FINAL NARRATIVE REPORT Project Short name: PARTICIPATORY APPROACH FOR COMMUNICATING RISK IN VOLCANIC PRONE COMMUNITIES case

study: Merapi Volcano, Yogyakarta

Authors (team leader & contributing members): Team Leader: Saut Aritua Hasiholan Sagala and Team Members: Ari Susanti Luluk Khoirul, Aria Mariyani, Muhibuddin Usamah Abstract This study deals with identification of disaster preparedness at community levels and the development of appropriate risk communication. We took the case study at communities living on the southern slopes of the Mt. Merapi. Two main works were carried out in this study: a field survey and a participatory approach. The questionnaire data from field survey to 322 respondents were analyzed using statistics and subsequently this information was used in developing the participatory approach. The findings indicate that people have been adjusted to the volcanic risks that they are dealing with. For example, they depend much on their hazard-related factors (disaster experience, hazard proximity and natural signals of disasters). The challenge for the risk manager is to incorporate this knowledge and to understand how the development of emergency planning can help to reduce the volcanic risks. Other important finding is the existence of bottom-up risk communication developed by the local people. This initiative was supported by a local NGO that helps to install the means to communicate over radio communication.

Background and Purpose This research study deals with the problems of volcanic risks in local communities living on the slopes of Mt. Merapi (see figure 1). Mt Merapi is a stratovolcano which is located at the border of Yogyakarta and Central Java provinces. According to Thouret and Lavigne (2005) and Chester et al (2001), at least half a million people are potentially affected by Merapi Volcano. In addition to that, four cities are located nearby the volcano including Yogyakarta, Sleman, Muntilan and Magelang and make them potentially at risk. However the population keeps increasing indicated by the land use changes that take place nearby the volcano (Lavigne and Gunnel 2006). Mt. Merapi has also been regarded as one of the most active volcanoes in the world since it has erupted very frequently (Lavigne and Thouret 2000). The return period of the volcanic eruption is about once in 4-5 years. Therefore the residents live in this volcano are expected to have been experienced with the volcanoes in during their life time. The question remains whether the community living in this community has been resilient against the volcanic risk or simply in vulnerable condition when the volcano erupts.

MT. MERBABU 3145 m MAGELANG

MT. MERAPI 2965 m

06 20 PF

PF

BOYOLALI

STUDY AREA

MUNTILAN

94 19

BOROBUDUR

Progo River

Boyong River PAKEM

Krasak River Gendol River KLATEN Opak River

SLEMAN

Kuning River

N
JAKARTA JAVA ISLAND

MERAPI

200 km

YOGYAKARTA

4 km

YOGYAKARTA

City / urban area

urban area

and ash flow deposits

Uninhabited area Uninhabited area near the volcano Main road main road river
main river

study area

City / Urban area block and ash flow deposits Pyroclastic flow Study area

Uninhabited area near the volcano main road main river

area

Figure 1 Mt. Merapi and The Study Area In 2006, the communities living at the southern flanks of the volcano had to evacuate during the eruption for two months. Some problems occurred during the evacuation process since a large number of the people were not willing to evacuate as reported by national media (BBC, 2006, Kompas 2006a, 2006b, 2006c). According to the media, the reasons for evacuation vary from economic reason, feeling of safety and the traditional cultural beliefs. The problem appears to be on the risk communication that was made by the emergency managers or government officials who are in charge with the evacuation order or process. Nevertheless, it is always difficult to decide on when the evacuation should begin? (Keller and Blodgett 2008). Our case study areas are located at two sub-districts: Pakem and Cangkringan subdistricts (see figure 1). These two sub-districts are located in Sleman district, at the north part of Yogyakarta province. They were selected for our study since these two sub-districts were mostly affected by the eruption in 2006 and many inhabitants in these sub-districts evacuated in 2006 eruption. Hazard Zone As also mentioned in our earlier mid-term report, these two sub-districts are part of Merapi Volcanic hazard zone 2 and hazard zone 3. Hazard zone 3 is located near the hazard source, and frequently affected by pyroclastic flows, lava flows, rock falls and ejected rock fragments (Hadisantono et al 2002). According to the standard by

