Why is it desirable to correlate runway names with the Earth's magnetic fields? Is it because the navigational equipment breaks / becomes inaccurate when there's a field shift?
It seems like it could be confusing over time to keep updating runway names as flight paths are repeatedly flown by generally the same sets of pilots, and as paths becomes familiar. For fancy autopilot setups most of the in-the-moment thinking is avoided and this may become a very minor detail.
Edit: Thanks for the quick info everyone! Now this makes sense.
Happy holidays :)
It's so that in the event of mechanical failure a pilot can always rely on a magnetic compass for navigation, without having to know or look up the magnetic variation of the airport.
And also so when they get cleared for “runway 15” they know which side to come on without having to consult a chart.
Also so there’s no confusion, each direction is named, so a runway that runs due North-South will be Runway 36 in the northern direction, and 18 to the south.
In practice it is rare for all but the largest airports to have more than 3 or 4 runways so it isn’t like there will be a Runway 2 and a Runway 3 at the same airport. If they need more capacity they’ll build parallel runways which will be named with a suffix, like 9L and 9R.
(Also, magnetic compasses aren’t that accurate in the first place, since there’s iron in the engines if nothing else. In theory there should be a compensation/calibration done with the results on a card, but, no guarantees that it’s done or up to date).
> If they need more capacity they’ll build parallel runways which will be named with a suffix, like 9L and 9R.
There are even some airports that have three runways with the same heading, so they have "L", "C", and "R". (Two examples that I've personally took off or landed on all three of are 17/35 at DFW and 01/19 at IAD.)
KDEN also put their runways so close together that it's not uncommon to for aircraft on approach on the parallel to set off another airplane's TCAS. I think most airline pilots consider that airport ripe for a midair.
Atlanta has five parallel runways: 27R, 27L, 9R, 9L, and 28/10. It appears the main four that are close to each other are set alternating east landing and west landing as they're not marked on both ends like 28/10 is.
There would be no (or extremely rare, at a minimum) operational reason to alternate directions for takeoff and landings on parallel runways. (There are rules for simultaneous operations to parallel runways and having them be in opposite directions would be problematic as you have to assume some aircraft will execute a missed approach and would then be climbing into opposite direction landing traffic with only a small lateral offset.)
About the closest I can think of to that is some mountain airports that do all takeoffs one way and all landings the other. I’m not aware of any of those with more than one runway and traffic is usually extremely limited.
27L and 9R are the same runway, always. 27R and 9L are the same runway. In Atlanta, besides 27L/9R and 27R/9L, there's also 26L/8R and 26R/8L, as well as the 28/10 you mention. See https://aeronav.faa.gov/d-tpp/2213/00026AD.PDF
This is the same dodge that ORD uses, as DocTomoe pointed out: pretending some of the runways are a different heading when they're actually not. I'm not sure how that can be avoided with four runways, though, since I don't see how to fit a fourth option in with "L", "C", and "R".
In fact, taking this dodge into account, the DFW example I gave actually has five parallel runways; the two they label 18/36 are actually the same heading as the three they label 17/35. I wonder if they'll ever need to go to six and add an 18/36C.
There is also the following suffixes for runway identification:
S = STOL runway, G = glider runway, W = water sealane or waterway, and U = ultralight runway. For example runway 8G / 26G in KBDU (Boulder) and 3W / 21W in KSFF (Spokane).
You want to have a consistent designation for opposite direction runways. (9R is always the same piece of pavement as 27L)
Doing that for 4 would need one direction to be A B C D and the other to be B C D E (which is workable, but doesn't seem particularly more elegant than just using left, right, and center [and in fact is worse in some ways, since from one direction, B is the left-most and from the other, it's A])
Also, EDDF/FRA, in which one runway got renamed to C when they opened a smaller one in the 2010s.
