of theeruption's effect at village level. In this questionnaire,we collecteddata related to crisis management: e.g., knowledge and effective-ness of the CVGHM alert system, means of transport, and evacua-tion of livestock. As in the previous surveys,we also inquired if therespondent returned to village during the evacuation period andwhy. The questionnaires were completed by Chiefs of the 543 se-lected villages located in the KRB III, II and I zones.
4. Results
4.1. Volcanic activity and evacuation chronology
The 2010 Merapi eruption forced people living near the volcano totemporarily to leave the danger zones, where access was restricted.The evacuation processes were executed progressively due to the in-creasing danger of eruption and in response to the alert levels (Fig. 5).Accordingly, increasing numbers of evacuation camps were establishedover the course of the crisis (Fig. 6). Five major periods are distin-guished with respect to extension and then contraction of the dangerzones (Fig. 7).
4.1.1. Period A (25 October–2 November 2010)The explosive phreatomagmatic eruption on 26 October produced a
12 km-high ash plume and pyroclastic density currents that extended8 km down the Gendol and Kuning rivers on the southern flank of thevolcano. Information related to preparation for evacuationwas releasedon 24October. The fourth andmaximal level of thewarning systemwasissued on the morning of 25 October and evacuations started on 26 Oc-tober, following the initial explosion. CVGHM confirmed that the 2010eruption beganwith an explosive cratering event; whereas, other erup-tions of the 20th century typically beganwith extrusion of lava. This un-usual beginning, continued escalation in seismicity and deformationand appearance and rapid growth of a new lava raised concerns thatthe eruption could continue and grow to become larger than those ofthe past century (Surono et al., 2012). According to the initial recom-mendation from CVGHM, there were 12 municipalities located in theKRB III with a population of 24,024 people to be evacuated: Sleman Dis-trict including Purwobinangun, Wonokerto, Girikerto, Hargobinangun,Umbulharjo, Kepuharjo, and Glagaharjo; Klaten District includingBalerante, Sidorejo, and Tegalmulyo; and Magelang District includingKemiren and Kaliurang (Fig. 8a). During this period, there were22,599 IDPs on 26 October; and 53,048 on 30 October.
4.1.2. Period B (3–4 November 2010)On 3–4 November, eruptive intensity increased with stronger
degassing and a series of explosions, some of which could be heard inYogyakarta (25 km from the summit). On 3 November at 16:05 (localtime), the CVGHM extended the restricted zone to 15 km along themain rivers (Woro, Gendol, Kuning, Boyong, Krasak, Bebeng, Bedog Riv-ers) and recommended 32 municipalities located in the KRB III with apopulation of 90,325peopleto be evacuated: Sleman District includingGlagaharjo, Kepuharjo, Umbulharjo, Hargobinangun, Pakembinangun,Purwobinangun, Girikerto, and Wonokerto; Magelang District includ-ing Kaliurang, Kemiren, Kamongan, Ngablak, Nglumut, Ngargosoka,Srumbung, Mranggen, Tegalrandu, Keningar, Ngargomulyo, Kalibening,Sumber, Krinjing, Mangunsoko, and Paten; Boyolali District includ-ing Tlogolele, Jrakah, and Klakah; and Klaten District includingTegalmulyo,Sidorejo, Balerante, and Kendalsari (Fig. 8b). As the
no surface deformation, cracking and potholes; A2, asphalt road with light cracking ordeformation and potholes; B2, cement road with no cracking and potholes; B3, mossyondition; D3, unpaved road. b, Illustration showing our method for mapping the change
evacuations atMerapi volcano, Journal of Volcanology and Geothermal
7E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010 evacuations atMerapi volcano, Journal of Volcanology and GeothermalResearch (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
Table 3Timeline for data acquisition and processing.
Before crisis During crisis After crisis
Field data collection Qualitative Interviews with key persons, stakeholders and communities at riskSecondary data: bibliography through local–national newspapers and televisions
Quantitive Road data collection Questionnaire-based surveys (before and after thelargest explosion of 4th–5th November 2010)
Survey of the impacts of eruption atvillage scale (Surdadu)
Secondary data on eruption monitoring andInternally Displaced Persons (IDPs)
Data processing Statistical database (population, refugee, questionnaire-based survey and Surdadu survey)Geo-database in ArcGIS platformDescriptive analysis of secondary data related to crisis management, eruption monitoring data, interviews
8 E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
evacuation progressed, there were 76,031 IDPs on 3 November and82,701 IDPs on 4 November.
4.1.3. Period C (5–13 November 2010)On 5 November 2010 at ~00.05 (local time) the paroxysmal erup-
tion took place. Ash fell in the region west and south of the vent(Damby et al., 2012) and pyroclastic density currents reached 16 kmfrom the summit in the Gendol river (Surono et al., 2012). Sustained ex-plosive eruptions during the night of 4–5 November generated an ashcolumn that rose to 17 km altitude. Several hours before the paroxys-mal eruption, CVGHM extended the danger zone to 20 km from thesummit and called for evacuation of people in this radius (Fig. 8c). Un-fortunately, the list of villages to be evacuated was not included in therecommendation letter and therewas no explanation as towhich sectorwas included in this new radius of 20 km. Consequently, there was amisunderstanding by some of the emergency managers in the rush toevacuate on 5 November and people living in the radius of 20 kmwere confused as to whether to leave their villages or not. Fortunately,many of them, especially in the southern sectors of the volcano, tried tomove away at least 20 km from the volcano. As the evacuationsproceeded the number of IDPs reached 239,618 IDPs on 5 Novemberand 381,696 IDPs on 13 November (Fig. 7). After 8 November, volcanicactivity started to decrease in intensity (Surono et al., 2012), althoughthe number of IDPs remained high and reached a maximum on 14November.
4.1.4. Period D (14–19 November 2010)As the volcanic activity decreased, the CVGHM recommended de-
creasing the radius of the danger zone during this time period as follows:maintaining a radius of 20 km for the sector between the Boyong andGendol rivers (mainly within Sleman District), but reducing the radiusto15 km for Magelang District and to 10 km for Klaten and Boyolali Dis-tricts (Fig. 8d). The pre-eruption population of this restricted area was
Table 4Key informant interviews.
Level Institutions Function Name
District Kesbanglinmas PBSleman (BPBD)
Secretary Asih Kushartanti
KesbangpolinmasMagelang (BPBD)
Secretary Moch Damil
KesbangpolinmasBoyolali (BPBD)
Secretary Usfal Pius-Wawan
KesbangpolinmasKlaten (BPBD)
Secretary Joko Rukminto
Department of PublicWorks
Chief of RoadMaintenance Section
Fauzan
Local Ngargomulyo municipality Head of municipality YatinBronggang Suruh village Head of village Sutaji
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010Research (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
464,328. The total number of registered IDPs decreased to 272,124 by19 November.