Volcanological Survey of Indonesia, permanent settlement in hazard zone 3 is not allowed. Hazard zone 3 is the most dangerous zone and therefore can be regarded as an exclusionary zones or controlled access area (Perry and Godchaux 2005). In contrast, in this area people have been living for more than a century. Hazard zone 2 is classified into two groups, namely areas which are affected by mass flow (pyroclastic flows, lava flows and lahars) and ejected material (thick dry volcanic ash fall, volcanic bombs and other ejected rocks) (Hadisantono et al 2002). In the case of increasing activity of the volcano, people living in hazard zone 2 should prepare for the evacuation. The classification of hazard zones was taken into account when selecting the location of the study area. In our study, the unit analysis was a hamlet and in this study we carried out survey to fourteen hamlets located in the two sub-districts. Five hamlets are located at hazard zone 3 and seven hamlets are located at zone 2. Study Purpose As mentioned earlier in the proposal, the study aims to investigate and develop an integrated participatory approach based risk communication in volcanic prone communities, taking communities in Mt. Merapi as a case study. a. To evaluate the effectiveness of existing volcanic warning systems as the means for risk communication b. To assess the evacuation process in year 2006 c. To evaluate the effectiveness of previous evacuation shelter in term of refugees points of view. Given the case happened in 2006, it may be assumed that current residents near Mt Merapi can participate to provide evaluation of the evacuation shelters location. d. To provide suggestions to increase the quality of warning systems, evacuation process and evacuation shelters location. e. To provide evidence in which risk communication could be improved with regards to include peoples knowledge in the context of participatory approach

Figure 2 Hamlets Surveyed in the Study Area

Methods and Partners The methodology of this research includes two approaches: (1) field survey and (2). participatory approach. These methods were conducted in a sequence, starting from the desk study at the preparatory stage prior going to the field, carrying out the fieldwork and applying participatory approach (PRA) The field survey included closed and open-ended questionnaire survey. The former was carried to by using a set of questions prepare beforehand while the latter was intended to get a deeper understanding on peoples perception. Both questionnaire survey was administered to the communities in Merapi Volcano in January February 2008 in fourteen hamlets on the southern flanks of Merapi volcano. The field survey was organized by research team and assisted by some volunteers from INFRONT. The legal process for survey permission was assisted by Research Center for Disaster, PSBA-UGM. During the field surveys we were also helped by some stakeholders. The respondents for the close questionnaire survey were selected randomly from the list given by the local leaders. In total, there were 322 respondents interviewed. This number of respondents represented about 15 percentage of the total households in the hamlets that were selected for the case study area. To make sure the respondents clearly understood our questions, in each meeting we had assistance from 5-6 facilitators guiding the respondents, and each respondent handled four to six respondents. The second survey method was carried out by doing in-depth interviews with 42 respondents. The participatory approach was carried out at two selected hamlets which are located very close to Mt. Merapi. The first hamlet is Turgo hamlet (see figure 1), a

hamlet which is about 6.5 km away from the volcano. This hamlet suffered from the eruption in 1994 where at least 60 people and many other inhabitants injured due to the pyroclastic flow. The second hamlet is Pelemsari hamlet which is about 4.5 km away from the top of the volcano. The eruption in 2006 occurred close to the position of this hamlet. Fortunately, the pyroclastic flow did not reach this hamlet and passed to the other direction. In general, the participatory approach was carried out on the following activities: analysis of problems (SWOT), analysis of history, analysis of information flow, analysis of institutional involvement, analysis of problems identified related with a volcanic eruption. Summary of Results First we discuss the summary from the field survey and then the application of participatory approach. Results from Field Survey During the field survey, we collected general information of socio-demographic variables: age, gender, income, education, type(s) of occupation, household size, type of house. Table in appendix 2 describes the characteristics of our respondents. The respondents were mainly male as our target for this study was the heads of the household. The main reason for this was we aimed to get information on the decisions made at the household level instead of individual decisions. This argument was supported by Lindell and Perry (2004) that in many situation, in evacuation or in an emergency condition, a family will do it together. Most of the respondents are considered poor people however they get many benefits by living close to Mt. Merapi. For example, the land is fertile and it is still easy to get grasses to feed their cattle. The respondents also mentioned they got benefits of the materials released after an eruption which they could use for their building construction as well as for business because they can sell the materials to the city (see figure 3a and 3b). Additionally most of their houses (79%) were built from the materials in Mt. Merapi.

