akar.earthStratification: Definition
What is stratification in forest carbon inventory? How dividing a project into similar strata cuts sampling error and cost, with Indonesian tips.
Stratification is the division of a project area into sub-areas, called strata, that are relatively uniform in carbon stock or rate of change, so that each stratum can be sampled and estimated separately. It is a standard step in forest carbon inventory design.
Why it matters
Forest carbon varies a lot across a landscape. If you sample a mixed area as one unit, the variance is high and you need many plots to reach the precision methodologies require. Stratifying into similar groups lowers within-stratum variance, which reduces sampling error, narrows the confidence interval and cuts uncertainty deductions. It also makes results easier to interpret: you can see how each forest type or planting cohort performs.
How it is done
- Choose stratification variables: land cover, forest type, planting year, species, soil, peat depth, slope or biomass maps.
- Map strata: using GIS, satellite imagery, drone data and field knowledge.
- Allocate plots: usually in proportion to stratum area and variance, with a minimum per stratum.
- Estimate per stratum: mean carbon and variance per stratum.
- Combine: area-weighted totals and combined uncertainty for the project.
Ex-ante strata can be updated ex-post if land conditions change, following methodology rules. Our plot sampling design guide and remote sensing biomass guide explain how maps support stratification.
Indonesia angle
Indonesian projects often mix forest types and land uses: degraded lowland forest, secondary regrowth, peat swamp, smallholder agroforestry and new plantings across many villages. Useful strata might separate peat from mineral soil, mangrove zones, or planting cohorts by year and species. Grouped smallholder ARR projects may stratify by planting year and agroforestry system.
Common mistakes
- Creating too many strata, each with too few plots.
- Using outdated maps so plots fall in the wrong stratum.
- Changing strata between monitoring periods without documenting why.
Related terms
See sampling error, confidence interval and control plot, or explore the dMRV hub.
Frequently asked questions
Why stratify a carbon project area?
Carbon stocks vary across a project. Grouping similar areas into strata reduces variance within each group, so fewer plots are needed to reach the required precision.
What variables are used to define strata?
Common ones are land cover or forest type, planting year, species mix, soil or peat depth, elevation, and canopy height or biomass from remote sensing.