SkyDemon
Important: These forums are for discussions between SkyDemon users. They are not routinely monitored by SkyDemon staff so any urgent issues should be sent directly to our Customer Support.


Great Circles


Great Circles

Author
Message
peter@luthaus.de
Too Much Forum
Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)

Group: Forum Members
Posts: 7, Visits: 15
Hello,
suggested by Tim I post this request in the forum for discussion. I noted (it was already mentioned in the forum) that SkyDemon uses generally rhumb lines to connect waypoints. I find this unsatisfactory and annoying. Industry standard and the wanted way is great circles. There are many reasons why great circles should be used (compared to none using rhumblines):
1) Great circle is the standard for GPS navigation, so built in certified GPS use it and Skydemon is not consistent with the track actually flown even producing off track warnings.
2) For planning the total distance is not so different for short legs, but the off course deviation can be quite large. E.g planning from EDLW to EPWR the rhumb line goes through EDDP Delta, the great circle passes it completely.
3) Most pilots use Lambert projection charts approximating the earth's convergency for planning. So the straight lines drawn on paper will be close to great circles and it is unexpected that skydemon shows a different route.
4) Radials are great circles
5) Airways are great circles
As I find the rhumb line really a nuisance and also other pilots I know do so I suggest changing to great circles. I hope this could be implemented soon.
Peter


T67M
Too Much Forum
Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)

Group: Forum Members
Posts: 166, Visits: 3K
This has been discussed at length in these forums before, and equally good arguments put forward for rhumb lines - the primary one being that SkyDemon is a VFR-only tool aimed primarily at pilots of SEP aircraft and helicopters flying relatively short routes. A rhumb line is defined by a single heading which is close to the heading a PPL will measure from a Lambert Conic Conformal chart (if they remember to use the middle of the line!!!) whereas a great circle route has a constantly changing heading and is an unnecessary complication for most of the pilots Skydemon is aimed at.

Previous discussions here have suggested that the choice of rhumb line vs great circle should be a configuration option in SkyDemon, although the default must be rhumb lines for consistency with the current interface. At some point Tim and the team may choose to implement this - if they can find a sensible way to present the constantly changing heading on a printed PLog. I did suggest that the PLog should show the initial and final headings, but I haven't worked through all the corner cases to see if this works and is sufficient. 

Incidentally, a straight line on a Lambert Conic Conformal projection is not actually a great circle route, although it is a good approximation for short routes. The error between a rhumb line and a great circle route is also insignificant within the DOC of most radio nav beacons, again because of the short distances involved.
peter@luthaus.de
Too Much Forum
Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)Too Much Forum (278 reputation)

Group: Forum Members
Posts: 7, Visits: 15
T67M - 7/27/2016 8:26:22 PM

I am very annoyed from the deviations even on relative short routes. I read through the posts and could not find a single argument favouring rhumb lines, also in this reply. I could also comment in detail:

"This has been discussed at length in these forums before, and equally good arguments put forward for rhumb lines - the primary one being that SkyDemon is a VFR-only tool aimed primarily at pilots of SEP aircraft and helicopters flying relatively short routes."
So it shouldn't matter to them, because they don't note a difference. Rhumb line or great circle is not flight rule dependent, I prefer to fly a great circle as well flying VFR, also my built in GPS doesn't note the difference in flight rules.

"A rhumb line is defined by a single heading which is close to the heading a PPL will measure from a Lambert Conic Conformal chart (if they remember to use the middle of the line!!!) whereas a great circle route has a constantly changing heading and is an unnecessary complication for most of the pilots Skydemon is aimed at. "
Also on a rumb line the heading is constantly changing (even the true one) because the wind will change along the route. Actually the course is constant for a rhumb line. And as already mentioned, there is no complication because people flying only VERY short distances will not notice it all, whilst people flying only moderate distances in order 100s NM east/west direction at out latitudes will be annoyed by rhumb lines and not by great circles.