Figure 3a A resident carrying grass

Figure 3b Residents work as sand miners

Despite the benefit from Mt. Merapi, they do face dangerous especially during the eruption period. An eruption may occur for several days or for months. In 2006 the eruption was from April June which made the residents must commute from the evacuation shelter to their land. In the morning they went to their land by truck provided by the local government and in the afternoon they went back to the evacuation shelter. To illustrate how far the residents needed to went and return

back from their evacuation shelter, we show the distances from their hamlets to their evacuation shelter (see table 1). Evacuatio Average distance from hamlet to Volcano n in 2006 Shelter (km) River (m) (km) (%) 1. Turgo 89 8.2 6 150 2. Ngepring 23 4.5 9 500 3. Kemiri 16 4.8 9 175 4. Boyong 35 3.2 9 100 5. Pelemsari 33 5.2 5 150 6. Pangukrejo 86 3.3 6.5 150 7. Balong 0 1.2 8.5 300 8. Karanggeneng 24 0.6 10 400 9. Kaliadem 80 4.2 5.5 200 10. Jambu 90 3 6.5 150 11. Kopeng 96 1.6 8 150 12. Kepuh 87 0.4 9.5 150 13. Kalitengah Lor 100 8.3 5 300 14. Kalitengah Kidul 94 7.3 6 250 Table 1 Distance from Hamlets to Evacuation Shelter, Volcano and River No Hamlet Table 1 also describes the distances from the hamlets to the sources of hazards (volcano and river). Using these data, we calculate the relationships between evacuation decisions and the distances to the volcano and the river (table 2). We found a significant relationship between both distances to volcano and river with the evacuation decisions. Therefore, we could conclude that the evacuation decisions correspond to the distance from volcano and the river. The closer the distance to the volcano, the higher the evacuation rate (%) while the closer the distance to the river the higher the evacuation rate. These findings confirm the previous research studies (Lindell and Perry 2004) that people take hazard proximity or distances to sources of hazards as their consideration when making evacuation decisions. Furthermore, this finding help to convince us that they people take their decisions are reasonable. This is somewhat different to what was reported earlier by most of the media that many people did not evacuate due to their traditional cultural beliefs. Mean SD
Evacuated in 2006 (E) Distance to volcano within 6.5km (DV) Distance to volcano within 150m (DR) Experience of 1994 eruption (E) Decision based on natural signals (NS)
0.63 0.52 0.52 0.09 0.65 0.48 0.50 0.50 0.28 0.48

E
1 -0.45* -0.21** 0.16** -0.06

DV
1 0.02 -0.32** -0.08

DR

DE

NS

1 -0.32** -0.12* 1 0.04 1

** correlation is significant at 0.01 * correlation is significant at 0.05 Table 2 Pearson Statistical Correlation of All Factors The relationship between evacuation decisions (E) and the disaster experience (DE) is found significant. This fact also implies that people take their disaster experience as their consideration while making evacuation decisions. For example, most of residents of Turgo hamlet, who experienced the deadly eruption in 1994, were willing to evacuate.

Results from Participatory Approach

The participatory approach was carried out at two hamlets which were located very close to the volcano: Turgo and Pelemsari hamlets (see figure 2 for the location of the hamlets). Problems identified by the local community in during the eruption are as follows: 1. The condition of cage for taking care the livestock while the residents were away for evacuating was not strong. Moreover the place was not safe and people were afraid of thief taking their livestock away. 2. Apparently, the evacuation shelter of school children were different from the evacuation shelter of their parents and they are far from each other. 3. The availability of vehicles during the evacuation time was very limited. This caused the parents and children were difficult to meet each other. 4. The third evacuation route at Turgo hamlet were not safe as the terrain is very steep and going up and down. 5. The communities lack the facilities to communicate during an eruption (e.g. handy talkie, megaphone, etc). 6. Need for more observation post to detect Mt. Merapi volcanic activities. During our discussion, the residents suggest the problem may be solved by the following sequence: 4, 2, 1 and 3. Despite the limited facilities, the school activity to the children could be carried out regularly using a house belongs to one of residents which is located at a safe place. The space for the evacuation shelter was 7x7 m 2. However the residents felt sufficient with this situation. From the workshop carried out with the residents in Turgo and Pelemsari Hamlets we found got the following information flow (figure 4). It describes the sources of information in general are from government agencies, a siren and monitoring by the local people.

