Two numbers describe the same piece of sky above an aerodrome, and only one of them moves. The transition altitude is published, stable and belongs to the aerodrome; the transition level is calculated, changes with the pressure of the day and belongs to the air traffic service unit. Between them sits a layer whose only purpose is to keep two altimeter references from being compared. Almost every reference confusion on a flight deck happens within a few minutes of that layer, and it is structural rather than careless.
Two ways of measuring the same vertical distance
An altimeter is a pressure instrument with a bias control. It converts static pressure into a height above whatever datum the subscale has been told to use. Set the local QNH and it reads an altitude above mean sea level. Set 1013.2 hPa — 29.92 inHg — and it reads a pressure altitude, expressed as a flight level, above a datum plane that has no fixed relationship with the ground at all.
Both readings are honest. They answer different questions. An altitude answers how far am I above the sea, and therefore roughly above the terrain and obstacles that were surveyed against that same sea. A flight level answers which pressure surface am I on, and therefore am I separated from an aircraft on another pressure surface. The first question is about hitting the ground. The second is about hitting other aircraft.
The separation argument is the reason flight levels exist. If everyone in the upper airspace flew on a local QNH, two aircraft crossing a pressure gradient at the same cleared level would sit on different pressure surfaces and would drift together or apart for reasons neither crew could observe. One arbitrary datum turns vertical separation into a purely relative measurement, independent of the weather. The price is that a flight level, on its own, says nothing about your height above the ground.
Somewhere the system has to change from one question to the other. That changeover is not a line. It is a layer, and the two boundaries of the layer are governed by completely different mechanisms.
One number is published, the other is computed
The transition altitude is an airspace-design product. It is fixed by the State, published in the AIP and reproduced on approach and departure charts, and it does not move with the weather. It sits above the terrain and obstacles of the area and above the altitudes at which arriving and departing traffic is normally levelled off. In much of Europe it is an aerodrome or terminal-area value, which is why neighbouring aerodromes in the same country can publish different transition altitudes, and why harmonisation to a common value inside a terminal area has been a standing project for years.
The transition level is an air traffic services product, recomputed as the pressure changes. It is the lowest flight level available for use above the transition altitude, derived from the current QNH by the responsible unit using a table in local instructions or in the State AIP, and broadcast on the ATIS where it is not passed with the descent clearance. On a stable high it holds for a week; on a fast-moving low it can change during a single arrival sequence.
| Quantity | What it is | Who fixes it | Does it move? | Where you read it |
|---|---|---|---|---|
| Transition altitude | The altitude at or below which vertical position is expressed as an altitude on QNH | The State, at airspace design | No — fixed until the airspace is redesigned | AIP, approach and departure charts |
| Transition level | The lowest flight level available for use above the transition altitude | The ATS unit, from the current QNH | Yes — with the pressure, several times a day if needed | ATIS, or with the descent clearance |
| Transition layer | The airspace between the two | Neither — it is what remains | Its thickness varies with the pressure | Implied; not normally published as a value |
That asymmetry is the root of the problem. Crews memorise the transition altitude of the aerodromes they fly to, because it never changes and it is on a chart in front of them. The transition level was heard once, twenty minutes ago, and may already have been superseded.
The transition altitude is a fixed number set by the airport, while the transition level changes based on air pressure. The transition layer is the space between them, where you can’t use either reference clearly. This is why pilots must be extra careful when flying through this layer.
Why low pressure pushes the transition level up
The 1013.2 hPa datum plane is not a fixed height. It is wherever the atmosphere happens to reach that pressure. When the sea-level pressure of the day is exactly standard, the datum plane sits at sea level and a flight level corresponds, near enough, to the altitude of the same number. When the QNH is below standard, the pressure has not yet fallen to 1013.2 by the time it reaches sea level, so the datum plane sits underneath sea level, and every flight level sits lower than its number suggests.
