Can big data analytics support local governments in Nepal to prepare communities against climate and non-climate shocks?
An interview with Sumit Dugar, a Frontier Tech Hub Pioneer.
Sumit Dugar is a civil engineer by training, but these days he prefers to call himself a ‘social engineer’, thanks to his work as a Disaster Resilience Adviser at the British Embassy in Nepal.
“It's been quite a journey from starting my academic life in a technical field, and then going to a policy and practice field,” Sumit tell me, explaining how a Master's in risk and natural hazards from the University of Durham led him to working on early warning systems for floods and strengthening climate services in the region whilst he worked for Practical Action in Kathmandu.
For the past five years Sumit has been at the Embassy in Kathmandu, working with the Nepalese government, UN agencies, NGOs and commercial suppliers to build the country’s capacity for disaster response and preparedness.
Early warnings are not enough
In April of 2015 Nepal was affected by a large earthquake that killed almost 9,000 people and injured around 22,000 more. Just two years later thousands of families were forced out of their homes as floods claimed lives, destroyed homes and ruined food stocks across 35 of the country’s 77 districts.
“In 2015 and 2017 we played a massive role in relief, recovery and reconstruction and building systems to help Nepal prepare against future shocks,” Sumit remembers, “and likewise during the very challenging times of COVID-19 for the past two plus years.” But, beyond these large-scale disasters, Nepal sees annual monsoon floods and regular earthquakes with a magnitude of between 4 and 6 on the Richter scale.
“Our programme is designed to respond to any crisis that may occur anytime at any part of the country,” Sumit explains. “We use technology such as climate services and satellite data, to forecast events and help communities adapt before any shocks happen.“
But early warnings are not enough. Once a warning message is relayed to people who might be at risk, their understanding of that information and ability to take action on it, can vary widely at a local level.
“Early warning systems have developed quite rapidly for these larger ecosystems,” says Sumit, “but for small river systems in Nepal the technology might not be there or there is limited capacity to engage with the communities. Even where the system is quite well developed, there may be a lack of understanding or lack of trust in the messages, and an uncertainty on what action you need to take.”
From cynicism to success
The inspiration for this pilot came from a 2018 paper that Sumit presented about Nepal’s need for a unified disaster information management system. In that paper he explored the idea of linking the various government ministries to create a unified repository across the country’s 753 local governments and seven sub national governments, all of which have their own roles and responsibilities with regards to climate change adaptation and disaster management.
Just a few months later the Youth Innovation Lab (a group that focuses on the role of open data, data driven advocacy, youth engagement and volunteerism on disaster risk reduction and climate change) came to the FCDO and Sumit with a similar proposal.
“They said they wanted to build an integrated disaster information management system that would have all the real time information that can help us look at the flooding situation or extreme rainfall situation.” Sumit says. “Plus, they were also linking with academics, showcasing maps that tell us which areas might be at risk of earthquake, flooding or landslides.”
Sumit was initially wary. He and his department had seen similar projects fail before. “We have seen very fancy systems become unsustainable after two or three years,” he says, “and then there’s no funding and the system dies a slow death. Or it’s a system which is so complex that, if something happens, then we need to contract somebody externally or bring some expensive consultant to do capacity building and troubleshooting.”
Sumit gave the group the benefit of the doubt however, and when he realised they were only asking for an initial grant of $5,000, he decided to take a chance on YILab.
“I thought, ‘Let's take a risk! If they fail, at least we can learn’. I wasn’t too optimistic. But two weeks after our initial meeting we had a prototype that linked with all the data systems for which they had access. Even though they had only five staff, ten volunteers and just $5000.”
Just three months later that prototype had been developed to the point that it was being publicly endorsed by the Nepalese government at an international conference - Global Platform for Disaster Risk Reduction (GPDRR) in Geneva.
Fast forward two years and that working prototype has grown to include more than 500 gigabytes of data, including validated data from the UK Space Agency, Imperial College London and the University of Durham and has datasets on climate change endorsed by the Ministry of Forest and Environment that showcases change in temperature and precipitation pattern under various climate change scenarios.
Members of Youth Innovation Lab looking at the data platform
It is, however, still a national system. What Sumit and YILab want to do with this pilot is make sure that data can be understood and acted upon at a local level.
Influencing and informing
“We have a good, well developed national system,” Sumit says. “But how can we ensure that this data is well understood by local governments? It’s about influencing and informing. How can we influence them to use data on climate change and disasters to make informed planning decisions? And can we inform them so they have the capacity to understand the data and then make the planning decisions with the limited fiscal resources that they have?”
To do this Sumit and the team are planning to work with two local authorities (one in a flood prone area and one affected by local governments), to test their ‘visualising risk’ tool, which Sumit describes as “the local version of the national system”. The task, as he sees it, is to avoid creating “a fancy information management system that nobody uses,” and instead make the data accessible and useful for those making local decisions.
When I ask Sumit how this ‘influencing and informing’ of local governments might eventually impact the local communities, he's ready with a long list of examples. “The construction of safe shelters in locations where people can evacuate if there's a flood warning, the resettlement of people from high risk landslide affected areas to low risk areas, construction of water supply or irrigation canals in areas that are at lesser risk of getting damaged by floods or earthquakes to improve irrigation systems and agricultural produce for people living in the floodplain or living in the hills…”.
But he’s also keen to stress that these longer -term impacts will only come about after a deliberative process of working with authorities that have many other issues to deal with. “Disaster risk reduction might be the 10th thing that they are thinking about,” Sumit remarks, “because local governments tend to be more interested in infrastructure development, health, education and agriculture. But we want to learn about those priorities because they’re the things we often miss while having workshops within ourselves. We're looking forward to knowing something new, something that we might take for granted or we just don't understand today.
“We can use the learnings to ensure that we approach this in the right way. I think we need to start somewhere and I think this is the right time to start.”
If you’d like to dig in further…
👀 Follow this pilot’s journey on their pilot profile page (regularly updated)
📝 Dig into our portfolio evidence on IoT Sensors for Impact (short read)
🎧 Listen to the Data Frontier to hear more about how data is being harnessed for international development and diplomacy (7 episode podcast)