Indigenous knowledge and science often seem poles apart, but meshing them can curb disaster risk, reports Smriti Mallapaty.
[KATHMANDU] In a poem about an earthquake that killed thousands of people in Nepal and India in 1934, Nepali folk poet Lok Nath Pokharel described the conspicuous and widespread death of snakes, despite the earthquake occurring in the winter when snakes usually hibernate.
Four decades on, in 1975, people living in and around the city of Haicheng, in northeast China, noticed an unseasonal increase in the number of snakes. Three months later, an earthquake struck.
On 26 December 2004, an earthquake off the west coast of the Indonesian island of Sumatra triggered the devastating tsunami that killed around 230,000 people. 
No official tsunami warning system was in place to prepare countries for the disaster. However, several indigenous communities in Indonesia and Thailand, as well as India's Andaman and Nicobar Islands, survived because folk tales they had listened to all their lives alerted them to the dangers of shaking ground and the eerily retreating sea.
Among them, the Moken (or Morgan) people living on Surin Island, 55 kilometres off the Thai coast, recognised these signs as indicating the impending arrival of giant and deadly waves known as "seven rollers". They consequently fled to higher ground, and all survived.
Snakes and folk tales — coincidence, or experience and knowledge which, linked to the insights of science, can offer people around the world significant protection against disasters?
Increasingly, the answer from scientists is: yes, we can learn from indigenous knowledge.
Jiba Raj Pokharel, professor of engineering and director of the Centre for Disaster Studies, Nepal, certainly has. He draws many of his ideas for early warning systems from local knowledge, including snake alerts.
Pokharel proposes constructing snake yards across the Terai plains that run parallel to the lower ranges of the Himalayas. Diurnal and nocturnal snakes would be used to signal earthquakes, which are expected in the near future, after an 80–year lull.
A two-pronged approach
"In disaster management, particularly for earthquakes, I have focused on indigenous knowledge that looks at the behaviour of animals, because conventional science has not been able to develop an early warning system," says Pokharel.
But it's not just a matter of taking local knowledge and inserting it into scientific preparedness plans.
Traditional knowledge doesn't always reduce communities' vulnerability to natural disasters, and may not adapt fast enough to changing social and climatic dynamics, points out London-based risk reduction specialist and co-founder of Secure Futures, Jessica Mercer.
Can snakes and folk tales offer people protection against disasters?
And scientific knowledge may 'clash' with local understanding of disasters, and thus be rejected by communities.
"People have died as a result of depending on spirits or cultural stories," and ignoring warnings from scientists, says Mercer. For example, before an eruption of the Mount Merapi volcano in Indonesia, official evacuation orders based on detailed scientific monitoring clashed with the advice of the volcano's spiritual gatekeeper, Mbah Marijan.
Marijan had received no premonition of a volcanic eruption. Residents therefore thought it unnecessary to abandon their homes and livestock, and instead chose to stay. Mbah Marijan was later killed in a 2010 eruption, having insisted on remaining close to Merapi.
Therefore "the most effective strategies from each [knowledge base] need to be identified to generate a 'safety culture', as opposed to relying on one knowledge base alone to reduce risk among 'at-risk' communities," suggests a paper by Mercer and her colleagues, published in Environmental Hazards in January. 
For Mercer, an important turning point for reforming risk reduction strategies was the 2004 tsunami. Among disaster risk reduction specialists, the tsunami sparked interest in indigenous knowledge, with a focus on integrating indigenous knowledge with modern science, she says. Mercer has now developed a theoretical framework to "combine and integrate the best of both worlds".
Communicating risk — the need for lucidity
One benefit of using both traditional and modern scientific knowledge is that it enables scientists to make the communication of forecasts more accessible to local communities — a problem not confined to developing countries. Communicating risk in a format that people will understand is a perennial challenge.
The UK-based nongovernmental organisation Practical Action, with support from the European Commission, has improved community flood forecasting by introducing scientific tools to existing monitoring systems in riverine villages of Nepal.
Until recently, people would receive telephone calls from relatives upstream warning them about heavy rains and possible flooding, says Yuwan Malakar, a project officer at Practical Action Nepal. "We improved on that method," Malakar explains, by installing sirens and establishing downstream flood thresholds based on upstream river gauge stations, established by the Department of Hydrology and Meteorology.
"When we improve on what [communities] are already doing, they own it," says Malakar. It is also easier to "build [mechanisms] into their daily lives".
The result? When Banke district along the West Rapti River, in Nepal's Mid-Western Region, saw the highest floods for 35 years in August 2012, there were no casualties or injuries, even though it was four years after Practical Action had left the area having established an early warning system there.
Practical Action Nepal has improved flood warning systems in the country
Flickr/The Advocacy Project
When water levels rose from three to five metres at the Kusum hydrological station, they automatically triggered a siren. Over 20,000 people in Banke district were alerted by text messaging, sirens, megaphones and radio broadcasts. They had six hours to bury their valuables and run to temporary shelters, and raised platforms that had been built in the inundation zone.
In southern Nepal's Chitwan District, an area affected by flash floods from smaller rivers, the organisation used local communication methods to improve access to scientific warnings. Besides the conventional sirens, local messengers (katuwals) have been trained to spread information about forthcoming floods, based on rainfall levels in gauges installed by Practical Action.
