A research prototype for mapping Auckland’s urban forest.
This model maps candidate individual trees across the Auckland metropolitan area and attaches modelled dimensions, growth form, a root-space scenario, and exploratory ecosystem-service sensitivities. Trees are detected from LiDAR canopy models and imagery, then matched to council and open inventories. These outputs are not field measurements, tax assessments, or arborist risk advice.
Tree detection, height and crown modelling, trunk-diameter allometry, growth-form classification, ecosystem-service scenarios, root-space scenarios, internal consistency diagnostics, and outstanding validation gaps are documented in the methods set. Model code and documentation.
Current-state model coverage spans the metropolitan extent, but some detections have no usable point-cloud returns. The multi-epoch layer is valid only where historic LiDAR exists and still needs independent removal/establishment validation. Trunk diameter and roots are modelled scenarios. Machine output currently predicts four growth-form classes rather than species, and should not be interpreted as a taxonomic identification.
Developed by Dr Timothy F. Welch, University of Auckland (Waipapa Taumata Rau). Built on Auckland Council open data, LINZ LiDAR and imagery, and i-Tree species references.