Three Airports, Twenty-Five Route Flags, Zero Uncorroborated: How Three Different Hazards Pass Verification


Executive Summary: A Clean Batch Is the Exception, Not the Rule

A route-risk engine that flags city pairs is only useful if someone checks the flags. On 24 October 2026 the engine behind this batch produced 25 flags across three airports, and a separate third-party corroboration pass scored each airport's evidence on a four-level scale. [source-1] [source-2]

The three airports did not score the same. Kansai (KIX) carried an alert-level flag on nine city pairs and earned a corroboration score of 3 — the top of the scale — with two official sources and two date-bearing sources among the results. Incheon (ICN) carried a watch-level flag on ten city pairs and scored 2, with zero official sources in the set and four date-bearing ones. Pudong (PVG) carried an alert-level flag on six city pairs and scored 2, with one official source and three date-bearing ones. [source-1]

The aggregate line is the one that matters: across the three airports, strong corroboration 3, weak corroboration 0, and engine-only claims 0. [source-1] Nothing in the batch asks the reader to take the engine's word for it.

That result is worth examining rather than celebrating, because the architecture behind it explains why it happened and what would have to change for it to fail. The three flags rest on three hazards that are not comparable in kind: a volcanic ash advisory issued by a meteorological agency, clear-air turbulence diagnosed against a flight information region, and a typhoon whose operational expression is an airport closure. Those three hazards have three different issuing authorities, three different verification paths, and three different probabilities of being wrong. [source-1] [source-3] [source-4] [source-5]

This article separates what was verified from what was merely plausible, and states where each of the three flags would break first.

Airport	Engine level	Route flags	Corroboration	Official sources	Date-bearing sources	Verdict
KIX	alert	9	3	2	2	Corroborated at top of scale
ICN	watch	10	2	0	4	Corroborated, no official instrument exists
PVG	alert	6	2	1	3	Corroborated by carrier announcement

1. Kansai: An Ash Advisory 439 Kilometres Away

The Kansai flag is the strongest of the three and the least intuitive, because the number that accompanies it looks like a reason to dismiss it. The hazard source is recorded as an eruption of Mount Sakurajima, Aira Caldera, with the ash advisory filed against Fukuoka FIR (RJJJ) and the source stated at 439 kilometres from the airport. [source-1] [source-6]

Four hundred and thirty-nine kilometres is far. An ash cloud 439 km from an airport does not close that airport in the way a typhoon does. A reader who stops at the distance concludes the flag is a false positive, and the four-level scale would have recorded zero corroboration if that conclusion were right. It recorded three instead. [source-1]

The reason is that the flag is not a distance claim. It is an advisory-area claim. A Volcanic Ash Advisory Center publishes a forecast polygon computed from the eruption column height and the wind field, valid for a stated period, and operators route around the polygon rather than around the crater. [source-7] [source-8] A polygon can extend hundreds of kilometres downwind, and a route that crosses it is affected regardless of how far the vent is.

That is why the corroboration score came back at the top of the scale: the claim "the Kansai approach may be crossed by an ash advisory area" is checkable against a published document with a stated validity window, and it checks out. [source-1] [source-6]

The operational consequence differs from a closure. An ash-affected airport usually stays open. What changes is the routing and the level at which aircraft will accept a clearance into the area, decided by the operator against the advisory, flight by flight. [source-9] A traveller checking a departure board therefore learns less from the board than from the advisory's validity window.

2. Incheon: Turbulence With No Document to Point At

The Incheon flag scores lower — 2 — and the composition of its evidence explains why: zero official sources in the set, four date-bearing sources, and the strongest single item a regional news report rather than a notice. [source-1] [source-10]

The hazard is severe clear-air turbulence, filed against Incheon FIR (RKRR) with the diagnosed distance recorded at 7 kilometres from the airport. [source-1] Seven kilometres means the phenomenon is being described at the airport itself, which makes it a more immediate claim than Kansai's and, in verification terms, a harder one.

Harder because of what clear-air turbulence is. It does not appear on radar, it has no moisture signature, and no agency issues an advisory polygon for it in the way a VAAC does for ash. [source-11] The evidence that it occurred is retrospective: a pilot report, an incident, a turbulence encounter logged after the fact. Ahead of the event there is a forecast, and a forecast of clear-air turbulence is a probability statement over a volume of airspace, not a document with a validity window. [source-12]

That asymmetry is the whole reason for the score of 2. The claim "severe turbulence is diagnosed in Incheon FIR" can be supported by a dated report, but it cannot be tied to an issuing authority in the way an ash advisory can, because no authority issues it. [source-1] [source-10] The engine is not being doubted; it is being scored on a rubric where the maximum score requires an official document and no such document exists for this hazard class.

A traveller should read that as a different kind of warning rather than a weaker one. An ash advisory tells you where not to fly and for how long. A turbulence diagnosis tells you that the smooth-air assumption is wrong over a specific region today, and the only available mitigations are seatbelt compliance and schedule flexibility. [source-13] Neither is a booking decision; both are in-flight and same-day decisions.