"Katuwals know in which houses there are disabled persons, pregnant women, and lactating mothers," explains Malakar. This ensures that everyone, including those most vulnerable, are protected.
Rolling out strategies in Africa
Similarly, the Kenyan branch of the Intergovernmental Authority on Development's Climate Prediction and Applications Centre (ICPAC), together with the Kenya Meteorological Department (KMD) and several other partners, has tried integrating local knowledge into climate models with the aim of making seasonal forecasts understandable to farmers.
Gilbert Ong'isa Ouma, a senior lecturer in meteorology at the University of Nairobi, says that for six seasons, the KMD worked with professional weather forecasters from the Nganyi community of western Kenya. The Nganyi traditionally keep a sharp eye on tree and animal behaviour, as well as wind changes.
According to Ouma, the "married" forecast, combining conventional science and local knowledge, produced finer details that were more relevant to local communities, with more specific dates for the onset of rains, the likely length of a season, and indications of where and when it would rain.
Bridging divides, building trust
Trust between scientists and local communities is important, and can be developed through close engagement.
"If we incorporate indigenous practices and try to link them with scientific early warning systems, then the reliability and acceptance among local communities increases, and that's how you can really breach this communication gap," says Rajib Shaw, associate professor at Kyoto University's International Environment and Disaster Management Laboratory.
Building trust is particularly urgent in the emerging context of climate change, Shaw emphasises, as farmers cannot account for drastic and abrupt changes in weather.
"The biggest challenge is validating knowledge — scientific [knowledge] for the local communities and local [knowledge] for the scientific communities," says Mercer, a process that can be "extremely time-consuming."
Regions face common risks, such as drought
Flickr/CIAT Neil Palmer
There are, however, a growing number of people making the time to do this.
Shane Cronin, professor and director of Volcanic Risk Solutions, a research centre at Massey University in New Zealand, offers an illustration from his work with communities on Ambae, an island containing an active volcano in the Pacific country of Vanuatu.
Cronin describes a misunderstanding between locals and scientists. In response to a 1995 eruption scare, French authorities sent small warships to evacuate the Ambae islanders. But the traditional warning signs — changes in the colour of Lake Vui, birds migrating outwards, ants surfacing and colonising vegetation, and other unusual animal behaviour — were not apparent. The strange, and to them, unwarranted, appearance of ships only scared the local people, who left their homes and ran uphill: "Everyone was really angry about the incident afterwards," says Cronin.
Cronin initiated a dialogue in the local Bislama language, through a series of workshops held in two communities. Participants created village histories, volcanic hazard maps and disaster response timelines that incorporated scientific information with cultural beliefs. Among other exercises, they matched local eruption timelines to exact years based on radiocarbon dating. The interactions allowed for "people to see that scientists weren't always going to be individuals they disagreed with, and in fact, that they could learn some things from the scientists, as well as them teaching scientists other things".
After 90 years of silence, the Ambae volcano finally erupted in November 2005. By then, relations between communities and officials had warmed, so when hazard levels escalated to 'red', communities coordinated the evacuation of a third of the island's population for over a month. The awareness raised by the eruption has led to communities requesting the installation of a permanent monitoring station.
Sharing knowledge, forging frameworks
These and other examples show the benefits of an interdisciplinary approach. But Mercer is concerned that efforts still only take place on an ad hoc basis, without interconnection.
She says that although countries such as Papua New Guinea increasingly recognise and state the importance of integrating indigenous knowledge into disaster management, they don't get down to actually applying it: "The 'how' is not there".
Given that many regions face common risks — drought, flooding and hurricanes for example — she proposes that countries develop shared policies or frameworks for how to combine the two knowledge bases in practice.
Cronin believes that international organisations working on disaster management have already begun to internalise this new way of working. This, he suggests, is reflected in the general shift, over the last decade, from relief-oriented work towards the "prevention, risk reduction, and risk-mitigation side of things, where people are focusing on community needs". Roles that used to be filled by ex-military or ex-emergency services staff are now becoming the preserve of geographers, human scientists and physical scientists.
Shaw says he has noticed an increased interest in indigenous knowledge at UNESCO (the UN Educational, Scientific and Cultural Organization). He is advising the organisation during a three-year project, funded by the Japanese government, in Indonesia, the Philippines and Timor Leste, that aims, among other things, to integrate indigenous and scientific knowledge into disaster risk reduction and climate change adaptation practice.
But he adds, "I wouldn't say that we are doing very well in the area of indigenous knowledge. There is still a lot of work that needs to be done."
Environmental anthropologist Lisa Hiwasaki agrees. A programme specialist for Small Islands and Indigenous knowledge in UNESCO's Jakarta office, Hiwasaki says that many people agree on the importance of incorporating local and indigenous knowledge in disaster education, but that in practice little has been done.
"The number of international organisations incorporating [such knowledge] in their work is very, very, limited," she adds. Hiwasaki attributes this to a shortage of three key issues: good examples, methodologies, and time.
"Disaster preparedness is a lengthy process that requires long-term commitment, so it's often overshadowed by 'more urgent' issues," she says.
But when disaster strikes, nothing is more urgent.