Pole overflight with GPS and IRS

[Hardy Heinlin]

Have you guys got any experience with flights over a geographic pole? GPS is still not reliable within circa 100 nm of a pole, is it?

Even if it's necessary to ignore GPS in that very high polar latitudes, why does the FMC apply split IRS opereration (IRU L + R) there instead of proceeding with the usual mixed IRS mode (3 IRUs) including the previously accumulated bias related to mixed IRS position?

Or are GPS and even mixed IRS mode reliable up there, but there's just a risk of mathematical errors, and so the system uses the most trivial position calculation during the pole crossing?


|-|


Edit HH: Thread split from http://aerowinx.com/board/index.php?topic=5278.msg56825#msg56825

[Jeroen Hoppenbrouwers]

I think GPS itself is as reliable over the poles as anywhere else, as the satellites are in orbits that cover the whole globe and they don't care about the coordinate system deployed down there. However there definitely are mathematical challenges representing near-polar coordinates. It's more a presentation problem than a fix problem.

https://en.wikipedia.org/wiki/Global_Positioning_System#Space_segment

https://www.quora.com/Does-GPS-work-at-the-North-Pole


Hoppie

[Hardy Heinlin]

But what special math needs the FMC to do over the pole if GPS remains reliable? Maybe the FMC ignores GPS for safety in case it fails during the pole crossing and the IRS is then required. And when the IRS is used in that area, it perhaps should use the split mode (because of mathematical risks) -- like it did in the early years before GPS was installed on the 744.


|-|ardy

[John H Watson]

I think GPS itself is as reliable over the poles as anywhere else, as the satellites are in orbits that cover the whole globe and they don't care about the coordinate system deployed down there. However there definitely are mathematical challenges representing near-polar coordinates. It's more a presentation problem than a fix problem.

If you keep following the links on those pages, you will see this one.

https://www.allaboutcircuits.com/news/navigating-the-arctic-why-gps-might-fail-you/?utm_campaign=4ffaa9206e-EMAIL_CAMPAIGN_2018_07_25_04_57_COPY_01&utm_medium=email&utm_source=All+About+Circuits+Members&utm_term=0_2565529c4b-4ffaa9206e-272887169/

There definitely seems to be a problem at the poles in one way or another. The aircraft should know if the GPS is faulty from the signals sent, so will switch to split IRS as required. The GPS is not just sending position data to the FMC, but HFOM, HIL, satellite coverage data, fault data, etc.

[Jeroen Hoppenbrouwers]

THAT was enlightening. Thanks!

I had been biased with the Iridium constellation, which has very near polar orbits and actually works better over the Poles than anywhere else -- so well that they can switch off most of the satellites when near the Pole as there are plenty left.


Hoppie

[Mariano]

"Even if it's necessary to ignore GPS in that very high polar latitudes, why does the FMC apply split IRS opereration (IRU L + R) there instead of proceeding with the usual mixed IRS mode (3 IRUs) including the previously accumulated bias related to mixed IRS position?"

I've wondered about this for many years. Could it be due to heading reference coming from IRUs? If both FMCs are using triple mix IRU signals, as the geographic pole is crossed, one or two IRUs could be (for a moment) feeding almost opposite heading signals to the FMCs than the third IRU might. Averaging slightly different positions (coordinates) is a lot easier than averaging opposite or almost opposite heading signals (which make no practical navigation sense for a using system).

Having an IRU feed its respective side FMC eliminates this issue (one FMC might be "heading" North and the other "heading" South, causing no issues other than temporarily differing ND displays (and RDMIs), wind and TAS calculations, drift angle, etc.).

Just my (probably very dumb) non-expert guess.

Best regards,

Mariano

[Hardy Heinlin]

That's a good theory, I think. It makes sense. There is no "average heading" at all, as far as I remember. The FMS incorporates a complex logic of automatic heading source selection for the left and right EFIS (modeled in PSX). Heading is always from a single source, never from a mix.

