Historical urban tree cover change and the potential implications for sustainable urban planning

Urban trees cool streets, manage stormwater and support residents' well-being, but long-term records of how a city's tree canopy has changed are rare. This project maps urban tree canopy change in Malmö, Gothenburg and Stockholm from the 1960s to the present by applying deep learning to historical black-and-white aerial photographs.

The result is a set of city-wide canopy change maps showing where tree cover has been gained and lost over six decades, and what this means for sustainable urban planning. The maps are used in workshops with urban planners and are intended to support evidence-based urban forestry policy in Sweden.

The project builds on an SLU environmental monitoring and assessment (FOMA) pilot study that tested open-source deep learning models for detecting urban trees in historical aerial imagery.