Urban environments are changing rapidly. People who live in cities are faced with risks and vulnerabilities that depend as much on socioeconomic factors, like access to health care, as they do on exposure to climate change and variability. Many of the programmes designed to improve urban resilience to disasters and climate change try to incorporate data — such as estimates of summer temperatures by 2050 or number of frost-free springtime days by 2030 — as a key element in science-based decision making. Traditionally, scientists feed such data into impact models, expecting the results from these models to be used in risk reduction and adaptation programmes. But, increasingly, scientists and development practitioners are calling into question such ‘top down’ efforts to use climate data for urban planning. This is partly because they do not account for rapidly changing conditions in cities or the multiple needs and capacities of different decision makers. Read full article by Sarah Opitz-Stapleton.