Canopy cover benchmarking and change measurement is critical to urban greening strategy, planning, execution and verification. Previously, cost and effort has meant the intervals for assessment have often been 5 years or more with data based heavily reliant on sampling or averaging techniques, largely limiting the work to an historic point-in-time % measurement.  Using fine-scale (0.5m resolution) satellite imagery, taken at a specified time intervals and interpreted with AI (artificial intelligence – a trained, convolutional neutral network) to identify trees from visual and NIR spectral data, year on year data was analysed.  In a surprising result, individual removed and hazard (dead) trees, tree growth and pruning/decline were clearly visualized.  Information was then compared to climatic data (abiotic) and human effort/pest/disease (biotic).  Integrating with a display interface (ArcGIS), canopy cover could be determined dynamically, measuring and segregating the % canopy cover based on and any viewable, subset area.  By overlaying the data on cadastre, clear accountability could be allocated to by property ownership (e.g. private and LGA).

This approach provides a powerful management tool with hard, timely, visualised data and clear stakeholder accountability.  It changes the urban greening strategy conversation from having a “tree planting” bias (because this has been the simplest measurable) to strategies encompassing a holistic view (including water, nutrients and retention) which, for the first time, can be measured, demonstrated and verified.