C
OMMUNITY
R
ISK
A
SSESSMENT AND
A
CTION
P
LANNING
P
ROJECT
Compendium of Case Studies – Guidance Notes
Participatory methods of incorporating scientific with
traditional knowledge for volcanic hazard management
on Ambae Island
Bibliographical reference
Shane J. Cronin, David Gaylord, Douglas Charley, Brent V. Alloway, Sandrine Wallez, Job W. Esau,
Participatory methods of incorporating scientific with traditional knowledge for volcanic hazard
management on Ambae Island, Vanuatu, Bulletin of Volcanology 66 (2004), pp. 652-668.
Abstract
A team of international volcanologists, local government officials, and community representatives
worked together on this Pacific island to develop a warning system for volcanic eruption as well as
shelter and evacuation plans. Ambae island is the nation of Vanuatu’s largest active volcano. Some
9,500 inhabitants live beneath a lake perched at 1,340 m in the volcano’s summit caldera. Scientific
and local knowledge was combined in gender segregated as well as combined community meetings.
Care was taken to build confidence and trust. These measures including ritual gift giving and
welcoming speeches customary in island society. The inclusion of a high ranking chief from a
neighboring island as a technical officer and a local woman as a facilitator were key confidence
building measures. Classic methods of Participatory Rural Appraisal (PRA) were employed (see
below), and are nicely reviewed in the journal paper (pp. 655-6).
Other hazards were identified besides mudflows and gas emission associated with volcanic eruption.
These included a large cyclone with a frequency of 15 years and smaller ones every 2-3 years; large
landslides (90 year frequency); earthquake (90 yrs.); drought (10 yrs.); annual floods; frequent
malaria and diarrhea.
Sub-group mapping exercises, subsequently presented to the plenary, revealed the distribution of
population and potential safe areas in the case of cyclone related floods, eruption induced mudflows
and gas emissions. Particularly dangerous valleys were identified. Discussions also revealed past
difficulties with warnings of cyclone and volcanic eruption.
The outcome was the establishment of a disaster committee. At follow up meetings a warning and
evacuation plan was developed.
Vanuatu
Location:
Ambae Island
Date:
2001
Sector focus: Natural hazard preparedness & warning
Spatial focus: Island communities
2 / 4
This case should be of interest in many parts of the world, and not just isolated small islands. The
culturally sensitive and patient process of combining external specialist knowledge with in depth
local knowledge is superb, and the empowerment resulting is a good example of effective
implementation.
Technical description
Hazard/risk type: Volcanic eruption and cyclone
Type of assessment: Natural hazard identification and analysis; vulnerability (exposure) mapping;
organizational/ communicative capacity analysis.
CRA process
Collection and digestion of background data. Community based action planning sessions.
Methods used: Timelines, table comparing changing village situation over 20 years, storytelling,
seasonal calendar, daily activity timetable, transect recording, mapping of areas exposed to hazards
and mapping safe areas, diagramming community organization and decision making, charting and
diagramming traditional warning and disaster management methods, ranking emergency
management needs, community disaster management action planning.
Was livelihood analysis used? Yes, but only to the extent that community mapping included
agricultural fields and paths connecting them to settlements. Livelihood security was not part of the
assessment from a functional point of view, and livelihood location – in the strict spatial sense – was
the only aspect integrated into the volcano hazard planning.
Was external specialist knowledge introduced? Yes: scientific knowledge of process of lahar
formation from volcanoes with crater lakes was combined with local oral history of past lahar flows.
A detailed French/ English topographic map was simplified and translated into the local language as
part of the community based disaster planning exercise.
Vulnerability analysis
Vulnerability was treated in this project in a limited way as identical with spatial exposure. Risk
mapping was a key tool. However, nuances of vulnerability according to age, gender, and wealth
were not featured. In a more egalitarian, small scale island society this may not have affected the
value of the outcome as much as it would in another context.
Capacity analysis
Resources available: Financial: UNESCO South Pacific Office funding with other support from
New Zealand aid and development agency (NZAID) and the South Pacific Applied Geoscience
Commission (SOPAC).
1
Individual international researchers also had foundation support. Human
resources: involvement of national, provincial, and local government officials. Local institutional
resources: hierarchical chieftanship system linked with local churches and women’s groups.
