Strip the terminology from greenhouse gas accounting and nearly all of it is one multiplication: how much of a thing happened, times how much emission that thing produces per unit. Understanding which half of that product you can actually improve tells you where a measurement budget belongs.
The Two Halves
| Activity data | Emission factor | |
|---|---|---|
| What it is | Litres of diesel, kg of nitrogen, head of livestock | kg CO₂e per litre, per kg N |
| Source | Your own records | IPCC defaults, national factors, or measured |
| Who controls it | You | Mostly published sources |
| Typical weakness | Incomplete, poorly attributed | Generic, not locally derived |
Activity Data Is Where Programmes Fail
The factor is published. The activity data has to be collected, and at smallholder scale that is genuinely hard.
Three recurring problems:
Incompleteness. Fuel purchased for the farm and used partly for household transport. Fertiliser bought in one season and applied across two. Records that cover the cooperative's purchases but not individual members'.
Attribution. A single input used across several crops, or several plots, one of which is in the project boundary. Splitting it requires an allocation rule, which is a decision, which should be documented.
Recall. Where records do not exist, data comes from farmer recall and recall of quantities over a season is systematically imprecise, usually in the direction of round numbers.
None of this is solved by a better emission factor. A precise factor multiplied by a guessed quantity produces a guess.
Emission Factor Tiers
IPCC organises factors into tiers, and the distinction is about specificity rather than sophistication.
Tier 1 uses global or broad regional defaults. Always available, always defensible, frequently unrepresentative of any particular place.
Tier 2 uses country-specific or region-specific factors, where a national inventory has developed them.
Tier 3 uses detailed models or direct measurement.
Higher tiers are better where the data supports them and worse where it does not — a Tier 3 model fed with poor activity data is less reliable than Tier 1 arithmetic on good records, and considerably more expensive.
Where Upgrading the Factor Is Worth It
Concentrate on the factor that dominates the total. In most tropical agricultural systems that is nitrogen, because the N₂O global warming potential multiplier is large enough that this term often exceeds everything else combined.
A locally-derived N₂O emission factor, where one exists for comparable soils and climate, is worth more than upgrading a dozen minor factors. A fuel factor refined from IPCC default to national specific changes almost nothing, because fuel is a small term.
This is the same principle that governs sampling design and allometry selection elsewhere on this site: identify the dominant term, improve that, and accept defaults for everything the total barely notices.
The Practical Sequence
Get activity data complete and attributable first. Use IPCC defaults for everything. Compute the total and see which factor dominates.
Then, and only then, consider whether a better factor for that one term is available and affordable. Programmes that reverse this order spend on factor refinement while their activity data remains the binding constraint, which is a common and expensive way to produce a number that is precise about the wrong half of the multiplication.
Move beyond estimates
Verifiers test sampling design, uncertainty and whether a number traces back to the field. TREEO dMRV captures that evidence in real time, in one auditable chain.