4.1.5. Period E (19 November–3 December 2010)After 19 November, local authorities again reduced the danger
zone to a radius of 15 km for the sector between the Boyong andGendol rivers, to 10 km for the rest of Sleman District, to 10 km forMagelang District and Klaten District, and to 5 km for Boyolali District(Fig. 8e). The pre-eruption population of this area was 190,902. Thenumber of registered IDPs dropped to 51,924 on 3 December and toless than 20,000 by 9 December 2010.
4.2. Number and distribution of IDPs
Before the major explosion on 4–5 November (periods A and B),there were approximately twenty IDP camps located less than 20 kmfrom the summit (Fig. 6a). Those camps were only prepared to accom-modate people living in KRB III as described in the contingency planmade in 2009. After 3 November, at least two IDP camps in Sleman dis-trict were closed and IDPs were transferred to other camps.
When the restricted zonewas extendedon4–5November, therewereno refugee camps prepared beyond 20 km to accommodate the IDPs.Therefore, most of the IDP camps utilized after the main explosion werepublic buildings (schools, hospitals, stadiums, village halls, and universi-ties) or even residents' houses or yards. On 13 November there were atleast 600 IDP camps registered by BNPB. These camps were scatteredaround Sleman, Magelang, Boyolali, Klaten and several other Districts inCentral Java Province and Yogyakarta Special Region Province (Figs. 1and 6).
The number of IDPs increased exponentiallywith the 4–5Novemberextension of the danger zone (Fig. 7a). Despite the evacuation order toclear the 20-km radius, the highest number of IDPs was only finallyreached on 14 November (399,403 IDPs) due to insufficient prepared-ness in handling the evacuation and IDPs' movement from one campto another. After 13 November, the number of IDPs then decreased sig-nificantly, i.e. in Boyolali District, in 13 November there were 51,756IDPs and only 29,780 IDPs two days later (Fig. 7b). And when the localauthorities decreased the radius of the restricted zone in this sector to10 km, people from villages that were still habitable and located lessthan 10 km from the summit also went home for good.
During Period D, the number of IDPs in the Sleman District de-creased slightly at first and then dramatically on 19 November,when the local authorities reduced the radius of the restricted zonefrom 20 km to10 km. At that time, more than 40,000 IDPs in theSleman district returned to their villages (Fig. 7b).
4.3. Community responses to evacuation
Public reactions to evacuation were analyzed through question-naireswith IDPs and in-depth interviewswith the headofNgargomulyo
evacuations atMerapi volcano, Journal of Volcanology and Geothermal
Fig. 5. Time line showing changes in alert levels, evacuation orders and restricted zones during the 2010 Merapi eruption. Periods are described in Section 4 of the text.
9E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
municipality and the head of Bronggang-Suruh village (one of theworst-affected villages during the 2010 Merapi eruption, located15 km southeast from Merapi, along the Gendol river; Fig. 2).
4.3.1. Preparations to deal with volcanic crisisCommunity education on volcanic disasters before the 2010 eruption
was conducted collectively by the members of BPPTK, local authorities,academic institutions and NGOs. Many of these activities took place be-tween April and August of 2010 and were described by Asih Kushartanti(Secretary of Badan kesatuan bangsa dan perlindungan masyarakat‘Kesbanglinmas’ —public protection andnatural disaster office of SlemanDistrict on 11 May 2010). As described in our interview with Asih, com-munity educationwas carried out in order to prepare the communities toface the next Merapi volcanic crisis. In addition, during 30 May–1 June2010, a series of volcanic crisis preparation and evacuation drills werecarried out in the Sleman District. These drills were led by the Police,the National Army, the local government agencies, and some militarypersonnel from the United States of America. More than 2000 peopleliving in Turi, PakemandCangkringan sub-districtswithinwere involvedin the evacuation drills (Fig. 1). However, these activities were onlyconducted with communities from the villages located in KRB III. Asrevealed by Sutaji (head of Bronggang-Suruh village — interviewed in
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010 eResearch (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
December 2011): “we did not know much about community educationon volcanic disaster risk before the 2010Merapi volcano eruption. How-ever,we knew that our village is vulnerable to lahar but not to pyroclasticflows.” These comments were typical of interviews with survivors of theeruption; those people who were involved in community education inthe year before the 2010 eruption were much better prepared thanthose who had not been recently involved in drills or other communityeducation programs.
Another preparation that aided the evacuations was related toroad quality improvements. The conditions of provincial and districtroads near Merapi are relatively good. Local roads linking each mu-nicipality are mostly asphalt and the roads in the southwestern partof the volcano are in the best condition (Fig. 9). Prior to the 2010eruption, local people worked collectively to upgrade the conditionof evacuation roads. This action was partially funded by the Depart-ment of Public Works at the district level by providing material forroad pavement (information from interview with Fauzan, chief ofRoad Maintenance Section, Department of Public Works, Sleman Dis-trict, on 10 May 2010). The maintenance of evacuation roads becamea major issue for post-crisis management following the 2006 Merapieruption (Mei and Lavigne, 2012). In response, local governmentsworked to enhance the quality of roads by paving them with asphalt.
vacuations atMerapi volcano, Journal of Volcanology and Geothermal
Fig. 6.Maps showing locations of IDP campsduring the 2010Merapi eruption. Individual panels show the status of IDP camps at the specifieddates. The relative population of the camps isindicated by the size of the circles as indexed in the legend. Danger zones (KRB I, II, and III) from Hadisantono et al. (2002) are shown in the background, along with 10, 15 and 20 kmradius circles.
10 E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010 evacuations atMerapi volcano, Journal of Volcanology and GeothermalResearch (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
Fig. 7. Graphs showing population of IDP camps over the course of the 2010 Merapi eruption. a, Total number of IDPs. b, Number of IDPs per district. Levels III and IV refer to alertlevels. Periods are defined in Section 4 of the text.
11E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
However, after several years, the condition of the roads worsened dueto trucks hauling volcanic material (De Bélizal et al., 2011b). Efforts tomaintain good road condition is inconsistent around the volcano. Forexample in Ngargomulyo, local authorities along with the communitybanned sand mining in their area because they were worried aboutthe consequences of the increasing truck traffic on the road (interviewwith Yatin, head of Ngargomulyo municipality on 28 July 2010). In con-trast, people and local government in the Klaten District gave less atten-tion to the effect of the truck traffic, because sandmining is seen as a keysector of their economy (Mei and Lavigne, 2012).