Figure 4 Information Flow We attempted to assess peoples understanding on the volcanic warning systems installed by the Indonesian agency for the volcanic eruption in 2006. The respondents reported that they mainly learned the news of the volcanic eruptions from natural signals (64%) and government officer (57%). This number indicates that, to a large extent, the respondents depend their decisions to the signals that they saw prior to an eruption. Sagala and Okada (in review) also reported that some respondents took natural signals as their source of consideration. These natural signals include tremor, temperature, shapes of the clouds (pyroclastic flow) over the volcano. In general, the information flow for a warning system during an eruption can be described by figure 4. Three sources of warnings were identified: government agencies, an early warning system installed at some places on the slopes of volcano and voluntary monitoring by the local people. The respondents were asked to assess the effectiveness of the warnings in terms of understanding the source of information, understanding the language, and quality of the instruction (see appendix 3.a, 3.b, 3.c). In term of understanding the source of information some 63 percent respondents mentioned that the sources are clear. However this number is far lower than those who mentioned the language use for the warning is easy to understand (77 percent) and those who mentioned the instruction is clear (73 percent). In general, the most of the respondents who live in hamlets close to Mt. Merapi (most of them are in hazard zone) evacuated during the evacuation in 2006. The

evacuation was arranged by the government officials. Trucks were provided in several places close to the hamlets which are located in hazard zone 3. The effectiveness of evacuation shelters provided by the government were assessed using several questions including: quality of the shelter, cleanness of the shelter, healthiness of the shelter and whether the respondents can sleep or not. The results describe respondents opinion over the evacuation shelters as listed in appendix 4 (4.a,4.b,4.c,4.d). Only a few respondents said that the evacuation shelter was convenient (37%). However in term of cleanness about 53% respondents said that the shelter was clean. Similarly, about 50% of the respondents said that the shelter is healthy. About 20% of the respondent mentioned that they could sleep well in the shelter. Due to these reasons, some respondents said they preferred to go home despite the risks from the volcanic eruptions. Therefore, it is important for the government to increase the quality of the shelters. Discussion Our findings above have shown us some identified problems faced by the residents of Merapi when the volcano is about to erupt. However, as mentioned earlier, there are some benefits from the volcano that occur after an eruption or before the next eruption comes. After an eruption the residents will get the benefits from the volcanic materials which they could use for constructing their houses and for the sources of income. Prior to an eruption, as described earlier, the residents enjoy the benefits from the land and mild temperature for tourisms. These conditions are described in the volcanic risk phases in figure 5.

Figure 5 Volcanic Risk Phases Figure 5 explains that there are phases where the residents get benefits from the volcano, which are before and after an eruption while they get the risks most of the cases during an eruptions, except for secondary hazards such as lahars, debris flow and landslides. This figure help us to understand the difference nature of volcanic risk as compared to other risks. Furthermore, this explains why people choose to stay close living with volcanoes despite the risks. This also applies in the context of Mt. Merapi. However in most of the cases, the government and communities put the attention more during the volcanic risk cases. What is important is to increase the preparedness part as the government and the communities need to realize that a volcanic eruption may occur at anytime. Preparedness maybe done by understanding the evacuation plan, preparing tools

such as masks, strengthening housing structures, etc. One option could also be done through purchasing insurance far before a disaster occurs. It is important for the government to explore the current informal institutional setting in the community when planning to increase the level of preparedness. A separate study by Sagala et al found the relationships between personal, community and institutional factors that contribute to people willingness to carry out preparedness. In their findings, the roles of collective efficacy or the intrinsic capacity of the community and community participation are important to motivate people to carry out preparedness. This is similar to what was observed in the field that the communities were more motivated when they worked together instead of individually though the knowledge is provided.