Then you have airports like ORD which have more than three parallel runways - but pretend some of them are ten degrees off to get more naming space. [1]
With respect, calibrating compass compensations is in fact very important in FAA approved aircraft. It's part of the inspection checklist signed off by the licensed mechanic.
My flight instructor once switched off the electricity to the airplane's controls and radios and told me to land. At an airport with a control tower. Using the magnetic compass, airspeed, barometric altimeter, RPMs, that's it. Nothing else works without electric power. There's a whole procedure for doing all that. Being able to navigate safely with an old-school magnetic compass is life-critical.
Not to doubt your story but being a flight instructor myself, I cannot imagine intentionally turning off power to airplane radios at an airport with a control tower (this is also illegal). As far as the pure mechanics of landing go, one does not need any instruments whatsoever: one can judge the airspeed by the sound and the wing angle. I routinely make my students do this when I teach tailwheel flying.
Not to mention that no one would ever be insane enough to turn off electronics and then say "right, now IFR your way in". You're flying VFR, you're going to have eyes on the runway.
> Not to doubt your story but being a flight instructor myself, I cannot imagine intentionally turning off power to airplane radios at an airport with a control tower (this is also illegal).
You just pre-arrange it with the tower and it's no big deal. Probably exceptions for busy class Cs or Ds, but it's no big deal to any airport near me.
I once took off with my pitot tube covered. I didn’t even notice while taking off it wasn’t until I was 100ft agl when I scanned my instruments and realized my airspeed indicator was dead. After a split second freak out I realized I had gotten here based on external indicators. So I calmly went right back into the pattern and landed using purely sound, wing angle, visual reference, and muscle memory.
I had a scuba instructor who used to sneek up behind and turn off our tank valves... in the pool. You dont do such things in the real world.
Also, in the real world, i would probably break out my smartphone. Short of an emp strike, it will still run on its own battery and can act either as a rough set of instruments or to phone the tower and declare an emergency.
It's a convention for A LOT of aspects of flying: air traffic control instructions ("turn left heading 090"), charts ("fly heading 070 after takeoff"), etc are all in magnetic coordinates.
Planes also a normal compass as their ultimate backup instrument, which of course is also magnetic.
Finally, while modern systems could operate off true north instead, not ALL planes have that.
Basic safety check. If I'm on RWY 31 at PAO, then my compass better read 310.
Also, winds (and just about everything else) are in magnetic degrees. If I have wind from 280 while I'm landing, then I immediately know that I have a left quartering headwind and what to expect.
For the uninitiated, PAO is a small general aviation airport in the city of Palo Alto aka the heart of Silicon Valley. Nearby is also SQL, adjacent to Oracle's (former) HQ. That name is purely coincidence!
Having just flown out of there yesterday, on the return it is also helpful, it's also helpful to know without having to look it up what I should set my OBS to in order to help align for practicing my pattern and approach.
In order to need compass, I will need to lose 5 GPS devices in the plane, 2 portable GPS (iPad and iPhone), 2 VORs, ADF, radar (yes it can be used for navigation), and 3 radios (ask for vectors). No I don’t care about compass or it’s accuracy.
One other thing not covered, on any airport chart, the exact alignment of the runway will be noted, usually down a tenth of degree.
So the numbering is a useful backup and discriminator, but the charts have much more detail, and modern navigational and instrument systems on runways obviate much of this for most flights anyways.
It’s super helpful and intuitive when you’re flying. Took me a while to get the compass visualized in my head but now if I’m making a visual approach and they give me a runway I can look at my current magnetic heading on my heading indicator - which you periodically sync to your magnetic compass - and know exactly what direction relative to my current heading I’ll be landing. So that’s the fundamental reason why most things flight related are magnetic except rare items like winds aloft reports. Wind direction reports at the field? Magnetic, so that you know how many degrees off the wind is from your landing direction and can do a bit of mental math to get the crosswind component. If you’ve just taken off ATC may say fly runway heading - again an easy one using your heading indicator which is synced to your magnetic compass.