4.3.2. Evacuation alertAccording to the results of our questionnaires of IDPs (Fig. 10), most
people received evacuation alerts directly from the head of village(54%). Other means of communication included: sirens (16%), neigh-bors (11%) or indirectly using telephones, kentongan (wood gongs)and two-way radios (walkie-talkie) (19%). Kentongan are traditionallyused for warning the public (Sumiyati, 2007), notably in rural areas orduring an emergency period where electricity is not fully adequate.Every kentongan code has its ownmeaning. In case of natural (volcanic)disaster, the kentongan is beaten repeatedly and continuously with thesame tone (gobyog or titir tone in Javanese). It indicates that peopleshould immediately evacuate to a pre-determined location, which isusually a village hall. However, not all the villages in Merapi have oruse a kentongan during volcanic crisis. This was the case in Sawanganand Selo sub-districts where the kentongan is traditionally forbidden
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010 eResearch (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
during Merapi's eruptions (Fig. 11). As mobile phones are common inrural areas (Fig. 11), people also used this tool to inform others duringthe crisis. The role of local organizations was also important in crisiscommunication, as exemplified by the actions of JalinMerapi, a local or-ganization supported by several NGOs working around the volcano.This association was established in 2006 by three community-basedradio stations in Klaten, Boyolali and Magelang Districts. During theemergency response period in 2010, JalinMerapi used various electronicmedia to quickly and accurately convey important information and datato support the decision making process. JalinMerapi could be accessedthrough a website, social networks such as Twitter and Facebook,SMS, radio communications, telephone and through information postedin the field. JalinMerapi was managed by a voluntary network that op-erated 24 h a day during the crisis.
4.3.3. Public responses to evacuation orderBased on the questionnaires distributed before the main explosion,
81% of the respondents (Fig. 10) evacuated (without force) when thefirst pyroclastic flows were reported on 26 October at 05:30 PM (localtime), despite the evacuation order given earlier on 25 October at11:00 A.M (local time). This evacuationmay be classified as “evacuationwhen hazard is imminent” a phenomena that is common at Merapi. Asin previous eruptions, peoplewait to evacuate until a pyroclastic floworash fall event has taken place, even though the government had previ-ously informed them that the volcano was already at its highest warn-ing level. On 25 October local authorities had conducted an organized
vacuations atMerapi volcano, Journal of Volcanology and Geothermal
Fig. 8. Maps showing changes in hazard and restricted zones over the course of the 2010 Merapi eruption. a, Period A, 25 October–2 November. CVGHM recommended evacuation oftwelve municipalities located in KRB III (Purwobinangun, Wonokerto, Girikerto, Hargobinangun, Umbulharjo, Kepuharjo, Glagaharjo, Balerante, Sidorejo, Tegalmulyo, Kemiren andKaliurang (shown as white polygons)). b, Period B, 3–4 November. CVGHM recommended evacuation of thirty two municipalities in the KRB III (shown as white polygons). c, PeriodC, 5–13November. Several hours before the paroxysmal eruption, CVGHM extended the restricted zone to 20 km from the summit and recommended evacuation of people in this radius.However, a list of villages to be evacuatedwas not included in the recommendation letter. This caused amisunderstanding by some of the emergencymanagers in the rush to evacuate on5November. d, PeriodD, 14–19November. Radius of 20 km for the sector between the Boyong andGendol rivers (mainlywithin SlemanDistrict)wasmaintained. The radiuswas reducedto 15 km for Magelang District and 10 km for Klaten District and Boyolali District. e, Period E, 19 November–3 December. The radius was reduced to15 km for the sector between theBoyong and Gendol rivers, 10 km for the rest of Sleman District, 10 km for Magelang District and Klaten District and 5 km for Boyolali District (see text for details).
12 E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
evacuation in order to evacuate persons with special access and func-tional needs, the elderly, children and pregnant women. The evacua-tions were particularly intended for the villagers living in KRB III.Another type of evacuation was undertaken by the Army and Policesearch and rescue teams, who sought out and evacuated the dead andinjured from the 26 October eruption, and to locate people missing orwho had remained in villages that were affected by the eruption.
Despite evacuation orders and efforts of local authorities to evacuatepeople, some residents were reluctant to leave their villages. For exam-ple and as previously noted,MbahMarijan (the “gatekeeper ofMerapi”)insisted on staying in his house in Kinahrejo (a village located 7 kmsoutheast fromMerapi's summit). On26October,Marijan and34 othersvictims (some of whom were trying to persuade him to evacuate) diedwhen they were hit by the first pyroclastic flows. Another examplecame from theNgargomulyomunicipalitywhere at leastfive people de-cided not to leave their houses during the eruption periods. In this case,village officials had to return to this village every three days to ensurethat they were safe and had adequate food and water.
During the paroxysmal eruption on the night of 4–5November, evac-uations were taking place spontaneously. In some cases, people did notstart to evacuate from the new 20-km radius restricted zone until11.30 PM.Without detailed instructions, people tried to escape by them-selves, but many did not know exactly where to go. “When we had to
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010Research (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
evacuate, most residents did not know where they had to go to exceptto go away from Merapi” (Sutaji, head of Bronggang-Suruh village,interviewed in December 2011).
Due to an underestimation of the potential for larger-than-normalevacuation areas in the2009 contingency plan, no IDP camp had beenprepared outside the 20 km radius from the summit and all the IDPcamps were located inside the restricted zone (Anonymous, 2009a,b,c).Therefore, on the night of 5 November and in the aftermath of themain explosion, the local authorities of the Sleman district began prepa-ration of a new IDP camp in Maguwoharjo football stadium (located23 km from Merapi). This site then became the largest official IDPcamp in the Sleman District and accommodated more than 21,000IDPs. Due to the limited space and the fear of a bigger eruption in the fu-ture, IDPs also moved to other camps, notably community-based camps,or they stayed with family relatives. Because of disorganization and therapid pace at this stage of the evacuation, local authorities faced logisticaldifficulties, such as recording IDPs and distributing aid and assistance.
4.3.4. Capacity of local transport for evacuation purposesTrucks (39%) and motorbikes (37%) were the most used means of
transport for evacuation before the 4–5 November main eruption(Fig. 10). Motorbikes played an important role during the 2010 evacua-tion as each household inMerapi has at least onemotorcycle. However,
evacuations atMerapi volcano, Journal of Volcanology and Geothermal
Fig. 9. Map showing road conditions in the Merapi region. Map on left shows road quality. Map on right show road material.
13E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
those without access to a personal vehicle had to rely on transportationsupplied by others, including emergency officials or local authoritieswho ordered commercial trucks or cars.Many also had to rely on neigh-bors' vehicles to facilitate evacuation, as was the case in Ngargomulyo.Trucks and cars were mostly intended for individuals with specialaccess and functional needs, frail elderly people and those with limitedincomes. “We asked residents who had cars or trucks to carry thosewith limited access to transport. The municipality then gave supportby paying the fuel cost” (Yatin, head of Ngargomulyo municipality, 23January 2011).
4.3.5. Evacuation of livestock: a key problem in evacuation managementFor people living on the slopes of Merapi, livestock (mainly cows)
represent their main livelihood, mostly through milk production andsale. Cattle husbandry in the Merapi region utilizes stall-fed cows,which depends on the exploitation of grasslands located up-slope(Dove, 2008). Before the 2010 Merapi eruption, there were 4858dairy cattle in the district of Sleman, which is the regional center ofmilk production (BPS, 2009). This business was conducted by atleast 1605 farming households with an average daily milk productionof as much as14.2 tons. Because livestock plays such a prominent rolein peoples' lives at Merapi, residents were ready to face personal dan-ger to continue feeding their animals.