"Previous discussions here have suggested that the choice of rhumb line vs great circle should be a configuration option in SkyDemon, although the default must be rhumb lines for consistency with the current interface."
I even see no point for a configuration option because great circle would make everybody happy. For the PLOG I would not care to leave it as it is, the mean course and heading would be fine. Of course configuration would be nice to show the differences in lessons.

"Incidentally, a straight line on a Lambert Conic Conformal projection is not actually a great circle route, although it is a good approximation for short routes."
The great circle is a very good approximation compared with the rhumb line, you will use charts where the cone constant represents approximately the earth's convergency. So why using a rhumb line that is a much worse approximation when you can have the better great circle?

" The error between a rhumb line and a great circle route is also insignificant within the DOC of most radio nav beacons, again because of the short distances involved."
Why do you want to be inaccurate when you can be accurate?

The last points you mentioned I can summarise you also think great circle is better but you accept the inaccuracy. I wouldn't if I had the choice. I really see no complication at all for a user not understanding the difference, the opposite: They will not be surprised that the built in GPS shows a different route.


T67M
Too Much Forum
Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)

Group: Forum Members
Posts: 166, Visits: 3K
For the PLOG I would not care to leave it as it is, the mean course and heading would be fine. 


You can't have it both ways! The "mean course and heading" of a great circle route is the rhumb line to all intents and purposes. Are you now arguing that the PLog and the magenta line should be different? I see airspace busts ahead using that approach!

There are two scenarios here:
  1. The route leg being flown is short enough that there is no significant difference between the two methods. By definition the same heading will be used for both rhumb line and great circle navigation. This represents a significant portion of legs flown by people using SkyDemon.
  2. The route leg is long enough that the differences are noticeable. By definition, the great circle route must have different initial and final headings (if it didn't, it would be a rhumb line!)
Tens of thousands of pilots (i.e. the current users of SkyDemon) seem to be happy to use the single heading rhumb line offered by SkyDemon and as taught during PPL navigation training. Most of them won't even notice the difference, and those who notice mostly don't care provided that the magenta line shown on their SkyDemon device represents the heading shown on their PLog. I suspect that the majority of these pilots would be confused if SkyDemon suddenly offered them two (or more) headings for each leg. 

There are a small number of pilots would prefer the mathematical perfection and/or want to benefit from the fuel/time savings (typically less than 5%) of a great circle route. In return they accept complexity and higher workload of flying a constantly changing heading. I have no objection to them choosing to do so if they wish, but I would strongly object if this small number of pilots caused confusion to the tens of thousands of people who are happy with the simpler (albeit mathematically imperfect) solution of flying a rhumb line as they were taught during their PPL training. That is why great circle navigation has to be a configuration option - if it is ever supported at all.

At the end of the day, SkyDemon is Tim's product, and he will implement the features he feels satisfy the needs of the majority of his customers. Thus far he has been incredibly successful - thanks Tim!


Lars-Vater
Too Much Forum
Too Much Forum (101 reputation)Too Much Forum (101 reputation)Too Much Forum (101 reputation)Too Much Forum (101 reputation)Too Much Forum (101 reputation)Too Much Forum (101 reputation)Too Much Forum (101 reputation)Too Much Forum (101 reputation)Too Much Forum (101 reputation)

Group: Forum Members
Posts: 1, Visits: 0
Hey there,
i’ve been following this discussion quite a while. I can’t see the point in not willing to program it the right way. Every flight we do is a great circle. On small distances you might not see a difference. Flying from here to a place 100NM or further you’ll notice a difference. I cant get your point of view, saying it is a VFR only tool. How does it make a difference depending on the rules you fly? The problem is still the same isn’t it?
Talking to other pilots and friends, especially those who fly all over Europe, would appreciate if Skydemon will give an update, cause it is not a feature, it is a bug! Looking at Jeppesen and alle the other tools you’ll notice, that they all use great circles, cause thats what you fly out there. I really do get Peter's point of view and share it.
Please fix this bug, i really appreciate.
Safe flying aviators,
Lars
Tim Dawson
SkyDemon Team
SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)

Group: Moderators
Posts: 5.3K, Visits: 3.8K
I will add that since the vast majority of SkyDemon subscribers appear to be happy with the current method, we would need a mightily compelling reason to switch at this stage. The current method gives the pilot a heading that, if he/she flies it, will get him/her from A to B (subject to wind). That simplicity is worth a lot, and that is what the notion of switching to great circles per leg is competing with. If you're navigating using SkyDemon as your GPS (which a huge percentage of our customers are) it doesn't matter, because the navigation instruments will also be helping you along the rhumb line.