3. Pudong: The Flag With a Carrier Behind It

Pudong carried an alert flag on six city pairs and scored 2, but its evidence composition differs from Incheon's in the way that matters most: it has one official source, and that source is a carrier announcement dated 23 October 2026. [source-1] [source-14]

A carrier announcement is a particular kind of document. It is not a regulatory notice and it does not bind anyone outside the airline, but it is an operational commitment published by the party that operates the flight, and it is issued because the airline has already decided something. [source-15] For a traveller, an airline announcement is often more actionable than a government advisory, because the airline is the entity that will rebook them.

The Pudong flags are the ones most entangled with the typhoon. Six city pairs at alert level, including long-haul pairs to European and Japanese destinations, sit in the same batch as the regional cancellation picture around Typhoon Champi. [source-1] [source-4] Whether a given Pudong departure is affected by the storm itself or by the network disruption it causes is not resolvable from the flag alone; it requires the flight-level status. [source-16]

That is the honest limit of a route flag. It says a city pair is exposed. It does not say which flight, on which day, for which reason. The flag is a screening instrument, and the score of 2 reflects that the corroboration available is one carrier's announcement plus three dated items rather than a government instrument. [source-1]

4. Why Zero Engine-Only Claims Is Not the Normal Case

The headline number in this batch — engine-only claims: 0 — deserves a caution, because a clean result is easy to misread as a property of the engine when it is usually a property of the week. [source-1]

Three conditions have to hold for a flag to be corroborated. There must be an issuing authority for the hazard class. That authority must have published something in the retrieval window. And the retrieval must have found it. In this batch all three held for Kansai because volcanic ash has a dedicated advisory system with continuous publication. They held partly for Pudong because a carrier chose to publish. They held only in the weak sense for Incheon because the hazard class has no issuing authority at all. [source-1] [source-6] [source-7] [source-14]

Run the same check in a week with no active eruption and no storm and the picture changes: the flags that remain would rest on historical baselines and standing advisories, and those are exactly the claims that tend to come back at 0 or 1. [source-17] A zero-engine-only result in a quiet week would be a reason to suspect the check, not to trust it.

The other caution is about what "corroborated" means. A score of 3 means a third party published something consistent; it does not mean the traffic consequence followed. An ash advisory can be perfectly verified and produce no cancellations, because operators route around it. [source-2] [source-18]

5. Three Hazards, Three Clocks

The practical value of separating the three flags is that they decay at different rates, and a traveller who treats them as one category will prepare for the wrong one.

Hazard	Issuing authority	Verification path	Clock	What it changes
Volcanic ash	VAAC / meteorological agency	Published advisory polygon with validity window	Hours, reissued	Routing and clearance into the area
Clear-air turbulence	None	Retrospective pilot report, forecast probability	Same day	In-flight procedure, seatbelt compliance
Typhoon	Meteorological service plus airport operator	Storm warning and airport notice	Days, with a landfall time	Cancellations, airport closure, rebooking

The ash clock is the shortest and the most document-driven: advisories are reissued as the plume moves, and an advisory that was valid at 06:00 may be superseded by 12:00. [source-7] The turbulence clock is the shortest of all but produces no document, which is why it scores lowest and why it is the least actionable in advance. [source-11] The typhoon clock is the longest and the most operationally consequential, because it is the only one of the three that reliably produces the cancellation event travellers actually experience. [source-4]

Reading the three in the same table is the error the scoring rubric is designed to prevent. They score 3, 2 and 2 not because one is more real than another but because the verification infrastructure behind them is not equally developed. [source-1]

What that difference costs a traveller is measured in the ground side of the trip rather than the air side, because none of the three hazards is controllable from a booking screen. A fixed-price airport transfer turns the arrival leg into a fixed cost so that an ash-driven reroute to a secondary airport does not become an open-ended taxi fare, an eSIM data plan keeps the carrier and the airport operator reachable when a terminal's own connectivity is saturated by a bank of cancelled flights, and flight compensation is the mechanism that converts a disruption covered by a published advisory into a claim rather than a loss. [source-42] [source-43] [source-44]

6. Conclusion: The Score Measures the Paper, Not the Danger

The batch's clean line — 25 flags, 3 airports, corroboration 3 / 2 / 2, engine-only 0 — is a statement about documentation, and it should be read as one. [source-1]

Kansai scores highest because volcanic ash is the best-documented hazard in aviation: an agency issues a polygon, the polygon has a validity window, and anyone can check whether a route crosses it. [source-6] [source-7] Incheon scores lower because severe clear-air turbulence has no issuing agency at all, so the best available evidence is a dated report about something that already happened. [source-11] Pudong lands in between because a carrier published, and a carrier announcement is an operational fact rather than a regulatory one. [source-14]

The trap is to invert the reading — to treat the ash flag as the serious one because it scored 3 and the turbulence flag as noise because it scored 2. The turbulence is the hazard with no document and no forecast polygon, which makes it the one a traveller cannot monitor, and unmonitorable is not the same as minor. [source-12]

What the batch does establish is narrower: for every flag raised this week, a third party published something consistent, and no flag rests on the engine's assertion alone. [source-1] What it does not establish, and cannot, is which of the three will cost a traveller a day.