Limitations to capacity: Financial/ material: no telephones on the island and lack of money for
maintenance of radio connections – thus limiting availability of cyclone warning. Institutional:
Hierarchical local social organization and political competition meant that information did not
necessarily flow quickly and evenly in the community. Women were often excluded from decision
making.
3 / 4
Action planning and implementation
What actions were actually planned? Community volcanic eruption and lahar response plan.
Were all actions actually carried out? Yes. All maps produced by the project were posted in a
room in a church, and they served as the basis of on-going work dynamized by the newly-formed
community disaster committee.
Have these actions turned out to be sustainable? The authors (writing two years later) express
doubts without intermittent follow up visits.
Were there any unanticipated additional benefits of the actions? Increased visibility of women’s
contribution to the community and their knowledge.
Were there any unanticipated negative consequences of the actions? None.
Limitations on action/ sustainability of actions: Isolation and distance makes follow up visits
difficult and costly. There is also a natural decay of perceived urgency as the years continue to
elapse since the last eruption (1994).
Indicators
No lahar flows or eruptions have tested the system put in place so far (September 2005).
Contextual notes
Existence/ role of prior or contemporaneous conflict? None.
Role of displacement/ relocation? None.
Role of prior disaster & prior recovery attempts? There had been an eruption of the volcano in
1994, so the hazard was reasonable vivid and fresh in people’s minds.
Significant historical, geographic, economic, political, or cultural issues that influenced this
instance of CRA and its consequences? The social organization is quite hierarchical. This both
has advantages and disadvantages for decision making and warning systems. Geographical
isolation and low income present challenges to systems that require external inputs of information or
technical components (e.g. ICT, radios, etc.).
Strategic notes
How has this practice of CRA influenced change in policy and practice at the national level?
N.A.
How has this practice of CRA influenced change in policy and practice at local level?
Disaster planning, warning, and decision making changed. The focus was narrow, so that although
cyclone and malaria both came up in community discussion groups, no measures for strengthening
houses against wind or managing the habitat of mosquitoes resulted. The project may have laid
down the foundation for such similar actions, but did not initiate them.
How has this practice of CRA influenced the level of organization and solidarity in the locality
where it was carried out? The work was done in a way that reinforced the authority of the
traditional leaders.
Less divided along class, gender, age, ethnic lines? Because of the culturally sensitive way this
work was carried out, traditional male chiefs and assistant chiefs were supported and the existing
4 / 4
political system reinforced. Thus gender divisions were probably very little affected although there is
more awareness of women’s knowledge.
More divided along these lines? No.
Are the people living in this area more able to speak out on issues that concern them?
National plans for ocean evacuation by boat in the case of a large volcanic eruption were criticized in
group discussions because community maps revealed that the proposed pick up point could be
isolated by mudflows and floods. An alternative wharf location was suggested.
Have new civil society organizations been created directly or indirectly because of this
practice of CRA? An on-going community disaster committee was formed.
Lessons learned
It is feasible to integrate highly technical expertise into a CRA process.
Local social institutions are very valuable in providing an environment in which CRA leads to
action.
Keywords
Volcanic eruption, lahar, cyclone, community mapping, gender, traditional hierarchy, early warning,
evacuation, combination of scientific and local knowledge, participatory rural appraisal
Resource person(s)
S.J. Cronin
s.j.cronin@massey.ac.nz
; D. Charley & S. Wallez, Department of Geology, Mines and
Water Resources, Port Vila, Vanuatu, S.W. Pacific; B.V. Alloway, Institute of Geological and Nuclear
Sciences, P.O. Box 2000, Tampo, New Zealand; J.W. Esau, National Disaster Management Office,
Port Vila, Vanuatu, S.W. Pacific.
1
http://www.sopac.org/tiki/tiki-index.php
.
C
OMMUNITY
R
ISK
A
SSESSMENT AND
A
CTION
P
LANNING
P
ROJECT
Compendium of Case Studies – Update
Participatory methods of incorporating scientific with
traditional knowledge for volcanic hazard management
on Ambae Island
Background
A team of international volcanologists, village elders and local government representatives
developed a volcano hazard map and early warning system with Participatory Action Research
(PAR) methods in island workshop from 1999-2003. An atmosphere of trust and mutual respect was
established for local and traditional knowledge of Mount Manaro and outside specialist knowledge
that provided a hazard map as a complement to knowledge of danger signs (for example, migration
of birds off the island, unusual insect and animal behaviour, die-off of vegetation and changes in lake
color). A warning and evacuation plan was made, and committees were formed to maintain
awareness of the eruption hazard.