During the eruption, livestock wasmostly left behind in the villages,but some animals were evacuated either to IDP camps or to relatives'houses. In some cases, cattle were evacuated several times as a resultof the changes in restricted area. Individuals abandoned their livestockin cases inwhich therewere toomany cattle to transport, no evacuationsystem was available, or due to uncertainty concerning the cost fortransport and care of the animals at the camps. Following the ash fallevents, it was difficult to find clean grass and fresh water for the cattlethat survived. Therefore, cows were fed with ash-covered fodder,which reduced milk production and resulted in weight loss. Conse-quently, some farmers opted to buy grass from other regions and haveit transported it to their villages at high cost. Those who were not ableto buy grass for their cattle, opted to sell at low prices. For examplethe regular price for a female cow in the region was around seven toten million IDR (Indonesian Rupiah) (700 to 1000 US$) before the
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010 eResearch (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
eruption. During the eruption period the price fell to three to five mil-lion IDR (300 to 500 US$).
5. Discussion and recommendations
5.1. Evacuation during the 2010Merapi eruption: a successful evacuation?
Evacuation is a complex process. The success or failure of an evac-uation is often determined by the ratio between rescued people andfatalities. Thirty-five people were killed during the first explosiondue to evacuation refusal despite the order given by the local author-ities. During the night of 4–5 November, about two hundred peopledied from the impacts of PDCs. Most of the victims lived in the arealocated outside the initial KRB III limit (Fig. 2), e.g., Bronggang village(where 49 people died from the eruption) is located 15 km from thevent. The high death toll was caused by a lack of readiness to evacu-ate. Many of the victims were found dead in front of their houseswith bags full of clothes or near their motorcycles or cars. However,the total of 367 fatalities is relatively small compared to the numberof people who would have died without the evacuations (10,000 to20,000 people; Surono et al., 2012). From local authorities and volca-nologist' points of view, the progressive enlargement of the restrictedzone during the eruption was carried out as soon as possible given therapidly changing situation (Surono et al., 2012). Although any loss oflife is tragic, the relatively low number of victims compared to thedisaster's magnitude and potential much larger number of fatalitiesthat would have occurred without the warnings and evacuationsqualifies the 2010 Merapi response as successful.
However, some difficulties were identified during the evacuation pe-riod, such as the repeated displacements of IDPs and conditions of IDPsin the camps. The organization of IDPs worked smoothly in the early pe-riod of displacement (25 October to 3 November). Local authorities hadprepared well for the evacuation using the 2009 Merapi contingencyplan in each district. But, when the volcanic activity increased, the man-agement of IDPs faced several problems in adapting to the rapid enlarge-ment of the restricted zone. Peoplewere confusedwhen IDP campshad tobemoved. During the crisis, IDPsweremoved several times from one IDPcamp to another in response to government changes in the areas of the
vacuations atMerapi volcano, Journal of Volcanology and Geothermal
Fig. 10. Bar graphs summarizing results of replies to questionnaires by IDPs. In addition to the categories listed in the graph showing “Transmission of evacuation order” we knowqualitatively that many IDPs received word of the evacuation by way of social media and through the program of JalinMerapi.
14 E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
restricted zones. Movements of IDPs generated problems both for thelocal authorities and for the IDPs themselves. The former could not pro-vide adequate camps in such a limited time and with inadequate re-sources, and the latter felt stressed and frustrated. The 2009 Merapicontingency plan failed mainly due to the scale and rapid onset of the2010 eruption.
5.2. Community reactions on evacuation: a reflection of disasterpreparedness
From the evacuations during the 2010 Merapi eruption, we candistinguish two types of local communities' reactions:
(1) Those who were already prepared to face the crisis.People who were already prepared to face the crisis were domi-nantly from the high-danger zones KRB III and II, where programs
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010Research (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
of disaster risk reduction had been previously carried out eitherby the local governments or NGOs. Due to their experience of vol-canic crisis, most of the inhabitants located in KRB III were awareof the consequences of volcanic eruptions. Therefore, when theevacuation order was given by local authorities more of the peo-ple in these zones obeyed this order, even thoughmost of the peo-ple delayed their decision to evacuate until the first pyroclasticflows occurred. Several efforts were conducted collectively beforethe 2010 eruption by the local government and communities inupgrading road conditions to facilitate a successful evacuation.In KRB III, the dissemination of information about volcanic activityand evacuation orders was already well organized before the2010 eruption: themunicipality chief served as the key informantand communicator for alerting the local population; the villagechiefs and local organizations were intermediaries for the com-munities. People were informed where to evacuate, they knew
evacuations atMerapi volcano, Journal of Volcanology and Geothermal
Fig. 11. Distribution of sirens and kentongan-wood gongs (map on left), mobile phones and two-way radios (map on right).
15E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
the rallying points and the IDP camps had already been designat-ed for each village.However, not all the people whowere prepared responded to thedisaster as ordered by the local authorities, e.g.,MbahMarijan andhis followers' evacuation refusal despite several disaster reductionprograms conducted by BPPTK and NGOs in the village whereMarijan lived. The refusal to leave the danger zone has also beenobserved elsewhere. During the 1980 eruption of Mount St.Helens, a resident named Harry R. Truman refused to leave hishome in the danger zone due to his attachment to the volcanoand his belief that the volcano would not destroy his village(Tilling et al., 1990). Evacuation refusals at Merapi volcano andMount St. Helens show that even though local communitieswere prepared to face the eruption, in reality, not all membersof the communities at risk are prepared to evacuate and given achoice, some individuals will not leave. Unfortunately, in such sit-uations, these individuals may put many others at risk.
(2) Those who were not adequately prepared to face the crisis.A second type of community reaction is demonstrated by thosewho were not well prepared to face volcanic crisis. In the caseof Merapi, the scenario used for the emergency planning onlyconsidered the hazard that could be generated by “typical”dome-collapse PDCs. As the contingency plan was created in2009 and was based only on threats posed by “typical” Merapieruptions as shown on a hazard map (Hadisantono et al., 2002)that was updated after the 2006 eruption, several areas affectedby the 2010 eruption were not in the designated KRB III dangerzone, i.e. Bronggang-Suruh village. Consequently, because of alack of preparation to evacuate and lack of education about disas-ter risk reduction (which was focused on villages in the KRB IIIzone) many people in this village became victims of the 4–5 No-vember PDCs. For example, during evacuation several inhabitantstook the wrong evacuation route, a path parallel to the Gendolvalley's direction instead of taking the perpendicular path to getaway from the river.
