Choosing Allometric Equations for Tropical Trees

Updated 2026-11-02 · 4 min read · By the akar.earth specialist team

Allometric equation guide for tropical trees: pantropical vs local models, wood density, height, palms, bamboo and mangroves, and how to justify your choice.

For most tropical forest and tree-planting projects, the safest default is the Chave et al. (2014) pantropical equation using diameter, measured height and species wood density, unless a validated local equation exists for your species and size range. The choice matters: the equation is often the single largest source of systematic error in a carbon estimate, and verifiers check it closely.

What an allometric equation does

An allometric equation converts easy measurements (diameter at breast height, height) into hard ones (dry above-ground biomass). It is built by cutting, drying and weighing sample trees, then fitting a model. The equation is only as good as those sample trees: their species, sizes, site and number.

The Chave et al. (2014) model with height is:

AGB = 0.0673 × (ρ × D² × H)^0.976

where AGB is in kg, ρ is wood density in g/cm³, D is DBH in cm and H is height in m. Without height, the same paper provides a model using an environmental stress variable E derived from location.

Decision steps

1. Identify the life form. Dicot trees, palms (coconut, oil palm, sago), bamboo, tree ferns and mangroves need different models.
2. Check the size range. Never extrapolate far beyond the diameters used to build the equation. Many local equations were built on small plantation trees and fail on large ones.
3. Look for local or species equations. Indonesian forestry research has produced equations for some plantation species (for example acacia, sengon and teak) and forest types. Check sample size, site, size range and whether the original data is published.
4. Compare. Apply candidate equations to your own measured trees and compare totals. Big differences signal that one equation is out of range.
5. Default to pantropical with local wood density and measured heights if no local equation clearly fits.
6. Document and justify the choice in the project description, as most methodologies require.

Inputs that drive error

Wood density varies within species and between sites, so give the source and level (species, genus or family mean) for each tree. See CO2 from biomass for the conversion chain from biomass to tCO2e, including root-to-shoot ratios for below-ground biomass.

Special cases in Indonesia

• Mangroves. Use mangrove-specific equations, such as the common equations published by Komiyama and colleagues or species equations for Rhizophora, Avicennia and Bruguiera. Prop roots hold significant biomass and are missed by stem-only models. See Rhizophora mucronata.
• Fast-growing plantation species. Young sengon, jabon or acacia have low wood density and high height-to-diameter ratios. Pantropical models without height can overestimate them.
• Agroforestry crops. Coffee and cocoa shrubs are pruned and multi-stemmed. Use crop-specific equations or measure basal diameter, not DBH. Shade trees use standard tree models.
• Palms. Use palm-specific equations based on stem height. Oil palm and coconut need their own models.
• Bamboo. Clump-based estimation with culm counts and culm-level equations.

Building a local equation

If no suitable equation exists and the project is large, a local equation can be worth the investment. Destructive sampling requires felling permits, a range of tree sizes that covers the project population, careful weighing of stems, branches and leaves, and wood density samples. The sample should include the largest trees you expect, because errors at the top of the size range drive totals. Publish or archive the dataset so verifiers can review it.

Common mistakes

• Picking the equation that gives the most carbon. Verifiers look for this.
• Using a species equation from another continent without testing it.
• Applying an equation built on trees up to 30 cm DBH to 80 cm dipterocarps.
• Ignoring height in forests where height–diameter relationships differ strongly from the pantropical average.
• Mixing units (kg vs t, cm vs m) in spreadsheets.

What good evidence looks like

Give the verifier the equation with full citation, its original sample size and size range, why it applies to your species and site, the wood density table with sources, the height model and its fit, and a sensitivity test showing how results change with an alternative equation. If you developed a local equation, provide the destructive sampling data and fitting report.

Uncertainty from allometry

Model error is usually systematic: it does not average out with more plots. Some methodologies ask you to account for it explicitly, others implicitly through conservative choices. Either way, show you considered it. See uncertainty deductions.

Try the tree CO2 calculator for a quick estimate per tree, and see plot sampling design for how individual trees scale to project totals. More on measurement systems is on the dMRV hub.