As a side note, modern avionics like G1000 auto sync your heading indicator to magnetic. So if we started flying with electronic avionics that are auto synced and have a database of magnetic variation around the world, then it’s feasible to always show true on the heading indicator and have runways and everything else be true bearings. But that’s not the case and I still fly some of the many steam dial planes out there as do others, so we’re stuck with magnetic runways and magnetic everything else used in real-time flight and situational awareness for pilots.
The difference between them depends on where you are on the globe - it could even be 180 if you happen to stand in the middle of the two!
Magnetic compasses always point to magnetic north, but with earth being molten metals - magnetic north likes to move around a bit, so we work out the difference between the two and apply it to the directions we track.
For example, where I am - the difference is around 11 degrees. In order to head directly west, I can't just let my compass point north and go directly perpendicular port - I have to point north plus 11 degrees, and then go. My actual track will be 281 degrees magnetic, not 270.
What's more interesting is that most navigation charts have all their directions and bearings in "true" north, with the variance also charted, whereas when you actually fly you track the magnetic headings plus/minus the variance. If you go long enough - the variance also changes so you have to take that into account when planning your route.
When you take off / land, runway headings directly corresponding to your heading are a good sanity check - alarm bells go off if you're meant to be flying west but come across Runway 09 in front of you (the opposite!), not to mention we still rely on magnetic compasses - you don't want to be coming into land at a track for 281 degrees and see 27 painted on it; chances are something went wrong somewhere.
The first non-stop flights to Hawaii from mainland US were part of a competition put on by Dole leaving from Oakland.
The 1st place plane took off after the 2nd place plane. This happened because the navigator of the 2nd place flight didn’t take into account the great circle distance and flew straight headings.
In practice, there's not a specific distance but planning convenience.
For short manually planned tracks (eg VFR planning), most people will plan from one visually helpful waypoint to another, and measure the heading at the mid-point of each route segment. The midpoint approximates the GC heading, and anyway the discrepancy is (very) small on short segments.
For long tracks (eg IFR interstate or intercontinental flights), most plans comply with standard routes, between nominal waypoints, and the route segments together may approximate GC tracks. Between each waypoint, one normally flies a constant heading.
This is imperfect, but variations due wind and traffic are far bigger than losses due not quite conforming to a GC track.
Madrid-Los Angeles indeed goes all the way north of Newfoundland, even northern than the great circle. The exact route taken over the Atlantic may vary from day to day depending on weather conditions.
Currently true north and magnetic north are broadly the same in the UK - and over the next year or so, there is a point where grid north (Ordnance Survey maps) also coincides with both other “norths”. That point will track north over the UK.
The video shows Matt visiting Cornwall at a place where that point was.
My understanding is that the change in magnetic declination (caused by the change in magnetic north) is predictable in the short term (this is why maps state the rate of change but only for a certain time period) but is not predictable over the long term.
Thanks for the actual link, I see lots of physical details included in there, which I pretty much dig.
Related, during the last two weeks or so I've been immersing myself in exploring the topographical maps of Bulgaria the Soviets drew towards the end of the '80s (it was for a personal project of mine). For those interested the map can be found in here [1], by scrolling further down the page, a typical map-sheet looks something like this [2]
Interesting fact: the magnetic variation in Canada changes at a more frequent rate, and some northern airports got tired of updating their runway numbers every year or two so they now use true instead of magnetic headings (I believe indicated on the runway numbers like 17T to indicate true instead of 17 which would indicate magnetic).
Technically there's a place (well, line) where North is one way, but the compass points south.
Flying a great-circle route that intersects the North Pole, and magnetic North Pole, would mean travelling both north and south at the same time for a period.
It's perhaps unfortunate that we live in an era where they are "close enough" that we have one word (North) for two different directions.
> we live in an era where they are "close enough" that we have one word (North) for two different directions.
