Lucky navigation

I’m plodding along, getting to grips with learning to use celestial navigation. I did a test flight last night where I was too tired to complete it, so I went to bed leaving the sim running…

The route was from Adak Island in the Aleutians down to Midway Island. That is over 1500nm of nothing but ocean. I ran a SkyVector plan and set off in daylight. I had traveled over 550nm by the time it was dark enough to see stars for a position fix. When I took the fix, I estimated I was about 19nm to the right of course, but because it was a slow process to crunch the numbers, I estimated I was about 100nm farther on by the time I had completed the plotting, so I extended the course line another 100nm, measured the distance from the planned track, which was now 23nm off course. I then applied the 1 in 60 rule to figure out a course correction, which turned out to be only 4 degrees. I made the heading change, and went to bed… When I woke up, I had flown well beyond Midway, but the track shown on the Goodway map shows I flew right over the runway! I’d rather be lucky than good!

Planned route is the green line. The yellow line shows a direct line from Adak to the position fix, which was then extended by another 100nm to account for the time it took to work out the fix and plot it.

Some hours later, the airplane passed directly over the field.

The orange line is the track flown. The blue line is point to point from Adak to Midway, but unlike Google Earth, Goodway uses a flat map, which means a great circle route ends up being drawn as a curve.

Anyway, I thought it was neat… lots of luck involved when it comes to actually flying over the island. I would have been happy to have been within visible range of it.

My ride for the exercise…


That is seriously cool!
On a related side note.
My old Nav teacher at flightschool was a retired navigator with SAS. He flew on DC-4/6/7. Having been replaced by a INS gyro, he had a natural scepticism towards technology. He used Sine/cosine tables instead of a pocket calculator, for instance.
He gave us an introduction to astro navigation and I remember thinking that I was glad it was not in the curriculum…!


Very nice, and the better harder you work, the better luck you have!

There is certainly an element of truth to “you make your own luck”, at least in part. You can at least make it more likely that you succeed. There is always that random element of luck though. Chaos theory and all that :wink: .


It’s not too bad once you get a grasp of the basics. For the project I am working on with Casper (@TheAlmightySnark), the aim is to make the process as basic as possible, while still using the fundamental principles. For example, using Google Earth to plot the ground points of stars is a lot easier than having to use a universal plotting chart and look up tables. We also decided to eliminate the need for altitude correction, by making the view in Stellarium to be always at sea level. It is easier to find resources for Celestial Navigation aimed at sailors, so it seems to be a logical compromise. There is still plenty of work involved in putting a position fix together.

I also learned the challenge of getting the workflow completed in a timely manner. In the example above, it took me 30 minutes to take the sightings, work out the positions of the ground points of the 4 stars I used, and the distance I was from each ground point, then plotting these points and drawing range rings around the points with the calculated radius. In a boat, you might have moved a few miles in 30 minutes, but if you are travelling at 200 knots ground speed, you are 100nm away from the fix by the time you have finished with plotting it. It’s no wonder that a navigator was such an important member of the crew. I don’t think you could fly the airplane as a pilot and also do navigation this way.


One of our Navy pilots loves the saying “aim small, miss small”… :+1:

Nice work @PaulRix!


As of 1985, US Midshipmen (NROTC and USNA) were required to learn celestial aviation.

The key is to make the sightings as quickly as possible. You mark the star, read/ record the sextant and record the chronometer stopwatch (paused when you said “mark”, then goes 2 x speed to catch up again), then do the second and third reading. Then go back, do the look ups in the tables and do the calculations. You (should) come out with a nice triangle where you should have been inside of when you took your readings. As mentioned, if you are going 10-12 knots, you may still be inside that triangle. :grin:

Our course today everything is GPS, however, there is still one reflection to celesteral nav still in common use today. There is a term in the Navy called “Gundecking”. Back in the days of sail, the midshipmen aboard a Man of War were required to train on celestial navigation. To cheat, one of them would sneak a look into the navigation log to find the “answer”. Then they would head down to some out of the way place on the gun deck and work the equations backwards to get the sextant angles. At training time they would just call off the already calculated sextant angles.

Nowadays, any type of improper short cut–aka “pencil whipping”–is known as gundecking.


