[chris (OP)]

How do aircraft altimeters work? And when a pilot takes off from a higher altitude airport, like Johannesburg, and flies to a low-lying airport, presumably the altimeter needs to be recalibrated to compensate?

[Colin2B]

The altimeter will always be calibrated to read zero at sea level for ‘normal’ air pressure so will read the actual height of the runway at Jo’berg and at the lower airport. It’s up to the pilot to know what height the runway is at.
Using a standard pressure is useful because air traffic control can specify separation heights without worrying about local pressure.

Standard pressure is 29.92 inches of mercury or 1013 hPa, and a report from the destination airfield will give actual local pressure.

I assume @alancalverd  will be able to give a lot more detail.

[evan_au]

Traditional altimeters measured air pressure.

Today's, most altimeters use GPS, which don't need recalibration every time the weather changes.

See: https://en.wikipedia.org/wiki/Altimeter

[alancalverd]

Indeed.  The altimeter has a "subscale" which you can reset to pressure values given by Air Traffic Control or an automatic weather station. The numbers are preceded by a 3-letter code.

QFE  is the atmospheric pressure at the threshold of the runway in use.  Its use should be obvious! An altimeter set to QFE reads "height".

QNH  is the current pressure at sea level. This is important once you have taken off, because the terrain below you is mapped as height above sea level, and at low level (generally below 6000 ft in  the UK) everyone else will be flying with the same QNH, so  you use it for terrain and traffic avoidance. As atmospheric pressure  varies with time and place, ATC will transmit "regional QNH" from time to time. This is the lowest forecast pressure for the next 2 hours in the region, and it can vary quite sharply as you transit the regional boundary. An altimeter set to QNH reads "altitude".

QNE is the standard pressure (1013.25 hPa) setting that everyone uses above the "transition altitude" (6000 ft or whatever the national regulations state). The altimeter reading with QNE set on the subscale, is reported as "flight level" in hundreds of feet,so 12,000 ft is reported as "FL120". This regulates traffic separation in airways: you are well clear of the deck so you can rise and fall with the actual local air pressure, knowing that everyone else at your flight level will be going in roughly the same direction.

There is some debate about the safest way of organising things. As Colin says, some countries dislike QFE and leave it up to the pilot to aim for the published altitude of his destination, using QNH. This avoids possible confusion and reduces workload on departure, but you can't assume that the forecast regional QNH, which is an estimate for the whole region in 2 hours' time, actually applies here and now, so you call for local QNH when you are 10 miles out and end up twiddling a knob and doing mental arithmetic at the same time.

Just to add to the fun, some aircraft have terrain clearance radar that looks ahead, and military and survey aircraft (particularly helicopters) may also have look-down radar that reads actual height above the immediate terrain

[alancalverd]

GPS is great when it works, and one day we may all be flying at a precise height above the geoid, BUT

1.Most aircraft are old! 25 years is a reasonable first-owner life for an airliner, after which it can be refurbished and spend another 10 years in public transport and charter work. Cherished light planes still take people to work after 50 years, and one occassionally encounters "warbirds" comprised mostly of bits of WWI hardware in the sky. GPS? Some of them don't even have electricity!

2. Whilst the authorities want to replace radio navigation and approach aids with published GPS waypoints and approaches, not all GPS systems are certified for primary navigation and final approach. Progress has been glacial at both ends.

3. GPS is principally controlled by the military who can switch it off or jam it for their own purposes. Jamming exercises are held somewhere in Europe every day. At a time of tension it is quite likely that both GPS and GLONASS will be unavailable over Europe, and GALILEO is something of a political football at the moment.

4. If you have ever had a battery failure (twice in the last 2 years for me) you will know why everyone carries a map and stopwatch, and learns to fly on those funny little legacy instruments squeezed in as an afterthought under the GPS entertainment display.

5. For as long as anyone else is flying on pressure instruments, we have a "baro" knob on the GPS panel that replicates the Kollsman window on his altimeter, and we fly on QNH and QNE to keep out of each other's way.

6. By the time you have added two more batteries, a cooling system and a second GPS display to a light aircraft, it weighs more than the original with duplicate primary instruments! I had to scrub a trip a couple of years ago when the "steam clock" Cessna wasn't available, but the local club offered me a posh one with twin Garmins.  Problem was that with my 3 passengers, we were now overweight so I had to find a lighter pilot!

[Halc]

Cherished light planes still take people to work after 50 years, and one occassionally encounters "warbirds" comprised mostly of bits of WWI hardware in the sky. GPS? Some of them don't even have electricity!

They have a 1909 Bleriot near me, claimed the oldest flying aircraft in the western hemisphere. I've seen it in the air.  No, it isn't a reproduction, and indeed, it certainly doesn't have GPS or probably an altimeter of any kind for that matter.
Makes me wonder what the eastern hemisphere has got that is even older.

[chris (OP)]

Problem was that with my 3 passengers, we were now overweight so I had to find a lighter pilot!

I like the way you blame it on the pilot...

[chris (OP)]

Fascinating discussion; thanks.

[alancalverd]

Problem was that with my 3 passengers, we were now overweight so I had to find a lighter pilot!

I like the way you blame it on the pilot...

Everyone always does, and it was supposed to be me anyway! Having given up smoking  50 years ago so I could afford flying lessons, I now need to lose 20 kg to fly my next plane.

[Bored chemist]

I'm just waiting for the first fatality caused by some **** thinking he can navigate using his mobile  phone.

GPS is not very good at determining altitude and the problem is that, if your altitude is important, it can change jolly fast.

[evan_au]

GPS is great when it works...

I see that it will take a while for GPS to take over...