Is it possible for the magnetic poles to ever exist far from the geographic poles? (I know the magnetic field could have been the opposite polarity, or not existed at all, but in either of those cases, we still wouldn't have gotten another word.)
> Is it possible for the magnetic poles to ever exist far from the geographic poles?
The north magnetic pole has been south of 70 degrees as recently as the mid-1800s and the south magnetic pole is currently north of 65 degrees. Every half million years or so the whole thing goes nuts for a while and reverses.
During polarity flips it can be anywhere but you have to really well define your terms as to what you count as a pole.
I don’t think the exact mechanics of the magnetic field are super certain, but the general idea is that the rotation sort of pins the poles nearish to the axis but there’s been some significant variation historically.
How is magnetic north defined? In relation to negative electric charge, or just by reference to a standard compass? (I.e. "whatever is attracted to this side of my magnet is magnetic north" )
Those are the same thing. The Earth is a planet sized (highly irregular) magnet, and compasses are very simple measuring devices for finding the alignment of the (local) magnetic field. Magnetic North is the "point" (it's not actually a point at all) on the surface of the planet where all magnetic field lines that you can trace by walking towards compass-incidated North converge.
And those lines won't be even remotely straight =)
I got criticized for not just hard coding the runway names in a billing app. I said don't y'all know these are the end times and we are at the early stages of a pole shift. I had the last laugh a few years later when the runways changed.
When I was a kid I wondered how we knew space was a vacuum.
We have a good (excellent) idea of the density of the core. We also have an excellent idea of how density changes with pressure. Add in that the core must be magnetic and a fairly common isotope in the universe, and you're pretty much left with Fe.
I still wonder how spinning molten iron can generate a magnetic field. Spinning a piece of solid iron doesn't produce a magnetic field, so why does molten iron?
Summary: "The debate about changing from Magnetic to True navigation is now moving towards how to change, and when, with March 2030 as the proposed date."
There was an interesting event a few y̶e̶a̶r̶s̶ months ago around Long Island, where it appeared that the Air Traffic Control software got confused by the magnetic declination. Basically it was generating false radar targets in the wrong place: They were rotated 13 degrees, the magnetic declination in the area, from the actual aircraft.
A short summary is that sometimes when ATC radars see an aircraft, if it doesn't seem to have ADS-B (is in a position where no known ADS-B-equipped aircraft is), it will create a synthetic target, broadcasting it out so that aircraft with ADS-B In can see it on their traffic displays. That system is called TIS-B. Somehow the radar in Islip got confused by the magnetic declination: it was generating TIS-B targets that were rotated 13 degrees from where the aircraft really were. And it was doing it for pretty much every aircraft, whether or not it had ADS-B. I assume it wasn't able to correlate the known ADS-B position with the secondary radar returns because they were offset by 13 degrees.
Very weird way for a pretty important piece of ATC infrastructure to fail.
"A few years ago" would have made perfect sense .. the tweet, however, is dated Mar 29, 2022 and states "Seems like something weird was going on with ATC systems around New York a few days ago."
None the less I'd (semi educated) guess that someone at the rader site in question "updated" | "improved" the implementation of The World Magnetic Model [1] in use there and effectively dropped the correct corrections from the end result (for a few days until it was noticed and fixed).
Normally coefficient sets are updated every five years .. and frequently sites and software that use the WMM find themselves with a current IT support person that has not previously encountered a WMM update and then backtrace their way into an "understanding".
This can cause errors until noticed.
Critical infrastructure places (Military, Air Traffic Control, etc) tend to fare better and have longer periods of staff retention, etc. but it can still happen that "we no longer have anyone that was here the last time (five years ago) this happened"
However; given the last epoch update (WMM2020) was released 12/10/2019 (in 'merc date speak MM/DD/YYY (barbarians)) and remains current until expiring on December 31, 2024 we can rule out a simple epoch update FUBAR.
