I imagine that there are many flights that cross the Atlantic outside the NATs to optimise their routing. My question is the following: are the fixes used along these flexible routes "rounded coordinates" such as 4010N, 4015N, 4520N, etc....

OR

... are the routes perfectly great circle?

I hope I made myself clear...

Thanks!

"...are the fixes used along these flexible routes "rounded coordinates" such as 4010N, 4015N, 4520N, etc...."

Generally, yes. It is very difficult if not impossible to get a random track within the NAT assigned airspace on that given day in the direction the tracks are active. Going the "other way" presents fewer problems but you will still file via coordinates as you've suggested. It is very easy then to create virtually a great circle track. Wind could be a factor though and, for that length of flight, a track designed to maximize the "pressure pattern" advantage may be more suitable. Flight planning companies have computers which can generate Best Time Tracks or Best Fuel Tracks. Your company's preference will decide which is used.

Hope that answers your question!

Also, it is not unheard of for the southernmost NAT track not to be paralell to the rest, but, rather, angle down the middle of the Atlantic toward Latin America. This might further restrict random routes crossing it.

Quoting Mozart (Thread starter): outside the NATs

Quoting Yikes! (Reply 1): within the NAT

Aren't we talking about everywhere trans-Atlantic EXCEPT the NATS system? Say a private flight from the Canaries to Miami which should be well below the southernmost tracks.

Geoff M.

Okay, the short answer to your question is I don't know.

However, I do have a couple of thoughts on it.

One question to be answered for ATC purposes is where you are going to be at any given moment. Let us imagine a flight from Madrid Spain to Bogota Colombia. If the route is great circle from a coasting-out point to a coasting-in point, let us say from KOMUT off the coast of Portugal to Barbados - all pretty much bluewater flying you are still entering and leaving the airspace which is the responsibility of sovereign nations along the way.

I don't have a proper chart but you will be departing airspace controlled by Lisbon and transiting airspace controlled by Santa Maria in the Azores. From their airspace you will enter others. These control agencies will require exactly when and where you will enter and exit their airspace.

Problem is, you and they are looking at a grid-chart of an ocean which is really round. You plot a great circle on that chart and it appears as a curved line. Ultimately you cannot superimpose a curve on a grid. In my example it is simple, you are entering Santa Maria airspace at KOMUT which is N38 W15 but suppose you went great circle at the coastline. Now you miss KOMUT and enter Santa Maria at 15West, but at what latitude? Maybe 38^o 07' north or some such odd number? They might not accept that.

For these purposes you might want to acknowledge great circle, but find lat/long coordinates that are very near your route and flight plan point-to-point from one of those to the next.

Another consideration is this: Does your navigation system (FMS, GPS etc.) find great circle or rhumbline between waypoints? I've taught the system at one airline and not found the certain answer for that one.

Again, and in the absence of any better information, I think I'd plot point-to-point along an approximation of a great circle. The difference crossing the Atlantic, even on a diagonal shouldn't be more than a couple of minutes.

Now we can both sit back and wait for a responder who actually knows how it works. 

Quoting SlamClick (Reply 4): Another consideration is this: Does your navigation system (FMS, GPS etc.) find great circle or rhumbline between waypoints? I've taught the system at one airline and not found the certain answer for that one.

They all use (744/A320/757/DELCO/Litton) use great circle navigation between waypoints.

To answer the general question, no, they're not great circle waypoints. If you were to fly a random track by just entering the entry and exit point you would have one set of coordinates. However, ATC would not accept those coordinates as valid waypoints. You can approximate a great circle route by just entering valid waypoints close to the great circle route you obtain. There are several airlines NW being one of them, that during the winter it's not uncommon for them to operate a "random" track far north of the normal NATS. The route flown isn't at a constant latitude but at a varying one with the farthest north being about in the middle of the oceanic routing.

On "great circle" vs rhumb line--

Let's say you're at 50N 00W and you tell the black box to fly you to 50N 179W. "Great circle" would head almost over the pole and would cover 4817.4 nm; the rhumb line would head west along the 50th parallel and would cover 6968.3 nm. Surely the black box knows to fly north and not west?

Quoting Timz (Reply 6): Let's say you're at 50N 00W and you tell the black box to fly you to 50N 179W. "Great circle" would head almost over the pole and would cover 4817.4 nm; the rhumb line would head west along the 50th parallel and would cover 6968.3 nm. Surely the black box knows to fly north and not west?

Okay so far as it goes, but not a realistic example. In the first place, absolutely nobody ever plans a flight halfway around the world without a single enroute waypoint. In the second place, for the flight you describe I would bet that the initial heading would be easterly. Why? To take advantage of the prevailing westerlies in those latitudes. Flight plans are optimized for "nautical air miles" which includes the wind factor.