Figure 6.a Institutional Analysis made by the residents of Turgo

Figure 6.b Institutional Analysis Sketch

Next Steps In this section we discuss the future action that needs to be carried out to increase the community resilient against the volcanic risks. Following are several suggestions by people in Turgo hamlet for the improvement of disaster management: 1. Widening of enlargement route 3 which is used by people in RT 04; the present condition of this route is still considered dangerous due to its steepness and narrowness. 2. Evacuation shelter for children should be located near the main evacuation shelter (Barak Pengungsi); this will help to eliminate unnecessary worries of their parents who live in the main evacuation shelter. 3. Schooling activities should be done in the vicinity of main evacuation shelter; this can be done by building additional tent or using the nearest school from the main evacuation shelter as the evacuation shelter for children.

4. An evacuation shelter for elderly; this will give the elderly space and opportunity to socialize with one another without being disturbed by the activities in the evacuation shelter. Expected improvement in the future includes: 1. Improvement of the evacuation route number 3 by widening and reducing the slope steep. 2. Evacuation shelter for the school-children is assigned close to the evacuation shelter for their parents so that the parents may often meet and take care their children by themselves. 3. The school activity may be carried out near the evacuation shelter of their parents or at an assigned tent so that children do not to be far from the evacuation shelter. 4. A special place for an elder people is necessary as they need to interact with other elderly while not disturbing other people. The next research studies shall focus on the following: 1. Identify what ways can be done to increase the preparedness of the local people, including improvement the quality of organization during the eruptions among the stakeholders involved. 2. Identify the possible ways to reduce the evacuation time from the hamlets to the evacuation shelters. 3. Increasing the quality and quantity of disaster education on volcano hazard phenomena and preparedness to deal with the hazards.

SUMMARY OF REFLECTIONS ON RESEARCH PROCESS AND IMPACTS From Team Leader & Team Members: During the completion of this research, we learned that the local communities are unique. Unique means each community has some characteristics that need to be taken into account while making a disaster management plan. As discussed in this report, the responses from the communities to deal with the hazard varies depending where they are located and how the traditional culture has affected them. The communities also have developed their own ways (coping mechanisms) to deal with the volcanic risks. Through time and experiences the coping mechanisms have increased in term of tools (ways) and quality. For example, after the occurrence of 1994 eruption that caused the death of over 60 people, more and more community based organizations on volcanic risks and preparedness have been developed in Mt. Merapi. In our opinion, through time community learns how best they can cope with the volcanic risks. But, actually this can be faster when there are organizations that directly help the community to deal with their risks. From Mentor: This research identified how communities have dealt with the volcanic hazards in their environment. Additionally, this research has contributed by identifying what sort of activities that may work well. Some ideas from this research may be applicable in other communities. This research also contributes to the knowledge on social aspects of disasters which is still lacking in the case of Mt. Merapi. From Host Organization: The host organization was pleased to be involved in this research. Not only this opportunity enlarges our concern on disaster risk management but also the collaboration with people from different backgrounds (academics, NGO and communities) that are involved in this research enrich our experiences. Since our current field studies on forest and environmental management are also in Mt. Merapi, we believe we could integrate the knowledge obtain in this research and our own domain. For example, increasing the quality of the environment and managing the place for residential may reduce the risks that the community face. From Partner Organization And Individuals: The opportunity to explore the collaborative research has provided a good platform for us that research and actions can be carried out together. This is also strengthened by the involvements of several stakeholders such as NGO (Infront), research center (UGM), independent researchers and volunteers and the local communities. From Beneficiaries: Through the discussions during these activities, the local people got reminded of the needs to prepare against disaster risks. Generally, in normal situation, the residents tend to forget the impacts of the risks. However, after the discussions on the disaster preparedness and risk communication, the local people had some better ideas and were triggered to make some plan to deal with the risks they face in their environments.