During the evacuation of 4–5 November, people who lived in thenew restricted zone between 15 and 20 km from the summit, were
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010 eResearch (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
confused and in a hurry. This condition was a result of the limitedtime to prepare their belongings and to gather other family members.Most villagers did not know where to go to evacuate, they only un-derstood that they had to move away at least 20 km from the volcano.The lack of preparation and coordination resulted in many familiesbeing split-up during the rapid evacuation. The subsequent slownessin the government agencies to register IDP's and the migration of IDPsfrom one camp to another made finding missing people difficult. Themovements of IDPs were also affected by the insufficient space in thecamps, inequality in logistics and distributions of assistance, and theaspiration of IDPs to gather with their family or neighbors. Insuffi-cient space in the camps resulted in many IDPs moving in with familyor relatives.
Another major issue related to evacuations in Merapi was the re-turn to one's village within the restricted zones during the evacuationperiod. This action was done either explicitly with the help of localauthorities, or illegally. For example, people from Bawukan IDPcamp, located 14 km from the summit, went back every morning totheir village (Kalitengah Lor) located less than 6 km from the summitduring the hazardous period preceding the 4–5 November explosion(Fig. 2). This daily migration was made by truck and was coordinatedby local authorities for various purposes, i.e. feeding or milking thelivestock, taking a shower, or just checking the condition of theirhouses. Such behavior was exceptionally dangerous because of the in-ability of scientists to predict exact eruption times. Such behavior isnot unique to Merapi; a return to the most dangerous area also tookplace during the 2007 eruption of Kelud volcano, East Java (DeBélizal et al., 2011a).
In order to avoid such problems, our analysis of the 2010 evacuationshows that there is an acute need to educate more of the vulnerablepeople in the Merapi region about volcanic disaster preparedness. Toavoid future losses, communities must have a clear understanding andknowledge of volcanic hazards and the potential for disaster. CurrentlyatMerapi, some villagers assume that the 2010 disaster was a divine actthat could not be overcome. This perception is common, even in mod-ern societies that are responsive to logical and social explanations(Chester, 2005). Therefore, local beliefs must be carefully consideredin natural disaster management, e.g., by creating a “parallel practice”
vacuations atMerapi volcano, Journal of Volcanology and Geothermal
16 E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
in which actions in response to the “miraculous event” should be ac-companied by “rational actions” such as evacuation (Chester et al.,2008).
In addition, many inhabitants assume that evacuation is a scaryword; they fear that they cannot return to their homes, and that theymight lose their livestock and belongings. Our analysis of the responsesfrom questionnaires and interviews indicates that an effective strategyto address this concern would be for local governments and local orga-nizations in the Merapi region to work together to make people awarethat evacuation is part of disaster risk reduction and that evacuationsminimize loss of life and possessions. One solution offered by the headof Ngargomulyo municipality is the creation of “sister-municipalities”in which a village located in the hazard zone is paired with a villagelocated in the “safer” area in the same district for emergency planning.In this way, a shared sense of purpose could result in more effectiveevacuationof people and livestock from the village located in thehazardzone to the “safer” village during the period of evacuation.
5.3. Evacuation management: an integral part of disaster management
Our studies confirm that evacuation must take into account notonly the actions required to move people from a threatened area toa safer area, but also how to ensure their livelihoods during the evac-uation period. Furthermore, our analysis of the Merapi 2010 evacua-tion shows clearly that evacuation management needs to consideractions and responsibilities before, during, and after the crisis.
Actions before the crisis include developing a detailed evacuationplan and effectively communicating the plan to all of the potentially af-fected communities aswell as preparation of infrastructure and logistics(e.g., road improvements; IDP camps, arrangements made for food,water, emergency supplies, etc.). Evacuation plans need to be compre-hensive and must take into account: 1) who will be evacuated,2) when and how the evacuation alert will be announced, 3) wherethe evacuees will meet and be housed (e.g., meeting points, destinationpoints or IDP camps), 4) how the evacuation will be undertaken(e.g., what evacuation vehicles will be used, which roads will be usedas evacuation routes, what items evacuees can take with them (includ-ing plans for livestock), and what modes and capacities of transporta-tion are available), and 5) how long an evacuation is anticipated. InIndonesia, these actions can only be accomplished through a collabora-tive effort of: 1) BPBD as crisis management coordinator, 2) CVGHMwho gives the recommendations of evacuation order and restrictedzone, 3) Army, Police and Department of Transportation who provideofficial transport and organize the evacuation, 4) Social Department,Department of Public Works and Department of Health who provideIDP camps, arrangements made for food, water, emergency supplies,5) municipalities and villages who provide local transport and organizeevacuations at local scales, and 6) non-governmental organizations(e.g., Red Cross, Tagana, Boy scouts) and volunteers who facilitate theevacuation process and organize IDPs in the camps (Fig. 3).
In the case of the 2010 Merapi eruption, local authorities used the2009 hazard zonemap as a base tool for the evacuation plan until 3 No-vember. Afterwards, the local authorities used new evacuation plansthat had to change rapidly with the changing radius of the dangerzones. This type of evacuation plan is common in volcanic emergencyplanning, e.g., examples include the operation plan for Popocatépetlvolcano in Mexico (Sedena, 2011) and the contingency plan for theAuckland Volcanic Field in New Zealand (Ferner, 2002). In practice, anevacuation plan based on radius from the summit allows emergencymanagers to utilizemultiple escape roads and take into account admin-istrative boundaries in their plans. However, for such an evacuationplan to be effective, multiple scenarios have to be properly defined ascontingencies. This was not the case in the 2009 Merapi evacuationplan; it used only one eruptive scenario based on the relatively smalleruptions and limited extents of PDCs of the past few decades. It didnot take into account the possibility ofmuch larger “100-year”eruptions
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010Research (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
(Surono et al., 2012) and the resulting greater extent of PDCs and tephrafall. Our analysis indicates that a detailed hazard zonation map at asub-district or municipality scale that takes into account both smalland large magnitude eruptions is needed. Such a contingency planwould help the emergencymanagers, the public officials and the publicrespond to future eruption hazards and to better understand the prob-able location and distribution of areas that will be affected by differenttypes of volcanic hazards.
At Merapi, evacuations are organized at the municipality level: mu-nicipal leadersmobilize their entire villages to evacuatewith the help oflocal organizations. A registry of IDPs and good organization of the IDPcamps is required for effective distribution of supplies and to helpre-united family members whomay have become separated.When ex-ecuted properly, this system of community evacuation minimizes thepossibility of people being left behind or being separated from theirfamilies and neighbors and it facilitates the registration of evacueesand organization of logistics for the camps. However, if the systembreaks down, as did on the night of 4–5 November when evacuationswere not planned and had to be executed suddenly, many people be-came separated from their families and neighbors. The separation offamily members contributed to people not staying in their designatedcamps and along with other factors (such as taking care of livestock)to returning to the danger zone. Our analysis of interviews and ques-tionnaires also indicates that boredom while waiting to return homeand a desire to go back to work as soon as possible to earn money forone's family were major psychological issues experienced by IDPs. Toaddress the danger posed by people returning to the danger zone andthe issue of boredom and lack of income during the crisis we suggest:1) that future emergency plans require that evacuated people not re-turn to the danger zone without official authorization, and 2) thatgroup activities for IDPs in the camps be included in the evacuationplan. Activities that provide some income, such as training in food prep-aration or brick making are desirable.