The working rule of thumb near sea level is somewhere between 27 and 30 ft per hectopascal, and it is an approximation that depends on temperature and on the actual pressure profile of the day. Applied to a QNH thirty hectopascals below standard, it puts the 1013.2 datum some eight to nine hundred feet below sea level. An aircraft levelled at FL 45 that day is not at 4 500 ft above the sea; it is closer to 3 700 ft. If the transition altitude of the aerodrome is 4 000 ft, that flight level has just sunk below it.
This is the whole mechanism. The transition altitude is defined against sea level and stays where it is; the flight levels above it slide down as the pressure falls. To keep the lowest usable flight level above the transition altitude by the required margin, the transition level has to be raised. Low pressure, high transition level. High pressure, low transition level. Nothing else in the arrangement is dynamic.
The trap. The direction feels wrong the first few times because both words contain the idea of height, and the pressure is doing the opposite of what the number does. A deep low does not lower the transition level; it raises it, because the flight levels themselves have descended and more of them have become unusable.
At an airport with a transition altitude of 4,000 ft, a QNH of 983 hPa (30 hPa below standard) would push the 1013.2 datum plane about 900 ft below sea level. This means FL 45 (4,500 ft) is actually flying at around 3,700 ft. To keep the lowest flight level above 4,000 ft, the transition level would be raised to FL 50, ensuring safe separation.
The layer exists because the two readings cannot be compared
Consider an aircraft holding at the transition altitude on QNH and another crossing above it on the standard setting. The two numbers cannot be subtracted, because they are measured against different datums, and the gap between those datums is exactly the day’s deviation from standard pressure — small on a normal day, several hundred feet on an extreme one.
The transition layer is the answer to that. By requiring the lowest usable flight level to be positioned a defined minimum above the transition altitude, the system guarantees that an aircraft at the top of the QNH world and an aircraft at the bottom of the standard world are genuinely apart, whatever the pressure has done to the relationship between the two scales. ICAO provisions set that minimum and States implement it in their own instructions; a thousand feet is the value most commonly applied, but the figure that counts is the one in the State’s own text, and some publish more.
Traditionally the layer was transit-only: you climb through it, you descend through it, you do not cruise in it, because a level inside it cannot be given one unambiguous reference. Provisions have since evolved to allow level flight there where the buffer is guaranteed, and practice varies by State. Which model applies is an AIP question, not a general knowledge one — exactly the sort of detail carried across borders by habit.
Europe computes a level, the United States moves the floor
The European model keeps the transition altitude low and local, and makes the transition level the moving part. The American model does the opposite. In the United States the transition altitude is a single national value of 18 000 ft: below it, aircraft use the local altimeter setting; at and above it, everyone uses 29.92 inHg. There is nothing to look up on the ATIS and nothing that varies by aerodrome.
The variability has to go somewhere, and it goes into the lowest usable flight level. When the altimeter setting falls below standard, FL 180 sinks towards the 18 000 ft transition altitude and eventually becomes unusable; the FAA therefore publishes a table that raises the lowest usable flight level in steps as the reported setting falls. The physics is identical to the European case; only the variable that absorbs it has changed.
| European model (ICAO regional practice) | United States | |
|---|---|---|
| Transition altitude | Published per aerodrome or terminal area; varies widely | Single national value, 18 000 ft |
| What moves with pressure | The transition level | The lowest usable flight level |
| Where the crew gets it | ATIS or descent clearance | A published table; ATC applies it |
| Changeover on climb | At the transition altitude of the departure aerodrome | Passing 18 000 ft, everywhere |
| Practical consequence | A value to look up at every arrival | Nothing to look up; low levels are simply withdrawn |
Neither is safer. The European arrangement keeps terminal airspace on QNH, which suits terrain and the vertical profile of an approach, and pays for it with a number that changes and has to be transmitted. The American one removes the transmission problem and pays for it with a very large block of airspace flown on QNH, over which cold temperature errors accumulate.
The moments where the confusion is structural
Reference confusion is not evenly spread across a flight. It concentrates in a small number of situations where the clearance, the instrument and the mental model can legitimately disagree.

