The Annual Emissions Report is the document verification is actually about. An hour of structured internal review before it goes out prevents findings that cost days to resolve afterwards.
Review in This Order
Later checks depend on earlier ones. Reviewing the arithmetic before the flight list is complete tests the wrong thing.
- Entity and scope
- Flight list completeness
- Scope classification
- Fuel data and reconciliation
- Data gaps
- Calculation
- Fuels claims
- Plan conformity
- The reconstruction test
1. Entity and Scope
- Is the reporting entity the one holding the Air Operator Certificate under which these flights operated?
- If the group holds several certificates, is each reported separately rather than aggregated?
- Is the scope determination documented with its reasoning, not just its conclusion?
- Have any structural changes during the year — a new certificate, a restructure — been reflected?
2. Flight List Completeness
The check most likely to find something, and the one most often skipped because completeness is invisible by inspection.
- Does the flight count reconcile against an independent source? Slot data, ATC records, the schedule plus known ad hoc movements.
- Are positioning and ferry flights included? They generate no revenue and are routinely missing from commercially-sourced lists.
- Are diversions and technical stops handled consistently?
- Are charter and ad hoc sectors present?
- Does the count by month look plausible against known activity, with no unexplained dips?
A month with a suspiciously low count usually means a data feed failed, not that fewer flights operated.
3. Scope Classification
- Is every flight classified international or domestic, by a documented rule?
- Are aircraft below 5,700 kg MTOM excluded, and is the mass source recorded?
- Are exempt flights — humanitarian, medical, firefighting, State — evidenced individually rather than asserted?
- Is the covered route pair determination based on the participation list as at the relevant date, and is that list captured rather than linked?
That last point recurs: a URL is not evidence of what a published list said during the reporting year.
4. Fuel Data and Reconciliation
- Does the total fuel reconcile against finance records within an explainable margin?
- Where sources disagreed, was the documented rule applied?
- Are escalated cases recorded with the decision and its reason?
- Are there unexplained gaps between fuel uplifted and fuel burned across the year?
- Do any flights show fuel figures implausible for the sector length or aircraft type?
5. Data Gaps
- Is there a gap register showing every gap, its cause and how it was filled?
- Was the fill method pre-defined in the monitoring plan, or invented after the gap appeared?
- Was it applied consistently across all gaps of the same type?
- Is the total proportion of estimated data stated?
6. Calculation
- Is the correct emission factor applied per fuel type — 3.16 kg CO2 per kg for Jet-A and Jet-A1?
- Is fuel in mass rather than volume, and if converted, is the density basis documented?
- Do the totals add up? Sum the monthly figures independently and compare against the annual total.
- Are covered and non-covered emissions separated correctly?
The independent re-sum catches spreadsheet errors that no amount of process review will.
7. Fuels Claims
If CORSIA Eligible Fuels are claimed:
- Is there certification from an ICAO-approved Sustainability Certification Scheme, per batch?
- Is the chain of custody documented from production to uplift?
- Is the lifecycle value applied, with its basis stated?
- Do claimed volumes reconcile against fuel purchase records?
- Is the same fuel not claimed under another scheme?
A claim without complete documentation will be disallowed, and the obligation reverts to the unreduced figure — usually discovered after the budget was set on the reduced one.
8. Plan Conformity
- Does what was actually done match the monitoring plan version in force during the period?
- If the plan was revised mid-year, is it clear which version applied when?
- Have any deviations been identified and explained rather than left to be found?
Drift between plan and practice is among the most common findings and is entirely preventable by reading the plan.
9. The Reconstruction Test
The single check that predicts findings better than any other.
Take five reported figures. Hand them to someone uninvolved in preparing the report, with access only to the files. Ask them to explain how each was derived.
If they cannot, a verifier will not be able to either. That is the finding, identified while it is still cheap to fix.
Choose the awkward ones deliberately — a month with a system outage, a diverted flight, a figure that required a manual decision.