Finally, the “after crisis” phase of the evacuation comprises themaintenance of evacuation infrastructures, disaster risk educationand preparation for the next crisis. Evacuation roads, signs, and IDPcamps can be maintained collectively by the local authorities or com-munities, as illustrated by the Ngargomulyo municipality. To preventfuture loss of life during unusually large eruptions, disaster risk edu-cation and preparation are needed not only for inhabitants living inKRB III, but for all Merapi danger zones.
6. Summary and conclusion
The experience of evacuation during the 2010 Merapi eruptionshowed that: 1) the evacuation went smoothly during the first days oferuption even though some aspects of the eruption (e.g., explosiveonset)were not in accordancewith the scenario of the 2009 contingencyplan. However, when the eruption became much larger than expectedon 4 November, the evacuation process faced difficulties due to thelack of preparation for such a large event by the government and inhab-itants. From the perspective of the number of lives saved, the evacuationduring the 2010 Merapi eruption is seen as a success (Surono et al.,2012). Yet, several problems in evacuation management are evident,e.g., separation of family members during evacuation, return to villageof evacuees during the danger period, lack of adequate IDP camps and re-lated activities to accommodate the larger-than-expected number ofIDPs and duration of the evacuation period.
Several steps can be taken in the future to improve future evacu-ations, such as 1) preparation of a new comprehensive contingencyplan with multiple eruption scenarios and translation of the planinto emergency evacuation actions that are to be taken in each mu-nicipality, so that local leaders and people can be prepared collec-tively for future crises. Contingency plans require frequent review,revision and updates in order to reflect changes in the volcanic situ-ation at Merapi. 2) Educating people about volcanic disaster risk
evacuations atMerapi volcano, Journal of Volcanology and Geothermal
17E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
reduction and awareness not just for inhabitants of KRB III, but alsofor people in all volcanic hazard prone areas.
The 2010 Merapi volcano eruption provided another example ofwhat we call a successful evacuation. However, rating an evacuationas successful is not only based on the number of lives saved, but alsoon theway people behave and their vulnerability during the evacuationperiod, and our study has identified a number of areas for improvement.
We encourage further comprehensive studies of communities' per-ceptions on evacuation, evacuation impacts on population and evacua-tion modeling in order to improve our understanding of how volcanicevacuation management can be best practiced in the future.
Acknowledgment
This studywas held in the framework of theMitigate andAssess Riskfrom Volcanic Impact on Terrain and Human Activities MIAVITA underthe Work Package 5 concerning “Socio-economic Vulnerability and Re-silience.” TheMIAVITA project is financed by the European Commissionunder the 7th Framework Programme for Research and TechnologicalDevelopment, Area “Environment,” Activity 6.1 “Climate Change, Pollu-tion and Risks.” The article reflects the authors' views. The EuropeanCommission is not liable for any use that may be made of the informa-tion contained therein. The authors wish to acknowledge the FrenchEmbassy in Jakarta and the Directorate General of Higher Education,Ministry of National Education Republic of Indonesia. We thankDr. Surono (Director of the Center for Volcanology and GeologicalHazard Mitigation, Indonesia); Dr. Subandriyo, head of the BPPTKand his staff, especially Ibu Dewi S. Sayudi, Bapak M. Muzani andIbu Sri Sumarti; local authorities at district level (Sleman, Klaten,Magelang, and Boyolali); the Disaster Research Center and the Facultyof Geography Gadjah Mada University, Indonesia; and particularly thepeople from Bronggang-Suruh village and Ngargomulyo municipality.The authors especially thank Dr. John S. Pallister (USGS) for his remark-able constructive comments and his incredible amount of work on thisarticle.We also thank Dr. Philippe Jousset and all the reviewers for theircareful reviews of the manuscript.
References
Abdurachman, E.K., Bourdier, J.L., Voight, B., 2000. Nuées ardentes of 22 November1994 at Merapi volcano, Java, Indonesia. Journal of Volcanology and GeothermalResearch 100, 345–361.
Aizawa, K., Ogawa, Y., Mishina, M., Takahashi, K., Nagaoka, S., Takagi, N., 2009. Struc-tural controls on the 1998 volcanic unrest at Iwate volcano: relationship betweena shallow electrically resistive body and the possible ascent route of magmaticfluid. Journal of Volcanology and Geothermal Research 187, 131–139.
Andreastuti, S.D., Alloway, B.V., Smith, I.E.M., 2000. A detailed tephrostratigraphicframework at Merapi Volcano, Central Java, Indonesia: implications for eruptionpredictions and hazard assessment. Journal of Volcanology and Geothermal Re-search 100, 51–67.
Anonymous, 2009b. Rencana Kontingensi Bencana Erupsi Gunung Merapi. PemerintahKabupaten Magelang, Magelang (153 pp.).
Anonymous, 2009c. Rencana Kontingensi Erupsi Gunung Merapi. Pemerintah KabupatenKlaten, Klaten (94 pp.).
Blong, R.J., 2000. Volcanic hazard and risk management. In: Sigurdsson, H., Houghton, B.,Rymer, H., Stix, J., McNutt, S. (Eds.), Encyclopedia of Volcanoes. Academic Press, SanFrancisco, pp. 1215–1228.
BPS, 2008. Kabupaten dalam Angka. Biro Pusat Statistik, Yogyakarta Special District,Indonesia.
BPS, 2009. Kabupaten dalam Angka. Biro Pusat Statistik, Yogyakarta Special District,Indonesia.
BPS, 2010. Kabupaten dalam Angka. Biro Pusat Statistik, Yogyakarta Special District,Indonesia.
Camus, G., Gourgaud, A., Mossand-Berthommier, P., Vincent, P., 2000. Merapi (CentralJava, Indonesia): an outline of the structural and magmatological evolution, with aspecial emphasis to the major pyroclastic events. Journal of Volcanology and Geo-thermal Research 100, 139–163.
Carlino, S., Somma, R., Mayberry, G.C., 2008. Volcanic risk perception of young peoplein the urban areas of Vesuvius: comparisons with other volcanic areas and implica-tions for emergency managements. Journal of Volcanology and Geothermal Re-search 172, 229–243.
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010 eResearch (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
Charbonnier, S.J., Gertisser, R., 2008. Field observations and surface characteristics ofpristine block-and-ash flow deposits from the 2006 eruption of Merapi Volcano,Java, Indonesia. Journal of Volcanology and Geothermal Research 177, 971–982.
Charbonnier, S.J., Gertisser, R., 2012. Evaluation of geophysical mass flow models usingthe 2006 block-and-ash flows of Merapi Volcano, Java, Indonesia: towards a short-term hazard assessment tool. Journal of Volcanology and Geothermal Research231–232, 87–108.