This leaves the other semi regular source of WMM errors - programmers that don't like FORTRAN (esp. super opaque single letter variable 25th degree spherical harmonic functions that date from the 1960s) and feel the need to "do it better".
Having barely survived the prior epoch update and knowing that another will happen again, they carve out some space to build a better mousetrap so that next time will be easier.
This is often a, uhh, interesting intersection of ego and hubris.
This happened here in Tampa (TPA) just a few years ago. They shut down the runways to rename them from 18 to 19 to account for the magnetic shift. Must be a lot of repainting, it took a few days.
87 comments
[ 3.0 ms ] story [ 155 ms ] threadIt seems like it could be confusing over time to keep updating runway names as flight paths are repeatedly flown by generally the same sets of pilots, and as paths becomes familiar. For fancy autopilot setups most of the in-the-moment thinking is avoided and this may become a very minor detail.
Edit: Thanks for the quick info everyone! Now this makes sense. Happy holidays :)
Also so there’s no confusion, each direction is named, so a runway that runs due North-South will be Runway 36 in the northern direction, and 18 to the south.
In practice it is rare for all but the largest airports to have more than 3 or 4 runways so it isn’t like there will be a Runway 2 and a Runway 3 at the same airport. If they need more capacity they’ll build parallel runways which will be named with a suffix, like 9L and 9R.
(Also, magnetic compasses aren’t that accurate in the first place, since there’s iron in the engines if nothing else. In theory there should be a compensation/calibration done with the results on a card, but, no guarantees that it’s done or up to date).
There are even some airports that have three runways with the same heading, so they have "L", "C", and "R". (Two examples that I've personally took off or landed on all three of are 17/35 at DFW and 01/19 at IAD.)
https://aeronav.faa.gov/d-tpp/2213/00119AD.PDF
There would be no (or extremely rare, at a minimum) operational reason to alternate directions for takeoff and landings on parallel runways. (There are rules for simultaneous operations to parallel runways and having them be in opposite directions would be problematic as you have to assume some aircraft will execute a missed approach and would then be climbing into opposite direction landing traffic with only a small lateral offset.)
In fact, taking this dodge into account, the DFW example I gave actually has five parallel runways; the two they label 18/36 are actually the same heading as the three they label 17/35. I wonder if they'll ever need to go to six and add an 18/36C.
Doing that for 4 would need one direction to be A B C D and the other to be B C D E (which is workable, but doesn't seem particularly more elegant than just using left, right, and center [and in fact is worse in some ways, since from one direction, B is the left-most and from the other, it's A])
Then you have airports like ORD which have more than three parallel runways - but pretend some of them are ten degrees off to get more naming space. [1]
[1] https://upload.wikimedia.org/wikipedia/commons/a/aa/FAA_O%27...
My flight instructor once switched off the electricity to the airplane's controls and radios and told me to land. At an airport with a control tower. Using the magnetic compass, airspeed, barometric altimeter, RPMs, that's it. Nothing else works without electric power. There's a whole procedure for doing all that. Being able to navigate safely with an old-school magnetic compass is life-critical.
You just pre-arrange it with the tower and it's no big deal. Probably exceptions for busy class Cs or Ds, but it's no big deal to any airport near me.
Also, in the real world, i would probably break out my smartphone. Short of an emp strike, it will still run on its own battery and can act either as a rough set of instruments or to phone the tower and declare an emergency.
Planes also a normal compass as their ultimate backup instrument, which of course is also magnetic.
Finally, while modern systems could operate off true north instead, not ALL planes have that.
Also, winds (and just about everything else) are in magnetic degrees. If I have wind from 280 while I'm landing, then I immediately know that I have a left quartering headwind and what to expect.
Not always: "If you read it, it's true. If you hear it, it's magnetic."
https://aviation.stackexchange.com/a/62/51758
Actually, it's more like "310ish". Both 305 and 314 are acceptable readings for a runway 31.