Internally, SkyDemon uses both rhumb lines and great circles all over the place. In a very early beta, each leg in a route could be comprised of either. We had to ditch one of them because the choice was too confusing. We ditched great circles because a fundamental requirement of SkyDemon has always been to produce a PLOG with a heading to fly for a leg. You can't do that with great circles.

I'm open to re-introducing great circles into the mix, but not on a per-leg basis. I can imagine, for example, that if somebody planned an exceptionally long leg like the one shown above, where flying a great circle would give an appreciable fuel saving or where the midpoint different from the rhumb line midpoint to the great circle midpoint exceeded a given distance, we would ALSO draw the great circle on the map. This would give a visual clue that the leg could be "broken up" by the insertion of more turning points to create more rhumb lines. Or we could draw the great circle between takeoff and landing airfields for instance.
JulianScarfe
Too Much Forum
Too Much Forum (638 reputation)Too Much Forum (638 reputation)Too Much Forum (638 reputation)Too Much Forum (638 reputation)Too Much Forum (638 reputation)Too Much Forum (638 reputation)Too Much Forum (638 reputation)Too Much Forum (638 reputation)Too Much Forum (638 reputation)

Group: Forum Members
Posts: 12, Visits: 19
Tim Dawson - 8/5/2016 11:34:21 AM
I'm open to re-introducing great circles into the mix, but not on a per-leg basis. I can imagine, for example, that if somebody planned an exceptionally long leg like the one shown above, where flying a great circle would give an appreciable fuel saving or where the midpoint different from the rhumb line midpoint to the great circle midpoint exceeded a given distance, we would ALSO draw the great circle on the map. This would give a visual clue that the leg could be "broken up" by the insertion of more turning points to create more rhumb lines. Or we could draw the great circle between takeoff and landing airfields for instance.

Had you considered changing the map projection at smaller scales, or is that just too hard?
Tim Dawson
SkyDemon Team
SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)

Group: Moderators
Posts: 5.3K, Visits: 3.8K
The map projection doesn't have anything to do with whether legs are rhumb lines or great circles internally.
Tim Dawson
SkyDemon Team
SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)SkyDemon Team (345K reputation)

Group: Moderators
Posts: 5.3K, Visits: 3.8K
So this morning I set out with intentions of prototyping a feature where the GC line between your takeoff and landing airfields is drawn faintly in the background, if the GC is more than 5% more efficient than the rhumb line. It didn't work, because although I created routes spanning the whole of Europe, I couldn't find any where there was that much difference. In fact I barely managed a 1% difference.
T67M
Too Much Forum
Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)Too Much Forum (10K reputation)

Group: Forum Members
Posts: 166, Visits: 3K
Tim Dawson - 8/11/2016 10:10:02 AM
So this morning I set out with intentions of prototyping a feature where the GC line between your takeoff and landing airfields is drawn faintly in the background, if the GC is more than 5% more efficient than the rhumb line. It didn't work, because although I created routes spanning the whole of Europe, I couldn't find any where there was that much difference. In fact I barely managed a 1% difference.

While I agree that for all practical purposes, the difference between GC and RL is irrelevant for most users of SkyDemon, I think the original query was more concerned about the difference in track over the ground than the difference in track length. What does the result look like if the threshold is set to a cross-track difference of more than 1nm (25% of a small ATZ) at the mid-point of the leg? (And I wish I was clever enough to work this out myself!)
GO
Merge Selected
Merge into selected topic...



Merge into merge target...



Merge into a specific topic ID...





Reading This Topic


Login
Existing Account
Username:


Password:


Select a Forum....





SkyDemon Forums


Search