Aviation GPS units include a terrain map, so they not only know where you are, they know where the ground is (including any nearby mountain tops), and the location and direction of airports.

The new commercial GPS chips coming out this year should give enough resolution to land you on a runway - the typically 100m error of today's mass production GPS chips might get you to the general vicinity of the runway... But, as Alan says, this could change at the whim of the military, terrorists or tired batteries...

WWI hardware in the sky

The phrase "Flying by the seat of your pants" was first applied to Douglas Corrigan in 1938...

But it certainly applied to early aviators, who used the accelerometer in their buttocks to tell which way they were accelerating - along with the low-speed gyroscope in their ears and the real horizon - sometimes augmented by a bit of cotton on the windscreen to tell if they were about to stall...

QFE... QNH... QNE

These "Q Codes" were originally developed for use with Morse Code, because they could define standardised short messages quickly. But their use has continued today...

I suspect that Alan would have less use for the Q code which reports sighting of whales in the area...

[alancalverd]

Used sensibly, the mobile phone can be quite handy. As I mentioned above, the definitive navigation aid is an official paper chart, which has the advantage of no batteries or signal outages.For flights below "airways" level, the UK chart is an Ordnance Survey 1:500,000 map overprinted with aviation data like radio frequencies and controlled areas (including runway elevations and altimeter setting regions, which is where we came in). The problem, if you wander off track, is to know where you are on the map (unfamiliar green fields all look the same, and there are plenty of them)  and even if your phone only identifies the nearest town, it can save a lot of embarrassment.

[Bored chemist]

The problem, if you wander off track

It's better not to do that.

if your phone only identifies the nearest town, it can save a lot of embarrassment.

Yes, but that's not really altimetry- which  phones are bad at.

[evan_au]

GPS is not very good at determining altitude

As I understand it, GPS is just as accurate when measuring height above the geoid as it is at measuring latitude and longitude... An error of about 100m with current mass-production chips.

The GPS geoid ignores little things like any local mountain peaks and the current tides and wave height, but it is a reasonable approximation to average sea level.

Your GPS receiver chip has to solve 4 equations in 4 unknowns: Latitude, Longitude, Altitude and Time. 

If there are clear signals from 4 (or more) satellites, then your phone can solve these equations, and determine these 4 parameters with quite good accuracy.

I imagine that if you can only receive from 3 satellites (and no nearby WiFi to provide additional hints), the firmware would ignore the altitude, and just calculate the latitude and longitude as if you were on the geoid(?).

Getting direct line of sight to 4 satellites is much easier if you are up in the air, rather than in the canyons of a city street, or surrounded by mountains and trees.

[alancalverd]

Most folk in older light aircraft nowadays use Sky Demon (other brands are available!) software on a GPS-equipped tablet or a mobile gizmo like a car satnav that you stick on the windshield. This gives you position, track, and absolute altitude in 50 ft increments, overlaid on a standard chart. It's particularly handy for squeezing between aerodrome traffic zones and airways, my least favorite and most used gap being only 500 ft.

Position accuracy is less important at 140 knots: by the time it has registered on your brain, you are somewhere else! But the system does give you terrain and obstacle warnings within 2 miles.

This simple system is not certified for instrument flying but it will direct you to the extended centerline of a runway where you can pick up an ILS and "follow the needles" in extremis.

[Bored chemist]

As I understand it, GPS is just as accurate when measuring height above the geoid as it is at measuring latitude and longitude.

Garmin disagree.
https://support.garmin.com/en-GB/?faq=QPc5x3ZFUv1QyoxITW2vZ6

To get a good measure of latitude you need satellites North and South of you.
To get a good measure of longitude you need satellites East and West of you
To get a good measure of altitude you need satellites above and below you...
Think about it.

[alancalverd]

Satellite below me? In a Cessna 172? Only about a dozen men have ever been higher than a GPS satellite.

[Bored chemist]

Satellite below me? In a Cessna 172? Only about a dozen men have ever been higher than a GPS satellite.

That's my point.

[evan_au]

To get a good measure of altitude you need satellites above and below you.

That is a good point that I hadn't considered - you want a big spread of satellite altitudes to minimise errors in calculation.
- And the Earth does block out satellites "below" you
- A narrow spread leads to a mathematical problem called "ill conditioned equations".

However, if your GPS antenna is well positioned*, it might pick up all satellites 20° above the horizon.
- That includes satellites directly overhead (altitude 20,000km)
- And those near the horizon limits (altitude 7,000km, relative to your position)
- And any in-between
- That is still a reasonable spread of altitudes

the accuracy permitted by geometry considerations remains less than that of horizontal positions. It is not uncommon for satellite heights to be off from map elevations by +/- 400 ft.

If I remember my imperial conversions correctly, 400 ft is not that much worse than the 100m I quoted earlier (Google tells me that 400 ft is 122 meters).

*A good position for a GPS antenna would be on the flat top of the plane.
- A less-good position is on the windshield of a plane
- An even less good position is next to the passenger window of an aircraft or a ship at sea
(I tried the latter last week, and it took a few minutes before it found 4 satellites within view of the porthole...)
- The worst position has to be a car on the streets of New York, where you may only be able to pick up signals from 80° to 90° in elevation, which is only an altitude range of 19,700 to 20,000 km. This is a much lower altitude spread than you would get in a plane, and altitude errors might easily creep into the calculation.
- Fortunately, we have not yet deployed Jetsons-style flying cars, so today, an accurate altitude is not important for cars in New York!

[Bored chemist]

A plane flying high will have a better estimate of the altitude than when it is near the ground.
So the estimate is worst when it needs to be best...