Update Time Frame, Mode of Follow Up & Confidence Level
On 26 November, 2005, the volcano erupted, and the warning system and action plan were put to a
real test nearly seven years since project inception. Update information was collected by the original
case study co-author and contact, New Zealand volcanologist Shane Cronin, who supplemented a
written follow up report with emails in 2007. Confidence level is very high.
Sustainability
Committees have been maintained to the extent that within two days a successful evacuation could
be mounted on the basis of local resources.
Actions implemented
Two days after seeing black smoke coming from the volcano crater the committee was activated and
evacuation begun. Within eight days some 3,300 of the island’s approximately 10,000 inhabitants
evacuated to zones earlier agreed on the basis of community and specialist mapping to be safer – at
the two ends of the island (each having ports and an airstrip). Emergency response plans
implemented included actions and responsibilities agreed to in advance by the residents leaving
danger zones and those who were to be their hosts.
Vanuatu
Location:
Ambae Island
Date of update: 29 February 2008
2 / 3
Welfare/security results
This locally-based hazard response was accomplished with little cost and negative health and social
impacts on the evacuees and host communities. “Welfare was regulated by Nakamal chiefs of the
host communities. … Despite high levels of tension, overcrowding, inadequate facilities and poor
sanitation, there was little unrest in the camps. Food and supplies were donated throughout the
evacuation period by major Vanuatu NGOs, church groups and businesses, as well as by the
Penema Province communities of Pentecost, Santo and Vila. The Red Cross deployed personnel
and resources to help with managing welfare and supplying water” (Cronin et al. unpublished).
Replication of method/approach
Similar methods were used on islands around greater Vanuatu and eventually provided input to the
central national volcanic warning system (the Vanuatu Volcanic Alert level System – VVAL) as well
as to the process of developing a Penema Province Disaster Plan.
Lessons learned/open questions
1.
A blend or even fusion of local and external specialist knowledge can be highly effective in
providing the basis for local action and self-protection. In this case, the plan developed locally
with outside specialist input was effective.
2.
The PAR process can establish new institutions and enhance older ones in such a way that
risk awareness is raised in a sustainable way. Islanders had not experienced a large eruption
for 100 years until 2005. Nevertheless, the process begun in 2001 was effective in raising
awareness in such a way that people were able to respond to the warning and evacuation
system actively and rapidly.
3.
Since workshops involved only 3-8% of residents (depending on specific venue) and were age
and gender segregated, how much more knowledge and volcano hazard influenced cultural
practice might there still be untapped?
4.
To what degree does the PAR and action planning process incorporate skill building for
improvisation in the face of uncertainty and gaps in both Western scientific knowledge and
local knowledge? In this case the interactions of steam, island formation in the crater lake,
rising crater lake water level, and rainfall were too complex to predict lahar and mudflow
hazard with certainty. Local experience did not reach back far enough to include details of
deadly lahars and mudflows in 1870 and 1914.
5.
As there had been a long history of mistrust of Western scientific warnings (peaking in 1995
when residents misread externally provided hazard map), really how deep does the apparently
newly established mutual respect and trust go? How sustainable is the system beyond this
one trial? If volcanic impacts ever affect the “green” or safer ends of the island, is there an
adequate plan for further evacuation? The case study authors refer to the 2005 event as a
“beneficial false alarm” to the extent that the danger was less than expected. Is there not a
down side to even “partially false” alarms in terms of credibility and community cooperation?
6.
To what extent can this work be the basis for multi-hazard local mobilization and planning for
cyclone hazard, local ocean oil or toxic spill that would endanger the coasts, or mass casualty
event such as air disaster at one of the air strips?
3 / 3
Keywords
Volcano hazard, hazard mapping, evacuation planning, local government, traditional knowledge,
warning system.
Additional contact details
S.J. Cronin
s.j.cronin@massey.ac.nz
(details:
http://volcanic.massey.ac.nz/aboutus/staff/cronin.html
), Institute of Natural Resources,
Massey University, New Zealand (
http://volcanic.massey.ac.nz/index.html
).
Author of Update
Dr. Ben Wisner:
bwisner@igc.org
Document Outline