Thanks for the info Will. Do you normally take sightings on just three stars or is more preferable? Dropping down to just 3 would certainly save time, but it would be harder to detect a bad sighting/mistake in the tables.

I have to say it has been a fascinating learning experience so far.


We were taught 3 stars. As you mentioned, not as much of a rush on a ship. When you plot the three bearings out, it is pretty obvious if you messed something up and you haven’t moved that far is you need to do it all again.

I guess you could take a couple more and only work out the calculations if one of the first three is obviously in error.

Your point about having a navigator in addition to a pilot is a good one. Even in an older aircraft doing 120 knots, you are in a rush to get the calculations done before you are too far from the sighting point. Plus, just the distance flown in between the three sightings becomes a factor.

On the plus side, in relatively smooth air at altitude, taking the sightings can be easier than doing so from the pitching deck of a ship.

BTW, this is really cool stuff. :grin:


Sounds like a good plan. I’ll take 4 sightings but only reduce 3 of them unless it looks off.


That is very neat Paul! Glad to see it’s working!

For anyone else interested, it’s not quite ready for distribution yet, but I am open to having a few more people trying this out.

It’s a simple system, a small Python 3.7 application runs in the background and talks to X-plane, kindly asking it to send position data over UDP. It then formats this UDP data into a TCP packet and asks Stellarium(With the Remote Control Plugin) to change the position.

That’s all there is to it! I build the interface in Kivy(2 textboxes, a button and 2 labels) though recently I’ve been messing with Godot 3 for a bit and I’ve got half a mind to rebuild it in a game engine, multiplatform, supports networking and interface elements out of the box. Mostly because building this was easy, distributing it is a bit hard, right now the application is about 50 mb’s in size, which is a bit insane for 100 lines(give or take) in code and a few small buttons.

Though the most likely option is, is that I am quite rubbish at creating a proper distributable file! :wink:


This is some next level stuff, I love it!

I can see this rapidly becoming the “Xmas trip - hardcore mode” option for those who can do it :sweat_smile:

1 Like

My first airliner (DC-10) had a little window for taking star shots. I can’t recall if it was still usable. For some reason I never took a peak. When I was learning to fly, there was a PanAm 747 Captain who would stop by the FBO from time to time. His name was Jim Smiley. To us, he was a god. He said that in all his aviation career, from Air Force F-106s to wide-body Captain, nothing was as difficult as Pan Am navigator school. It was the only time in his life that he genuinely thought that he wasn’t going to get through the program. Kudus, Paul, for pushing the hobby into something new!


I think it makes a nice twist over how we all usually get from A to B in our sims. It also gives a greater appreciation for how it used to be done. There is certainly a lot more to it than pressing the ‘Direct To’ button. Having said that, if you break it down to a basic level, then you can be accurate enough for our purposes without making it too complicated.

1 Like

It’s been very interesting learning about how this is done. I still have a lot to learn of course. The nice thing is that there are apps available to us that make the process simple and relatively quick to crunch the data, or you can manually go to the look up tables. I like doing it myself, but I am much slower than the app. At the end of the day, you are still using sight readings from an accurate simulation of the night sky (Stellarium) to navigate, which I think is very neat.

The biggest kudos should go to Casper though. His app makes it all possible. I tried to write a similar app last year using Visual BASIC and failed miserably.

1 Like

I have a beautiful, functional sextant in a wood box on a shelf. I think I still have a 99 year digital tables calculator too.

This all got me to wondering if the stars shown in the Steam VR waiting menu are accurately placed… :thinking:

I was thinking about buying a Sextant to try my hand at some real sightings, but I figured that it would probably end up on a shelf in a nice box. It would be a great conversation piece though.

I used the sextant gauge in FSX for a Pole to Pole trip once. I got fairly decent with it, but it did a lot of things for you.

My brother got an antique sextant from his wife as a wedding gift (1840’s ish IIRC). He’s at sea a decent bit, but I don’t know if he’s had a chance to try it out.

A circumpolar trek might be an interesting challenge. :thinking:

We spent some time taking sextant shots at the Air & Space Museum today!

© 2020 | Articles Website | Forums Rules & FAQ