There are some demo situations in PSX that show the polar heading disagreement, GPS inhibit, split IRS etc.:

Training 001 - Starting polar navigation mode.situ
Training 002 - Overflying North Pole.situ
Training 003 - Terminating polar navigation mode.situ


Regards,

|-|ardy

[emerydc8]

If you keep following the links on those pages, you will see this one.

Thanks for the link, John. I've always wonder why they take the GPS out of the picture at the pole.

Doesn't the AFDS also maintain wings level when crossing over the pole? Maybe that's to prevent what Mariano brought up about there being such a small difference between north and south.

Jon

[Hardy Heinlin]

Doesn't the AFDS also maintain wings level when crossing over the pole?

I haven't found any hint on wings level in the manuals. Maybe you're thinking of the HDG SEL mode autochanging to HDG HOLD when the MAG/TRU reference changes?


|-|ardy

[emerydc8]

I think you're right.

[Phil Bunch]

An interesting subject, conceptually, scientifically, and mathematically...but since we have no absolute location for *anything* in the physical universe where we live, what can we do that is permanent and credible???  The material of the earth itself is in constant motion and is not stable, for example.  Nothing outside of the earth seems to have an accurate, absolutely defined position.  This problem makes developing high-precision time and length international standards seem simple, IMO.

I would think reverting to IRS-based navigation would be a good thing to do since presumably it can take you back to your home airport or any other airports.  Does an IRS system assume anything about having a spherical earth or make other simplified assumptions like that???  When they UK constructed a physical longitude reference line in Greenwich, they had the right idea!

--------------

<Attempted humor:  I recommend that we take all necessary steps to make the earth a perfect sphere and  keep it that way.  While we are at it, we might as well make its magnetic poles be stable and to agree with whatever definition of the poles is preferred.  Then all we need to do is to construct a physical object in Greenwich that defines the zero locations for both longitude and latitude!>

[Hardy Heinlin]

...but since we have no absolute location for *anything* in the physical universe where we live, what can we do that is permanent and credible???  The material of the earth itself is in constant motion and is not stable, for example.

I'll pass the question on to Torrence, the NASA expert ... My humble assumption is: We can at least consider the earth's rotational motion around its own axis to be a stable self-reference of its rotational axis. -- It is what it does, so to speak. -- I think, fluctuations of the rotation rate don't matter as long as the rotation rate is non-zero. -- The pole axis exists because there is a rotation. -- If there is no rotation, we could declare any point a pole, just like we put the 0° longitude on Greenwich. -- The geographic pole is at that point where an inertial gyro doesn't draw a track but just rotates. At that point we are clearly at a 90° latitude. The equator is where an inertial gyro draws a straight lateral track and doesn't rotate at all. The latitudes in between can be calculated by sine/cosine functions. This is how the IRS determines its latitude position on the planet during initial NAV alignment. It senses the rotational motion and tracking motion for some minutes, and then, based on the sensed values, calculates its local latitude.


Regards,

|-|ardy

[Jeroen Hoppenbrouwers]

Although small, you will need to take the rotation of the Earth around the Sun into account. I believe that the rotation of the Sun around the center of the Galaxy and the rotation (?) of the Galaxy around something else is too small to be relevant.

IRS is absolute to the actual motion in space, referenced to stars far far away and long, long ago. That's why you can use it to fly to the Moon.

Hoppie

[Hardy Heinlin]

The reference system for the attitude is not the earth's surface, but the outer space. Agreed. You need a weight to add the earth's surface as a platform reference to the attitude indicator in the long run.

But the platform reference for the latitude detection is the surface platform of the planet, not the platform of the solar system. On the other hand, you first need a stable platform for the latitude sensor (same for platform heading and track). So if the platform isn't mounted on the ground but flying on an aircraft, you need an attitude stabilization system (gyros, weights) to keep the platform parallel to the planet's surface, and so the attitude system is -- indirectly -- a factor again. Edit: Fixing the platform on the ground is of no help; the earth's pole axis is tilted relative to the earth's ellipse plane around the sun, so there's still an unknown disturbing factor.

As for the rest: It's all relative :-) It all depends on the observer's location and time ...

