Why foresters reach for crown diameter in the first place
When you've got a photo pass or a crown map instead of boots on the ground, crown diameter is the one tree-level measurement you can actually see from above. DBH isn't visible in an aerial image. So the conversion problem is as old as photo interpretation itself: turn a crown width into a trunk diameter, then run that diameter through a volume table.
The relationship isn't fixed across species or stand conditions, so any single equation you pull off the shelf needs caveats attached before you trust it on a cut estimate.
The basic crown diameter DBH relationship
Most published equations take a linear form:
CD = a + b × DBH
or inverted for our purposes:
DBH = (CD − a) / b
where a and b are species- and region-specific coefficients. The regional crown width equations cruisers cite most often (Bechtold's among them) fit this for species groups across different forest inventory regions. A loblolly pine in a plantation and an open-grown white oak in a fencerow do not share a coefficient, and using one set across both will put your volume estimate off by a margin that matters on a cut deal.
Three things move the coefficients around in practice:
- Species. Conifers generally carry narrower, more conical crowns relative to DBH than hardwoods do. Run a hardwood crown model against a pine stand and it overstates diameter at almost every crown size.
- Stand density. Crown-to-DBH ratios shift with stocking. A tree growing in the open develops a wider crown for its diameter than the same species shoulder to shoulder in a dense stand, because it isn't competing for light from its neighbors. An equation fit to open-grown trees overestimates DBH when applied to a closed-canopy stand.
- Crown class. Suppressed and intermediate trees carry smaller crowns relative to their diameter than dominants and codominants. If your crown delineation only picks up the visible crowns in a closed canopy, you're sampling the upper crown classes and missing the suppressed stems, which skews a stand table toward larger trees than the stand actually holds.
From DBH to a volume estimate
Once you've got a DBH per tree, the volume side is the familiar part. Most cruisers already run local volume equations, Scribner, Doyle, or a regional cubic-foot table, and all of them want DBH plus either total height or merchantable height as inputs. Height is a separate derivation problem, usually pulled from a photogrammetric surface model, and it isn't solved by the crown diameter conversion at all.
Here's the practical limit people run into: the crown-to-DBH conversion carries its own error, and that error compounds with whatever error already sits in the volume equation. A cruise plot with tape-measured DBH feeds volume tables with published error bounds. A crown-diameter-derived DBH stacks another layer of uncertainty on top, one that rarely gets quantified in the vendor material selling the conversion. If someone hands you a stand volume number built entirely on crown diameter without naming the equation and species group behind it, ask.
Where this breaks down
Overlapping crowns in a closed canopy are the obvious failure point. Where two crowns touch or interlock, a segmentation pass has to decide where one tree ends and the next begins, and that boundary call changes the crown diameter measurement for both trees. Dense, even-aged stands with full canopy closure are the hardest case for this reason, which is also exactly the stand type where a timber company most wants a reliable stem count instead of a sampled guess.
That's the gap a full-stand crown inventory is built to close rather than paper over. Tree Crown Mapping delineates each crown individually from a sub-10cm pass and reports per-tree crown size across the entire stand, so instead of stretching one regional coefficient over every stem, you've got an actual crown measurement for every tree in the stand, not an extrapolation from a handful of cruise plots.
If you're putting together a cut estimate and want a full crown count behind it instead of a sampled plot average, that's the use case this product is built for.