Nighttime satellite data reveals global flood response patterns
By harnessing repositories of satellite data, scientists have discovered that human settlements with low flood protection levels tend to resettle further away from rivers after catastrophic flood events compared to settlements with more protective infrastructure. The findings shed light on the various strategies that communities adapt to cope with flood risk, a key topic in an age when booming populations mean that more people live in flood-prone areas. Floods have often been some of the most destructive natural disasters in both ancient and modern times, causing over $1 trillion in economic damages and more than 226,000 casualties worldwide between 1980 and 2014. These costs have significantly increased compared to previous decades due to urbanization and population growth near flood-prone rivers, making it imperative to understand how people adapt and respond to the risk of flooding events. Here, a research team led by Johanna Mård examined satellite data on the distribution of artificial lighting in 16 countries from 1992-2013, and deduced how the distribution of human settlements changed in response to flooding events. The scientists also investigated in more detail four "hotspot areas" with differing hydrologic and socio-economic contexts, such as the Mekong River in Southeast Asia and the Brisbane River in Australia. The large-scale analysis showed that human proximity to rivers increased for most countries during the time period, and the increase in distance correlated with the number of flood fatalities; Mozambique, which experienced the most flood fatalities, had the largest proximity increase at around 1,000 meters. This relocation tendency was muted in countries such as the U.S. and The Netherlands that harbor more structural flood protections such as levees and reservoirs. Mård et al. say their research provides a framework for other scientists to improve simulations of flood risk and further characterize global patterns of flood response.