Common Findings This Review Prevents
| Finding | Caught by |
|---|---|
| Flights missing from the report | Check 2 |
| Positioning legs omitted | Check 2 |
| Exempt flights not evidenced | Check 3 |
| Discrepancy resolved with no rule | Check 4 |
| Gap fill method invented after the fact | Check 5 |
| Arithmetic or aggregation error | Check 6 |
| Fuels claim without chain of custody | Check 7 |
| Practice drifted from the plan | Check 8 |
| Figure nobody can explain | Check 9 |
Reviewing the Supporting Pack
The report is the visible output; the pack behind it is what verification actually examines. Review both.
The monitoring plan, every version. With approval dates, so it is unambiguous which applied when. A revision mid-year means two versions matter.
The reconciliation rule set. Is it current? Rules written two years ago against systems since replaced are worse than none, because practice will have diverged from them silently.
The decision log. Every manual intervention with its reason, date and author. Check it is complete for the period rather than trailing off after March, which is the usual pattern.
The gap register. Every gap, its cause, the method applied. Cross-check the total estimated proportion against what the report states.
Captured evidence, not links. The participation list as at the relevant date, programme approval status, certification documents. A URL is not evidence of what a page said during the reporting year, and this is the single most common evidence weakness.
Third-party records. Handling agent dockets, fuel supplier statements. These are the ones most likely to be missing, because obtaining them depends on somebody else.
Previous findings and their closure evidence. Not just the assertion that a finding was closed, but what was actually changed. A repeat finding is treated far more seriously than a first, and the closure evidence is what prevents one.
A useful discipline: assemble the pack as though handing it to a stranger. If any item needs a verbal explanation to make sense, that explanation belongs in writing.
Who Should Do the Review
Not the person who prepared the report. Familiarity is exactly what makes problems invisible.
For a small operator, a colleague from a different function is enough — the reconstruction test in particular works better with someone genuinely uninvolved. For larger operators, internal audit or an external reviewer is worth the cost, provided the external reviewer is not the intended verifier.
Sense-Checking the Numbers
Before the structural checks, a few minutes on the figures themselves catches errors that process review will not.
Compare against last year. A change of more than a few percent should have an explanation — network growth, a route change, a fleet change, a method change. An unexplained shift is usually a data problem, not an operational one.
Check emissions per flight hour. Divide total covered emissions by covered block hours. The figure should be broadly stable year on year and plausible for your fleet. A sharp move points at a fuel data or scope classification problem.
Check the covered proportion. What share of total international emissions is covered? It should match your understanding of the route network and the participation list. A large unexplained change usually means the route-pair logic changed or the participation list was applied inconsistently.
Check month-to-month plausibility. Plot monthly totals. Dips usually indicate a failed data feed rather than fewer flights; spikes usually indicate a duplicate.
Re-sum independently. Add the monthly figures in a separate sheet and compare against the annual total. This catches aggregation errors that no amount of process review will find, and it takes two minutes.
None of these prove correctness. All of them find the kind of error that is embarrassing to have a verifier discover.
Frequently Asked Questions
How long should the review take? Half a day for a small operator, two to three days for a large one with sampling. Far less than resolving findings afterwards.
Can our consultant review it? Yes, and they can also help prepare it. They cannot then verify it — that requires independence.
What if we find an error after submission? Tell the authority and the verifier promptly. A self-identified and corrected error is handled far better than one discovered by the verifier.
Should we sample or review everything? Sample, but choose the sample badly on purpose — awkward cases rather than representative ones.
What is the most common thing this review finds? Missing flights, usually positioning or ad hoc sectors, and reconciliation decisions with no documented rule behind them.
Does the verifier repeat these checks? Broadly yes, which is precisely why doing them first is worthwhile.
What comes after submission? Verification, then the offsetting requirement once ICAO publishes growth factors. See preparing for a CORSIA audit.
Need CORSIA compliance that survives verification? DSTechnoverse builds monitoring plans, data pipelines and reporting processes for Indian operators, and supports project developers through eligibility and placement. See our CORSIA carbon credit services. We are based in Indore, Madhya Pradesh and work across India.
Apply as a CORSIA buyer or seller
Talk to our carbon markets team about your position.