Edit: forgot 2 HSIs ;)
So the numbering is a useful backup and discriminator, but the charts have much more detail, and modern navigational and instrument systems on runways obviate much of this for most flights anyways.
As a side note, modern avionics like G1000 auto sync your heading indicator to magnetic. So if we started flying with electronic avionics that are auto synced and have a database of magnetic variation around the world, then it’s feasible to always show true on the heading indicator and have runways and everything else be true bearings. But that’s not the case and I still fly some of the many steam dial planes out there as do others, so we’re stuck with magnetic runways and magnetic everything else used in real-time flight and situational awareness for pilots.
[0] https://www.youtube.com/watch?v=qD6bPNZRRbQ
[1] https://www.youtube.com/watch?v=jfOUVYQnuhw
The difference between them depends on where you are on the globe - it could even be 180 if you happen to stand in the middle of the two!
Magnetic compasses always point to magnetic north, but with earth being molten metals - magnetic north likes to move around a bit, so we work out the difference between the two and apply it to the directions we track.
For example, where I am - the difference is around 11 degrees. In order to head directly west, I can't just let my compass point north and go directly perpendicular port - I have to point north plus 11 degrees, and then go. My actual track will be 281 degrees magnetic, not 270.
What's more interesting is that most navigation charts have all their directions and bearings in "true" north, with the variance also charted, whereas when you actually fly you track the magnetic headings plus/minus the variance. If you go long enough - the variance also changes so you have to take that into account when planning your route.
When you take off / land, runway headings directly corresponding to your heading are a good sanity check - alarm bells go off if you're meant to be flying west but come across Runway 09 in front of you (the opposite!), not to mention we still rely on magnetic compasses - you don't want to be coming into land at a track for 281 degrees and see 27 painted on it; chances are something went wrong somewhere.
Example, San Francisco US, SFO (37°37'08"N 122°22'31"W), to Granada Spain, GRX (37°11'19"N 3°46'38"W) would go north of Newfoundland Canada:
http://www.gcmap.com/mapui?P=SFO-GRX%0D%0A&MS=wls&DU=mi
The 1st place plane took off after the 2nd place plane. This happened because the navigator of the 2nd place flight didn’t take into account the great circle distance and flew straight headings.
(looks at map of Pacific)
Hang on ... where did the earlier ones stop?
For short manually planned tracks (eg VFR planning), most people will plan from one visually helpful waypoint to another, and measure the heading at the mid-point of each route segment. The midpoint approximates the GC heading, and anyway the discrepancy is (very) small on short segments.
For long tracks (eg IFR interstate or intercontinental flights), most plans comply with standard routes, between nominal waypoints, and the route segments together may approximate GC tracks. Between each waypoint, one normally flies a constant heading.
This is imperfect, but variations due wind and traffic are far bigger than losses due not quite conforming to a GC track.
https://www.flightradar24.com/data/flights/ib6171#2ebdae12
Press GPS direct button and don’t bother ;) ;) ;)
If you are in the UK you might be interested in Standup Maths’ video https://youtu.be/HcFvegnQpPo.
Currently true north and magnetic north are broadly the same in the UK - and over the next year or so, there is a point where grid north (Ordnance Survey maps) also coincides with both other “norths”. That point will track north over the UK.
The video shows Matt visiting Cornwall at a place where that point was.
Ref: https://www.sco.wisc.edu/learning-center/magnetic-declinatio...
Note it doesn't work with an adblocker enabled. (I don't think the API to show the maps is free.)/
Related, during the last two weeks or so I've been immersing myself in exploring the topographical maps of Bulgaria the Soviets drew towards the end of the '80s (it was for a personal project of mine). For those interested the map can be found in here [1], by scrolling further down the page, a typical map-sheet looks something like this [2]
[1] http://web.uni-plovdiv.bg/vedrin/index_en.html
[2] http://web.uni-plovdiv.bg/vedrin/details/large/K-35-016-3.jp...