The average segment of the computer generated flight plans I've seen (even for flights of ten hours or more) is maybe 300-350 nautical miles. Many segments might only be twenty to fifty. At such distances rhumbline will equal great circle to the mile. Furthermore, published airways are rhumbline, magetic.

88N 00W to 88N 179W: 241.23 nm "great circle", 376.76 nm via rhumb line.

"Furthermore, published airways are rhumbline, magetic."

You've probably noticed the airway radials at adjacent VORs are often a degree or three away from reciprocal? True, that could be due to the change in the magnetic deviation-- I'll see if I can find a counterexample.

[Edited 2005-07-24 01:59:16]

Quoting Timz (Reply 8): True, that could be due to the change in the magnetic deviation-- I'll see if I can find a counterexample.

I've always assumed that it was due to small differences in variation (in the US, deviation means something else) but the isogonic lines on the charts are a full degree of spread and so, anomalies of less than a degree may exist between them. Hard to guess.

Anyway, I've played with the FMC at the gate and plotted reverse course between waypoints and gotten reverse headings that were quite a bit different from the natural reciprocals. I don't think it does the math the same way every time.

Looks like we can forget about airways being rhumb lines. Couple examples:

J82 runs FSD to RAP with no apparent kinks; at about the midway point is VIVID. Airnav gives the lat-lons for all, and also says VIVID is on the 267.29 radial of FSD. At FSD, true "great-circle" direction to VIVID (based on the lat-lons) comes out 276.293 degrees and to RAP 276.293 degrees. Rhumb line FSD to VIVID is 275.20 degrees.

Airnav gives the 1965 magnetic variation for FSD: magnetic north is 9 degrees E of true north. I guess they set VOR-north 9 degrees east of true north in 1965, and VOR-north doesn't change since then, even tho magnetic variation does?

Eastward from GTF, J203-J204 run to HILGR and split there, to MLS and BIL. Airnav says HILGR is on the GTF 076.54 radial; true "great-circle" to HILGR at GTF is 92.5344 degrees, so they didn't quite get that one right-- 1990 mag variation 16 E.

It says HILGR is on the BIL 333.00 radial; true BIL to HILGR is 347.004, with the expected 14 E variation (in 1990 again).

It says HILGR is on the MLS 280.18 radial; true is 295.182, mag var 15 E (1965).

So far I've been too lazy to compute the J203-204 rhumb courses. I did another one, tho: from DLH to SSM J140 shows no kinks. I'm too cheap to get a current chart, but it's probably still the 086 radial of DLH (var 05 E) and the 281 radial of SSM (var 04 W). True "great circle" directions are 91.230 deg at DLH (checks) and 276.230 deg at SSM (doesn't quite check, but better than the rhumb line). Rhumb course comes out 94.118/274.118, so the rhumb line reaches a maximum of 8 nm from the direct line.

Okay, I think I need to clarify my question again:

Let's assume a plane goes from Paris to Caracas. On that day, there are no NATS routes that make sense for such a flight.

Let's then assume that the flight plan is one that leaves the European airway system at some FIX west of Britanny, say "KOLEK". It re-enters the airway system somewhere around Barbados, say at NDB "BGI".

Now, what interests me is what happens between those two points: the airline will of course try to fly as close as possible to Great Circle, and since this is a random route, it has the choice. I know that they plan for segments of something between 300-400 nm each. But what I don't know: are those segments linking "obvious" fixes that are close to the GC route, such as "1020N", "1030N", and so on

OR

will the flightplan generate "perfect GC fixes", i.e. fixes that are perfectly on the GC line, and that could have coordinates like "1217N" and "1632N" and so on.

In other words:

- is the route defined by linking such "obvious" fixes
- or are the fixes defined by what is the perfect GC?

Thanks

Quoting Mozart (Reply 11): Now, what interests me is what happens between those two points: the airline will of course try to fly as close as possible to Great Circle,

First things first, as others have said, the airline will NOT of course try to fly as close as possible to Great Circle -- they'll want to fly as close as possible to the minimum air-distance with winds included.

That said, it doesn't change the nature of your question, since airlines will indeed want to fly a "random" route over open water. I think your question is, "given that an airline can file a completely random route, is it really 'fly wherever they want' or is it limited to some structure of points they can file to ATC to describe their route?'"

Currently most, if not all, such plans are filed with points that are every 10 degrees of longitude and whole degrees of latitude. While the NATS sometimes include a point that's not a whole degree of latitude, I've never seen a random route point that's not whole degree latitude.

Hope that answers your question.

It does answer the question! THanks a lot!