Nothing is absolute, Isaac.


|-|

[Phil Bunch]

There is an interesting discussion of GPS errors and their correction at Wikipedia:

https://en.wikipedia.org/wiki/Error_analysis_for_the_Global_Positioning_System

I haven't reviewed this or related articles recently but from my failing memory I think a lot of the corrections that produce good GPS accuracy depend on having extremely accurate clock(s) in the satellites.  Also, Einstein's General Relativity is needed to correct for the effects of the (rotating) mass of the earth on time:

https://en.wikipedia.org/wiki/Error_analysis_for_the_Global_Positioning_System#Special_and_general_relativity

An excerpt:

"The effect of gravitational frequency shift on the GPS due to general relativity is that a clock closer to a massive object will be slower than a clock farther away. Applied to the GPS, the receivers are much closer to Earth than the satellites, causing the GPS clocks to be faster by a factor of 5×10^(−10), or about 45.9 μs/day. This gravitational frequency shift is noticeable."

-------------
Do airliners still carry a sextant and a pilot who is well-practiced in using it for navigation?  I think airliners still have an observing port for using sextants...or is this only for older airliners?  What do celestial navigators do if their airliner is only flying in the daylight and they need to do some celestial navigation?

[Jeroen Hoppenbrouwers]

Do airliners still carry a sextant and a pilot who is well-practiced in using it for navigation?

I think airliners still have an observing port for using sextants...or is this only for older airliners?

What do celestial navigators do if their airliner is only flying in the daylight and they need to do some celestial navigation?

Nope.

Only very old ones... 707... I believe 747 Classics?

No celestial navigators any more  :-/

However you still have diurnal nav hobbyists:
https://en.wikipedia.org/wiki/Cessna_188_Pacific_rescue


Hoppie

[Jeroen D]

Very few people know how to use a sextant these days. Being a former merchant marine engineer I have used them at the start of my career. We are talking 1979. No GPS in those days. Heck the most sophisticated piece of electronics on board was a very simple radar. And we still had a radio officer, sparks, using morse to communicate.

Whereas navigation, and thus using a sextant, was really the area of the navigation officers, I learned from them. In fact a few weeks ago I took a refresher course in astro navigation. Purely for fun. Two days brushing up on theory, one day aboard a small yacht practicing, shooting the sun.

It is usually yachtsmen that are still interested in astro navigation and have sextants. Purely for fun. But some of these guys sail around the word solo and still see it as the most reliable back up to all electronic and battery powered stuff. Our teacher similar story. 10 years of solo sailing. Two days into a crossing of the Atlantic, his generator packs up. So back to old school astro navigation.

It is good fun. History of astro navigation is hugely interesting too!

Jeroen

[Will]

There's a chapter in Ernie Gann's memoir Fate is the Hunter in which he describes using celestial navigation while flying an old American Airlines airplane. I forget which type. That would have been in the 1950's, I'm guessing.

Meanwhile, I eagerly await Torrence's input, but it seems to me that the center of mass of the earth would be easy to find, would it not? Would that be a good reference to use for GPS corrections?

[Hardy Heinlin]

I think the planet's rotational axis intersects the center of the planet's mass. I assume if you want to detect that center point by sensing gravity forces, you need to do this at multiple very distant points in space to get 3D data. If you do it by gyro systems, you can detect it within a small local area in a few minutes. But the problem is not only the planet's additional motion around the sun; there's also our moon that disturbs the planet's axis location in space. The common center point of the rotation pair "Earth & Moon" is not at the center of the earth but just some hundred miles below the earth's surface. It's eccentric ...


!-|ardy

[torrence]

Don't have time to address all of the above issues right now.  Quick answer is that if all you need is the old 'chronometer' standard: ~ 1 nm longitude error in a month of sailing, you can do that with a chronometer and a sextant.  Plus you could use copies of the Nautical Almanac and Bowditch.  If GPS goes off line globally while you're in the air - ?
Cheers
Torrence

Added comment - the old Polynesian navigators could do much of this with just wind/wave and stars. 
T