Chester, D.K., 2005. Theology and disaster studies: the need for dialogue. Journal ofVolcanology and Geothermal Research 146, 319–328.
Chester, D.K., Duncan, A.M., Dibben, C.J.L., 2008. The importance of religion in shapingvolcanic risk perception in Italy, with special reference to Vesuvius and Etna. Jour-nal of Volcanology and Geothermal Research 172, 216–228.
Damby, D.E., Horwell, C., Baxter, P., Delmelle, P., Donaldson, K., Dunster, C., Fubini, B.,Murphy, F.A., Nattrass, C., Sweeney, S., Tetley, T.D., Tomatis, M., 2012. The respira-tory health hazard of tephra from the 2010 centennial eruption of Merapi with im-plications for occupational mining of deposits. Journal of Volcanology andGeothermal Research. http://dx.doi.org/10.1016/j.jvolgeores.2012.09.001.
De Bélizal, E., Lavigne, F., Grancher, D., 2011a. Quand l'aléa devient la ressource: l'activitéd'extraction des matériaux volcaniques autout du volcan Merapi (Indonésie) dans lacompréhension des risques locaux [When the hazard becomes the resource:block andsandmining aroundMerapi volcano (Indonesia). Cybergeo: European Journal of Geogra-phy, Environment, Nature, Paysage 525. http://dx.doi.org/10.4000/cybergeo.23555.
De Bélizal, Édouard, Lavigne, F., Gaillard, J.C., Grancher, D., Pratomo, Indyo, Komorowski, J.-C.,2011b. The 2007 eruption of Kelut volcano (East Java, Indonesia): phenomenology, crisismanagement and social response. Geomorphology 136, 165–175. http://dx.doi.org/10.1016/j.geomorph.2011.06.015.
De la Cruz-Reyna, S., Tilling, R.I., 2008. Scientific and public responses to the ongoing vol-canic crisis at Popocatépetl Volcano, Mexico: importance of an effective hazards-warning system. Journal of Volcanology and Geothermal Research 170, 121–134.
Dove, M.R., 2008. Perception of volcanic eruption as agent of change on Merapi volca-no, Central Java. Journal of Volcanology and Geothermal Research 172, 329–337.
Ferner, B.C.H., 2002. Contingency Plan for the Auckland Volcanic Field. Auckland Re-gional Council, Auckland.
Fielding, E., 2002. BBC: Volcano hell. www.bbc.co.uk. Online: January 17, 2002. Last accessed:July 09, 2012. http://www.bbc.co.uk/science/horizon/2001/volcanohelltrans.shtml.
Fowler Jr., Floyd J., Cosenza, C., 2008.Writing effective questions. In: deLeeuw, E.D., Hox, JoopJ., Dillman, Don A. (Eds.), International Handbook of Survey Methodology European As-sociation of Methodology series. Lawrence Erlbaum Associates, Inc., pp. 136–160.
Frank, D.G., Meier, M.F., Swanson, D.A., 1977. Assessment of increased thermal activity atMount Baker Washington March 1975–March 1976: a summary of geothermally in-duced changes, progress report on geophysical and geochemical monitoring and analy-sis of hazards atMount Baker. U.S. Geological Survey Professional Paper, 1022A (40 pp.).
Gaillard, J.C., Le Mason, Virginie, 2007. Traditional societies' response to volcanic haz-ard in Philippines. Mountain Research and Development 27, 313–317. http://dx.doi.org/10.1659/mrd.0949.
Gaudru, Henry, 2005. Potential Impacts of Eruptions on Volcanic Islands: Global Ap-proach for Volcanic Risk Mitigation. SVEUROP, Genève.
Gavilanes-Ruiz, J.C., Cuevas-Muñiz, A., Varley, N., Gwynne, G., Stevenson, J., Saucedo-Girón, R., Pérez-Pérez, A., Aboukhalil, M., Cortés-Cortés, A., 2009. Exploring the fac-tors that influence the perception of risk: the case of Volcán de Colima, Mexico.Journal of Volcanology and Geothermal Research 186, 238–252. http://dx.doi.org/10.1016/j.jvolgeores.2008.12.021.
Green, M., Perry, R., Lindell, M., 1981. The March 1980 eruptions of Mount St. Helens: cit-izen perceptions of volcano threat. Disasters 5, 49–66. http://dx.doi.org/10.1111/j.1467-7717.1981.tb01129.x.
Hadisantono, R.D., Andreastuti, S., Abdurachman, E., Sayudi, D., Nurnusanto, I., Martono, A.,2002. Volcanic hazardmap ofMerapi volcano, Central Java and Yogyakarta special prov-ince. Scale 1:50,000. Center for volcanology and geological hazard mitigation, Bandung.
Hall, M.L., 1990. Chronology of the principal scientific and governmental actions lead-ing up to the November 13, 1985 eruption of Nevado del Ruiz, Colombia. Journal ofVolcanology and Geothermal Research 42, 101–115.
Hartmann, M.A., 1934. Der Grosse Ausbruch des Vulkanes G. Merapi Mittel Java imJahre 1872. Naturkundig Tijdschrift van Nederlandsch — Indië 94, 189–209.
Hewitt, K., 1983. The Idea of Calamity in a Technocratic Age. Allen et Unwin Inc., London3–32.
Houtkoop-Steenstra, H., 2000. Interaction and the Standardized Survey Interview.Cambridge University Press, Cambrigde (13 pp.).
Lavigne, F., Thouret, J.C., Voight, B., Suwa, H., Sumaryono, A., 2000. Instrumental laharmonitoring at Merapi Volcano, Central Java, Indonesia. Journal of Volcanologyand Geothermal Research 100, 457–478.
Lavigne, F., De Coster, B., Juvin, N., Flohic, F., Gaillard, J.-C., Texier, P., Morin, J., Sartohadi, J.,2008. People's behaviour in the face of volcanic hazards: perspectives from Javanesecommunities, Indonesia. Journal of Volcanology andGeothermal Research 172, 273–287.
Lu, Z., Wicks Jr., C., Power, J.A., Dzurisin, D., 2000. Ground deformation associatedwith theMarch 1996earthquake at Akutan volcano, Alaska, revealed by satellite radar interfer-ometry. Journal of Geophysical Research 105, 21483–21495. http://dx.doi.org/10.1029/2000JB900200.
Mei, E.T.W., Lavigne, F., 2012. Influence of the institutional and socio-economic contextfor responding to disasters: case study of the 1994 and 2006 eruptions of theMerapi Volcano, Indonesia. Geological Society, London, Special Publications 361,171–186. http://dx.doi.org/10.1144/SP361.14.
