Redrawing Auckland’s canopy

What changed

Release 5 combines the 2024 LiDAR survey with aerial NDVI (normalised difference vegetation index) to redraw the canopy.¹ Each current crown is shown as one marker on the map; a crown can hold several trees, and a large tree can be split into several crowns. Source records and earlier detections remain separate records, linked through their positions and overlapping crown boundaries.

The new tree layer uses a 3 m height threshold. Separate layers show canopy 2–3 m tall, vegetation below 2 m, and crowns at least 5 m wide. Those lower layers do not contribute to benefit totals.

Crowns are not merged beyond 400 m²; a larger tree can be split into several crowns. Crown areas of 360 m² or more read “at or near the 400 m² segmentation limit; the tree may be larger” on tree pages and in Species maxima.

Earlier records

Every earlier tree ID remains available. Older dimensions, foliage profiles, neighbourhood statistics, benefit scenarios and survey histories are kept in the full record. Releases 4 and 5 redraw the same survey, so a change in the number of crowns does not establish tree loss, new planting or biological growth.

Areas outside the release 5 coverage keep their earlier crown records. Inside the coverage, earlier detections without a unique new crown can be switched on separately.

Tree identity and protected-tree positions

Records are linked using the actual crown boundary. This is spatial evidence: one crown may contain several trunks or trees. Unverified Notable Trees schedule positions stay at their source locations and are excluded from automatic crown matching.

ALTO retains its stable tree IDs when Council record IDs change. The source crosswalk records the old and current record IDs. The Devonport Primary plaque report is attached to the tree the observer selected; its link to the Notable Trees schedule is still marked for confirmation.

Dimensions and estimates

Dimensions and benefits

Tree panels and 3D illustrations use the new crown dimensions. Existing benefit equations are recalculated once per crown representative. Their assumptions, including uncalibrated trunk and carbon coefficients, remain visible. Earlier estimates are available alongside them.

Trunk diameter

Trunk diameter (DBH) is the diameter at 1.3 m above the uphill ground. No trunk in this release was measured: it holds two model estimates for each crown, one from crown size and height and one from the benefit equations. The tree panel and tree page show their median as the central estimate, marked Estimate, with the two models as its range. A trunk measured at a reviewed KYTE ground visit replaces both.

Root scenarios

Root extents remain scenarios. The current estimates use crown and estimated trunk size, with known paving where available. Neighbouring crowns are no longer treated as measured boundaries between root systems. Effective root extent remains unknown where paving inputs are missing.

The root protection radius is 12 times the trunk diameter, with the area capped at 707 m² (a 15 m radius). The tree panel and tree page draw it from the central trunk estimate, with the range the two trunk models give. It is a planning scenario, not a measured root extent.

Forecasts

Forecasts start at the current height and reuse the earlier fitted model. Its original benchmark is retained; its transfer to the release 5 redraw has not yet been validated. Nearby-history weighting awaits refreshed spatial matching. Older observed changes are kept separate from the new starting height.

What still needs checking

The small reference benchmark found fewer splits with release 5. It does not establish region-wide detection precision.¹ Large spreading and multi-stem trees can still be split or partly trimmed. A photographed Kermadec pōhutukawa remains a recorded example to investigate.

Summed crown area includes overlaps. It should not be read as a land-cover percentage or compared directly with a Council canopy cover percentage. The local board canopy cover and historical-change overlays retain their earlier methods and dates.

Three self-intersecting polygons were repaired and four recorded areas were aligned with their saved outlines. These small corrections are logged; the original segmentation files remain intact.²

Ground visits and contributions

Ground visits are recorded in KYTE: photos, notes and measurements made at the tree. A trunk circumference, a clear photograph of the whole tree and a confirmed species can help test the estimates. Saved visits keep their original tree IDs.