Deep data helps cities prepare for disaster
We’ll send you a myFT Daily Digest email rounding up the latest Environment news every morning.
By 2050, nearly 70 per cent of the world’s population will be urban, according to UN forecasts, up from 55 per cent today. Many of these 2.5bn new urbanites will have to adapt to life in cities that are particularly exposed to natural disasters and climate change. Jakarta, built on swampland, is sinking so quickly that the government is relocating to a new capital. Floods wreak regular havoc on Bangkok, Dar es Salaam and Nairobi.
Developing cities are also pollution hotspots. Simply breathing the New Delhi air on a smoggy day is equivalent to smoking 25 cigarettes. The effect of heatwaves is exacerbated in cities, where buildings and infrastructure trap heat and reduce the natural ventilation of wind and breeze. Beneath the earth, seismic disasters also loom, with the likes of Tehran, Kathmandu and Istanbul built on geological faultlines.
Reducing urban risk requires disaster-informed spatial planning, building codes and infrastructure, as well as clear plans for post-disaster response. This relies heavily on data that is often patchy or non-existent, especially in less-developed nations.
However, disaster response experts have begun utilising non-traditional data sources to assist with planning. These include high-resolution images from satellites and drones, localised temperature and seismic intelligence from microsensors, and open data repositories such as social media activity to street maps from thousands of volunteer digital cartographers.
In Santiago, Chile’s earthquake-prone capital, seismic experts have found that combing social media platforms for tremor-related posts produces a similar picture of seismic activity to advanced geological instruments that have been deployed for decades, says Niels Holm-Nielsen, global technical lead for resilience and disaster risk management at the World Bank. While not a replacement for such equipment, the overlap suggests social media analytics “could be of huge value in helping less-developed countries and cities to use these cheaper alternatives” if they lack access to high-end equipment, he says.
Microsensors are producing localised data to help guide public health planning. In Washington DC and Baltimore, a group of citizen scientists working with researchers at Portland State University and the National Oceanic and Atmospheric Administration affixed temperature sensors to their cars and bicycles to produce a rich data set that revealed differences of up to 8 degrees Celsius between areas in a city.
“At a policy level, that’s a huge difference when you look at the public health implications of, say, a 2-degree Celsius increase and what that means for a city,” says Mr Holm-Nielsen. Such data can help city planners and community organisations prioritise hotter areas for public health interventions, green infrastructure and environmental monitoring.
Microsensors are also allowing scientists to build better forecasting models to predict how disasters might play out by giving them access to alternative data sets. One initiative called Tomorrow’s Cities is distributing hundreds of sensors in Istanbul, Kathmandu and Quito to gather data on ambient vibrations, such as rumbling trains or waves hitting a coastline.
They use this to create models of how surface layer geology behaves under stress, says John McCloskey, chair in natural hazards science at the University of Edinburgh, as this can amplify or reduce the effects of an earthquake.
“Running multiple scenarios will allow us to identify areas of particular threat which can be avoided for critical infrastructure,” he says. Urban authorities could provide their development plans and, using this data, “we can assess the risk for scenarios such as earthquakes and floods”.
Tomorrow’s Cities also analyses earth observation data to run disaster scenarios. Using high-resolution satellite images, researchers create digital topography maps that show how artificial and natural structures might behave in a multi-hazard situation, such as landslides caused by an earthquake, which dump sediment into a river to cause flooding.
They can tweak variables, such as exploring what a heavier-than-usual monsoon might mean in a post-earthquake scenario. “Most risk assessment is done on an isolated-hazard basis,” says Hugh Sinclair, professor of surface geodynamics at the University of Edinburgh. “We are now able to combine different hazards that lead to cascades of risk,” he says.
Researchers argue that the end goal of using analytics and advanced technologies is to inform decision-making by citizens and affected communities — powerful allies in the data-gathering process.
Low-tech conveniences, such as mobile phones, can help ensure early warning systems have their desired effects at a local level, says Vivien Deparday, disaster risk management specialist at the World Bank’s Global Facility for Disaster Risk Reduction.
Collaborating with the Red Cross, the agency informed residents about the relationship between localised clogging of drains and flooding.
“They can now get a message that it is raining upstream, and check if their drains are clean,” he says. If they can ascertain that this is the case, “there is a much higher chance the city won’t flood when it rains”.
With the adoption of fifth generation mobile networks, AI and the Internet of Things, cities are becoming more connected by the day. How can ‘smart cities’ navigate the challenges ahead?