[0] - https://www.ngdc.noaa.gov/geomag/calculators/magcalc.shtml#d...
Flying a great-circle route that intersects the North Pole, and magnetic North Pole, would mean travelling both north and south at the same time for a period.
It's perhaps unfortunate that we live in an era where they are "close enough" that we have one word (North) for two different directions.
Is it possible for the magnetic poles to ever exist far from the geographic poles? (I know the magnetic field could have been the opposite polarity, or not existed at all, but in either of those cases, we still wouldn't have gotten another word.)
The north magnetic pole has been south of 70 degrees as recently as the mid-1800s and the south magnetic pole is currently north of 65 degrees. Every half million years or so the whole thing goes nuts for a while and reverses.
I don’t think the exact mechanics of the magnetic field are super certain, but the general idea is that the rotation sort of pins the poles nearish to the axis but there’s been some significant variation historically.
[0] https://en.wikipedia.org/wiki/Charity_Shoal_crater
[1] https://craterexplorer.ca/charity-shoal/
There might be a very large metal deposit close to the surface (like in Sudbury) that upsets the local magnetic domain.
And those lines won't be even remotely straight =)
We have a good (excellent) idea of the density of the core. We also have an excellent idea of how density changes with pressure. Add in that the core must be magnetic and a fairly common isotope in the universe, and you're pretty much left with Fe.
"Why aviation’s compass is shifting towards True navigation"
https://news.ycombinator.com/item?id=28985745
Summary: "The debate about changing from Magnetic to True navigation is now moving towards how to change, and when, with March 2030 as the proposed date."
Thread with details here: https://twitter.com/lemonodor/status/1508505542423064578
A short summary is that sometimes when ATC radars see an aircraft, if it doesn't seem to have ADS-B (is in a position where no known ADS-B-equipped aircraft is), it will create a synthetic target, broadcasting it out so that aircraft with ADS-B In can see it on their traffic displays. That system is called TIS-B. Somehow the radar in Islip got confused by the magnetic declination: it was generating TIS-B targets that were rotated 13 degrees from where the aircraft really were. And it was doing it for pretty much every aircraft, whether or not it had ADS-B. I assume it wasn't able to correlate the known ADS-B position with the secondary radar returns because they were offset by 13 degrees.
Very weird way for a pretty important piece of ATC infrastructure to fail.
None the less I'd (semi educated) guess that someone at the rader site in question "updated" | "improved" the implementation of The World Magnetic Model [1] in use there and effectively dropped the correct corrections from the end result (for a few days until it was noticed and fixed).
Normally coefficient sets are updated every five years .. and frequently sites and software that use the WMM find themselves with a current IT support person that has not previously encountered a WMM update and then backtrace their way into an "understanding".
This can cause errors until noticed.
Critical infrastructure places (Military, Air Traffic Control, etc) tend to fare better and have longer periods of staff retention, etc. but it can still happen that "we no longer have anyone that was here the last time (five years ago) this happened"
However; given the last epoch update (WMM2020) was released 12/10/2019 (in 'merc date speak MM/DD/YYY (barbarians)) and remains current until expiring on December 31, 2024 we can rule out a simple epoch update FUBAR.
This leaves the other semi regular source of WMM errors - programmers that don't like FORTRAN (esp. super opaque single letter variable 25th degree spherical harmonic functions that date from the 1960s) and feel the need to "do it better".
Having barely survived the prior epoch update and knowing that another will happen again, they carve out some space to build a better mousetrap so that next time will be easier.
This is often a, uhh, interesting intersection of ego and hubris.
[1] https://www.ngdc.noaa.gov/geomag/WMM/
This NavCanada presentation - gives a good overview of the issues in changing.
https://www.navcanada.ca/en/magnetic-north-vs-true-north.pdf