The descent clearance that arrives before the ATIS
A crew cleared to a flight level receives its arrival information late — workload, frequency congestion, a runway change — and the transition level in the briefing is the one from the previous listening. The defence is that ATC will not clear an aircraft to a flight level below the current transition level. The exposure is that the crew’s own picture of where the changeover sits is stale.
The go-around
The aircraft has descended through the transition level, both altimeters are on QNH, and the missed approach procedure specifies altitudes — correctly, since it is designed in the QNH world. If the climb continues above the transition altitude under vectors, the reference has to change back in a high-workload phase, competing with a configuration change and a frequency change. It is the one changeover nobody briefs, because the briefing was about the approach.
The hold, and the vector that ends in a level-off
Holding is where an aircraft spends the longest continuous time near the changeover, and a hold assigned as an altitude reads back almost identically to one assigned as a flight level. The deeper risk is time: a long hold lets the pressure be updated, a new ATIS be issued and the briefing numbers age.
The clearance whose reference changed but whose number did not
Phraseology separates the two worlds — a clearance to an altitude carries the word altitude and the QNH, a clearance to a flight level carries the words flight level — but it is only a defence if the readback is checked against the setting on the instrument rather than against the number already in the head.
| Moment of flight | Subscale | Level expressed as | Structural risk |
|---|---|---|---|
| Climb below the transition altitude | QNH | Altitude | Low — the reference is the one used on the ground |
| Passing the transition altitude | Change to standard | Flight level | An early or late change with a level-off inside the layer |
| Cruise | Standard | Flight level | Low, except where a State uses metric levels |
| Descent above the transition level | Standard | Flight level | Stale transition level in the briefing |
| Passing the transition level | Change to QNH | Altitude | Wrong QNH, or a QNH applied to the wrong aerodrome |
| Go-around and missed approach | QNH, then standard if the climb continues | Altitude, then flight level | High — an unbriefed changeover in a high-workload phase |
| Holding near the changeover | Whichever the clearance implies | Either | The pressure and the ATIS both age while you orbit |
Cold temperature corrections apply to altitudes and never to flight levels
The second altimetry error has nothing to do with the subscale. An altimeter is calibrated to a standard atmosphere; in colder air the true height above the ground is lower than the indication, and the error grows with height above the setting source. Hence the cold temperature corrections applied to published altitudes — minimum sector altitudes, procedure altitudes — from the tables in ICAO’s instrument procedure provisions.
Those corrections belong exclusively to the QNH world. A flight level is a pressure surface; correcting it would break the separation logic, because the aircraft above and below would not be applying the same correction. So the same day produces two behaviours: below the transition level, the crew may be flying deliberately higher than the published figure; above it, nobody corrects anything. It is a further reason the layer is a boundary rather than a line, and it is a real trap on a cold, low-pressure day when the transition level has been pushed up and the corrections are largest at the same time.
The right reflex. When the correction is applied to an altitude assigned by ATC, the controller has to be told: the aircraft is no longer where the clearance places it, and the separation picture depends on knowing that. Cold temperature correction is a crew responsibility, but on an assigned altitude it carries a declaration with it rather than being a private adjustment.
What this changes in daily practice
None of this makes the arrangement fragile; it makes it asymmetric. One boundary is a fact of the aerodrome and can be briefed once and trusted. The other is a fact of the day, with an expiry date, and belongs to the current ATIS rather than to the flight plan.
Three habits follow. Treat the transition level as perishable, refreshed with the pressure, not as a line in a briefing completed in the cruise. Brief the go-around changeover as explicitly as the descent one, because the climb back through the transition altitude is the only reference change that happens with the thrust levers forward. And on cold, low-pressure days, treat the two altimetry errors as one problem: the layer is thicker, the corrections are larger, and both push the same way. How individual States publish and update the level is a recurring thread on forums.fl330.com.

Leave a Reply
You must be logged in to post a comment.