Newhall, C., Bronto, S., Alloway, B., Banks, N.G., Bahar, I., del Marmol, M.A., Hadisantono, R.D.,Holcomb, R.T., McGeehin, J., Miksic, J.N., Rubin, M., Sayudi, S.D., Sukhyar, R., Andreastuti,S., Tilling, R.I., Torley, R., Trimble, D., Wirakusumah, A.D., 2000. 10,000 years of explosiveeruptions on Merapi volcano, Central Java: archaeological and modern implications.Journal of Volcanology and Geothermal Research 100, 9–50.
vacuations atMerapi volcano, Journal of Volcanology and Geothermal
18 E.T.W. Mei et al. / Journal of Volcanology and Geothermal Research xxx (2013) xxx–xxx
Ratdomopurbo, A., Beauducel, F., Subandriyo, J., Nandaka, A., Newhall, C.G., Suharna,Sayudi, D.S., Suparwaka, H., Sunarta, 2013. An overview of the 2006 eruption ofMt. Merapi. Journal of Volcanology and Geothermal Research (in this issue).
Roman, D.C., Power, J.A., Moran, S.C., Cashman, K.V., Doukas, M.P., Neal, C.A., Gerlach,T.M., 2004. Evidence for dike emplacement beneath Iliamna Volcano, Alaska in1996. Journal of Volcanology and Geothermal Research 130, 265–284.
Sayudi, D.S., Nurnaning, A., Juliani, Dj, Muzani, M., 2010. Peta Kawasan Rawan BencanaGunungapiMerapi, Jawa Tengah danDaerah IstimewaYogyakarta 2010 (Merapi Haz-ard Map, Central Java and Yogyakarta Special Region Provinces). Pusat Vulkanologidan Mitigasi Bencana Geologi, Bandung.
Schlehe, J., 1997. Reinterpretation of mystical traditions: explanation of a volcaniceruption in Java. Anthropos 91, 391–409.
Schlehe, J., 2007. Cultural politics of natural disaster. In: Casimir, M. (Ed.), Culture andthe Changing Environment: Uncertainty, Cognition, and Risk Management inCross-cultural Perspective. Berghahn, Oxford, New York.
Schramm, D., Hansen, R., 1991. Aim and Scope of Disaster Management, vol. I. DisasterManagement Center. University of Wisconsin—Madison, Wisconsin (70 pp.).
Sedena, 2011. Sedena National Defense Secretariat Mexican Army and Air Force. Lastaccessed: Fevrier 12, 2012 http://www.sedena.gob.mx/en/index.php/plan-dn-iii-e/geological-phenomena/664-operations-plan-volcano-qpopocatepetlq.
Southworth, Frank, 1991. Regional Evacuation Modeling: A State-of-the-Art Review.ORNL, Oak Ridge (75 pp.).
Sumiyati, 2007. Makna Lambang dan Simbol Kentongan dalam Masyarakat Indonesia.Historia Vitae, vol. 21. USD Press, Yogyakarta 2.
Surono, Jousset, P., Pallister, J., Boichu, M., Buongiorno, M.F., Budisantoso, A., Costa, F.,Andreastuti, S., Prata, F., Schneider, D., Clarisse, L., Humaida, H., Sumarti, S., Bignami,C., Griswold, J., Carn, S., Oppenheimer, C., Lavigne, F., 2012. The 2010 explosive erup-tion of Java's Merapi volcano — a '100-year' event. Journal of Volcanology and Geo-thermal Research 241–242, 121–135.
Suryo, I., Clarke,M.C.G., 1985. The occurrence andmitigation of volcanic hazards in Indonesiaand exemplified at the Mount Merapi, Mount Kelud and Mount Galunggung volcanoes.Quarterly Journal of Engineering Geology & Hydrogeology 18, 79–98. http://dx.doi.org/10.1144/GSL.QJEG.1985.018.01.09.
Texier, P., 2007. L'éruption du Merapi et le séisme de mai 2006 à Java-Centre: des crisesrévélatrices des réelles causes de vulnérabilité des populations. Reseau Asie — IMASIE,Paris.
Texier, P., Gaillard, J.-C., Lavigne, F., 2009. Les sociétés d'Asie du Sud-Est face aux menacesvolcaniques: Entre phénomènes extrêmes et contraintes quotidiennes. In: Bertrand,
Please cite this article as:Mei, E.T.W., et al., Lessons learned from the 2010Research (2013), http://dx.doi.org/10.1016/j.jvolgeores.2013.03.010
D., Rieutort, L., Thouret, J.C. (Eds.), Villes et Volcans: Relations, Représentations, Pra-tiques. Presses de l'Université Blaise Pascal, Clermont-Ferrand, pp. 71–95.
Thouret, J.C., Lavigne, F., Kelfoun, K., Bronto, S., 2000. Toward a revised hazard assess-ment at Merapi volcano, Central Java. Journal of Volcanology and Geothermal Re-search 100, 479–502.
Thouret, J.C., Gupta, A., Lube, G., Liew, S.C., Cronin, S.J., Surono, 2010. The 2006 pyrolasticdeposits of Merapi Volcano, Java, Indonesia: high-spatial resolution IKONOS imagesand complementary ground based observations. Remote Sensing of Environment114, 1949–1967.
Tilling, R.I., 2008. The critical role of volcano monitoring in risk reduction. Advances inGeosciences 14, 3–11. http://dx.doi.org/10.5194/adgeo-14-3-2008.
Tilling, R.I., Topinka, L., Swanson, D.A., 1990. Eruptions of Mount St Helens: Past,Present, and Future. Special Interest Publication, U.S. Geological Survey (56 pp.).
Tobin, G.A., Whiteford, L.M., 2002. Community resilience and volcano hazard: the erup-tion of Tungurahua and evacuation of the Faldas in Ecuador. Disasters 26, 24–48.
UN, 1985. Volcanic Emergency Management. United Nations: Office of the United Na-tions Disaster Relief Co-ordinator (UNDRO), New York (86 pp.).
UNOCHA, 2006. OCHA Field Situation Report 1–17 Indonesia-Volcano-Update MtMerapi Central Java Province 19 April–26 May 2006. United Nations Office forthe Coordination of Humanitarian Affairs, Jakarta.
Voight, B., 1990. The 1985 Nevado del Ruiz volcano catastrophe: anatomy and retro-spection. Journal of Volcanology and Geothermal Research 44, 349–386.
Voight, B., Constantine, E., Siswowidjoyo, S., Torley, R., 2000a. Historical eruptions ofMerapi Volcano, Central Java, Indonesia, 1768–1998. Journal of Volcanology andGeothermal Research 100, 69–138.
Voight, B., Young, K.D., Hidayat, D., Subandrio, Purbawinata, M.A., Ratdomopurbo, A.,Suharna, Panut, Sayudi, D.S., LaHusen, R.,Marso, J., Murray, T.L., Dejean,M., 2000b. De-formation and seismic precursors to dome-collapse and Fontaine-collapse NuéesArdentes at Merapi Volcano, Java, Indonesia 1994–1998. Journal of Volcanology andGeothermal Research 100, 261–287.
Wilson, T., Kaye, G., Stewart, C., Cole, J., 2007. Impacts of the 2006 eruption ofMerapi Volca-no, Indonesia, on agriculture and infrastructure. GNS Science Report 2007/07. (69 pp.).
