Big Data for Disaster Resilience

Globally, Nepal is one of the most disaster prone countries in the world. Every monsoon season, communities across Terai and hill regions suffer from floods and landslides and this year was not an exception. In addition, communities across the country also become victims of other recurring disasters including fires, thunderbolts, glacial lake outburst floods, epidemics and earthquakes. According to the 2015 Global Assessment Report on Disaster Risk Reduction, of the total reported human casualties between 1990 and 2014, landslides and floods accounted for 32.5% and 29.1% human casualties respectively. Similarly, 12.4%, 11%, 8.3% and 6.7% of the losses were due to electric storm, fire, cold waves and other disasters respectively. From economic perspective, floods accounted for 53.2% of losses, fire for 24.4% damages and landslides for 12.9% costs. The 2015 Nepal earthquakes – magnitude 7.8 Gorkha earthquake on April 25 and magnitude 7.3 aftershock on May 12 – also proved how vulnerable Nepal is to natural calamities. As per the available data, the quake and aftershocks killed about 9,000 people and injured close to 22,000 individuals. In addition, the calamity damaged 824,000 houses and had witnessed impairments in 39 districts. Of the 39 districts where damages were reported, Gorkha, Lamjung, Sindhupalchok, Kathmandu, Lalitpur and Bhaktapur suffered major losses. In total, the country suffered from an economic loss worth seven billion dollars. Between 2008 and 2017, Nepal witnessed two major floods – August 2008 flood and August 2017 flood. According to a paper published in the January 2017 edition of Geo Environmental Disasters, the August 2008 Koshi Flood had affected about 2.64 million people in Nepal and India, including 65,000 people and 700 ha productive land in Nepal. Moreover, the flood has also converted 25% of the fertile land in Shreepur, Harupur and western Kushaha villages in Sunsari district unfertile, which are still barren. Exactly nine years later, Nepal witnessed a major flood one more time. This year, the cloudburst made the situations further worse. As of August 18, the 2017 Flood has affected over 16 million people in South Asia. In Nepal, the flood has affected the country’s entire southern region disturbing most of the transportation networks, drowning hundreds of communities, claiming 135 lives, injuring 41 people and 30 individuals missing. In addition, the calamity has completely destroyed 79,812 houses and additional 104,425 houses have been damaged partially. As the flooding incidents wane, districts across Nepal’s Terai region have been reporting additional cases of human casualties and physical damage. Thus, the final damages – human, animal and physical – from the flood are expected to be higher than currently reported. All these unpredictable natural calamities and subsequent unfortunate damages clearly indicate how vulnerable most Nepali communities are towards repetitive disasters and how poor our authorities - be it government agencies or private institutions or non-governmental organisations, local and international or community stakeholders - are prepared for minimising damages from annual disasters. With the country still witnessing migration of large population from comparatively safer hilly districts to more prone Terai areas and the government planning to establish smart cities across the country, concerned stakeholders should pay sincere attention to damages so far from disasters and plan accordingly. Experiences from around the globe show that harnessing technology, specifically big data, could help local and international planners in Nepal to make country’s current and future communities safer. Big Data: Introduction and Scope The term ‘Big Data’ first appeared in a 1997 paper authored by NASA scientists. Since then, there have been many changes in the definition and scope of the term. In September 2014, University of California Berkeley School of Information had published 43 definitions of big data, including those coming from professionals at the university and others at Data.Gov, Ford Motor Company, Google, The New York Times and Columbia University. I find Oracle’s definition - A holistic information management strategy that includes and integrates many new types of data and data management alongside traditional data – of big data as one of the most comprehensive definitions. In addition, the company considers 4Vs: volume, velocity, variety and value as the key characteristics of big data. In modern connected world, data can be easily collected through users’ mobile phones, their travelling patterns, fitness activities, purchasing trends, email and internet browsing history and social media updates. The data thus collected through multiple electronic platforms has become integral to the infrastructure development, proper allocation of resources, planning and policy formulation within cities globally. Experiences from other countries In early 2015, a notable author Bernard Marr opined that world could save 13,000 lives a year from earthquakes by utilising big data for it helps organisations like Terra Seismic to predict future earthquake incidents with 90% accuracy. In fact, most of the high-income Organisation for Economic Co-operation and Development(OECD) countries have long been using big data in minimising risks from natural disasters and also tackling urban transportation and land use patterns. In recent years, emerging economies like China and India too have been exploiting advantages of big data for ameliorating their economies and also easing lives of their citizens. In 2011, damages in major world economies from natural disasters reached estimated 380 billion dollars. Most of the damages came from the magnitude 9.0 Tohoku Earthquake in Japan ($ 300 billion), magnitude 6.3 Christchurch Earthquake in New Zealand ($ 18 billion), flooding in Thailand ($ 12 billion) and Australia ($1.8 billion), two major tornados in the United States ($ 14 billion) and Hurricane Irene ($ 5 billion). In the following years, those economies have had identified inadequate data about people and physical infrastructures in the affected areas as one of the key reasons behind their failures to minimise losses from those disasters and have since been sophistically utilising big data to prepare themselves for worst future scenarios. Even the United Nations agencies have commenced working on big data with experts from governments, private sector and universities at the UN Global Pulse network’s high-tech labs in New York, Jakarta and Kampala and other UNDP country offices across the globe. There are many examples from emerging and least developed countries as well where big data has had played crucial roles while those economies were hit hard by disasters. For instance, the Netherland Red Cross used 510’s; a start-up that specialises in utilising data and artificial intelligence to address disaster risk reduction, resilience and response; big data for rapid shelter planning in worst-hit areas. Similarly, during the century’s most disastrous flood in January 2015, the Government of Malawi used big data collected through earlier community mapping works that helped government and communities to support local recovery undertakings. Big Data in the Nepali Context Generally, there are four major elements of disaster management - prevention, preparedness, response and recovery – and big data can influence all four elements in Nepal. However, damage patterns during the recent floods imply that government authorities and communities themselves ignored future floods ever after suffering huge damages from a destructive flood in 2008. Had they made full use of data collected in the aftermath of the 2008 Koshi Flood and relocated and upgraded most vulnerable communities and physical infrastructures, the net human and animal casualties and damages on houses, roads and other infrastructures would have been minimal this time. Moreover, with available data we could have also developed evacuation route planning for severely affected areas, eco routing for vehicles’ fuel efficiency in emergency locations, and crowd control plans for managing crowds in affected areas for avoiding potential hunger and health hazards among others. However, while observing the response and recovery works, I could find government authorities, private agencies, non-profits and volunteers best utilising big data, mainly gathered through social media feeds, in their activities. As we cannot stop most of the natural calamities from occurring, the only way to be prepared is to utilise what we have, huge data from two major and other annual floods, to minimise damages during such events and speed up post-disaster works in future..

---

Jaya Jung Mahat, an alumnus of the Lee Kuan Yew School of Public Policy at the National University of Singapore, is a Kathmandu-based public policy researcher. He writes extensively on issues that connect economics, politics and innovation. He can be reached at [email protected]