SUBSIM Radio Room Forums - Celestial navigation by day

SUBSIM Radio Room Forums (https://www.subsim.com/radioroom/index.php)

- Silent Hunter 4: Wolves of the Pacific (https://www.subsim.com/radioroom/forumdisplay.php?f=202)

- - Celestial navigation by day (https://www.subsim.com/radioroom/showthread.php?t=189378)

Quote:

Originally Posted by Rockin Robbins (Post 1784960)

In practice, at sea, in a small craft or a submarine, you're doing incredibly well to get within 20 miles of your true position. Larger errors than that are very common, leading to methods of navigation that take into account a presumed statistical error.

For example, you know you're likely within (don't you love that word likely? If you're a worrywart, it sure doesn't help the sleep process!) 30 miles, that's plus or minus 30 miles! That means there is a 60 mile range north/south and east/west that you could be in. You're approaching a coast and want to make sure you go pretty directly to a port.

Do you navigate to the position of the harbor? NO!!!!! That would be stupid.:arrgh!:

Why is it stupid. Well, your position is not a point but a probability distribution 60 miles long. When you achieve landfall there are as many possible positions north of the harbor as there are south. You ARE Shoedinger's Cat, and you've just been let out of the box. Quick, which direction do you turn to get to the harbor? Choose the choice of your choice, but either way there is a 50% chance that you have actually turned away from your destination! How cool is that?:rotfl2:

Now let's do it taking into account the error range of +-30 miles. If instead of aiming at the port, we aim for a point 55 miles to the north, we know two things. If we end up on the south end of the error probability range, we'll be in sight of the port and can easily turn whichever direction we need to. When we get there we'll brag about how accurate our navigation is.

If we come to landfall and can't see the port, we KNOW we're north of it. We turn south decisively, like we know what we're doing ('cause we do!) and sail directly into port, where we brag about how accurate our navigation is. In practice it is normal to aim 100 miles from your destination to ensure that you turn in the correct direction. If you wanted to end up at Daytona Beach you would aim for Jacksonville! Of course I'm ignoring that when you get within 50 miles of Daytona you could pick up a radio station and vector in on that. I'm assuming you are using only celestial navigation.

The gradations on the sextant are irrelevant in determining how accurately you can navigate. On a boat getting to within half a degree is darned good shooting. Like I said, atmospheric scintillation can make Sirius dance around like it hasn't been to the toilet for 24 hours. I've seen it move more than half a degree and there is no way you can fix its exact position. The accuracy of the instrument far exceeds the accuracy of the man.

These videos are not extreme examples. Note that Sirius also flashes all the colors of the rainbow while it does its "I gotta go" dance.
https://www.youtube.com/watch?v=jwn6..._order&list=UL
https://www.youtube.com/watch?NR=1&v=4f35gIWdCz0
https://www.youtube.com/watch?v=umvgCr32ssk

I understand that Stellarium does simulate this scintillation?

No program simulates scintillation that I am aware of. It would be a non-useful function. If there were still such a thing as a telescope simulator out there, then scintillation would be necessary.

Just another doubt I have for you guys using Stellarium with SH4/SH3.

When taking sights during sunrise/set exactly when do you mark the time? When the first glimpse of the upper limb of the sun shows up or when its lower limb is not touching the horizon anymore? How accurate is the sun movement and position located against Stellarium's?

Astronomically, sunset is defined as the moment of last contact between the sun's disk and the horizon. That actually puts my sunset time above 60 seconds too early, because I was solving for the midpoint of the disk as sunset. Hey that's what I get for not consulting any reference materials.......

Sunrise is the first contact of solar disk and the horizon.

Stellarium is accurate to the second on astronomical positions and timings. I prefer Cartes du Ciel for astronomical purposes, but Stellarium is just as good for navigation.

In practice, I time from first contact on sunset and last contact at sunrise. Why? Because those are easiest to time accurately. Then I apply corrections for the diameter of the sun and refraction to get my real time of sunset or sunrise.

Food for thought: How many times have you been out on a dark night (there ARE none in any of our urban areas), looked up and wondered at the countless number you see? It's a cliche from the time before telescopes, "as countless as the stars in the sky." Well it might surprise you that all the stars you can see with the naked eye are identified, their positions, brightness, variability, color precisely tabulated. And they are actually counted.

From a city, only the very brightest can punch their light through the excess lighting and the thickly obscuring atmospheric pollution. I'm not even going to hazard a guess there, but it could be less than a couple hundred. But lets get out of the city into a suburban area with no Wal-Mart within 10 miles. Here there are about 2,000 stars that are visible to the naked eye. You might find that number surprisingly low. Well the reality is half that because half are completely obscured by the Earth below your feet!

But what about the pristeen darkness many miles out to sea or in the middle of the Australian outback? How many stars can you see then? The total number of visible stars from such a location is only about 6,000, and you can only see half of them at any one time because the other half are below the horizon.

So countless? Not hardly! With nothing but your eyes you can only see a tiny number of stars, fewer by far than the population of any of our well-known cities. No matter where on Earth you are, if you could count every one you see, you wouldn't get to 3,500. Isn't that amazing?

Quote:

Originally Posted by Rockin Robbins (Post 1786012)

So countless? Not hardly! With nothing but your eyes you can only see a tiny number of stars, fewer by far than the population of any of our well-known cities. No matter where on Earth you are, if you could count every one you see, you wouldn't get to 3,500. Isn't that amazing?

The lesser artificial light to offuscate your vision, the more you can count (not to mention other factors, such as weather, pollution, air temperature etc. The atmosphere acts as a giant lens). What you are saying in terms of numbers is the absolute maximum amout of stars, in perfect sighting conditions?

At 05:13 7. sept '39 the uboat started from harbor Swinemuende with course 017° T and speed 05 kts.
At 06:42, at twilight, the commander asked for a position-fix because the executed heading and speed was incorrect.
The navigator's result:
http://i714.photobucket.com/albums/w...03_A_Testa.jpg
3-star position fix
(from SH5)

Quote:

Originally Posted by Daniel Prates (Post 1786637)

What you are saying in terms of numbers is the absolute maximum amout of stars, in perfect sighting conditions?

That's what he's saying. And I'm amazed, because I never knew that.

Quote:

Originally Posted by ETsd4 (Post 1786994)

I'm curious. I know how the three LOP with the stars are obtained, but how were the three light grey lines obtained?

The aim was to find the incenter of the triangle. It is the point of intersection of the triangle's three angle bisectors.

Quote:

Originally Posted by Daniel Prates (Post 1786637)

That is correct. Maximum number possible to see in the absolute best of conditions on the darkest spot on earth.

Quote:

Originally Posted by ETsd4 (Post 1787343)

The aim was to find the incenter of the triangle. It is the point of intersection of the triangle's three angle bisectors.

However the assumption that you would be at the center of that triangle is almost certainly wrong. What you hope is that you are somewhere in that triangle. The smaller the triangle is, the more consistent your observations are. But you could be anywhere within that triangle, IF you haven't made any mistakes, in which case you could be outside the triangle.

Celestial navigation isn't for sissies..... And it isn't for perfectionists because they go stark raving mad.

Quote:

Originally Posted by Rockin Robbins (Post 1787368)

However the assumption that you would be at the center of that triangle is almost certainly wrong. What you hope is that you are somewhere in that triangle.

yeah, tell me more. Report your "findings" to the public and write a new chapter in celestial navigation calculation. The incenter-point of the triangle should be your FIRST assumption where your actual position could be.

Assuming good observations, not a really great assumption at sea, and accurate calculations, a good assumption if your calculators are all working and you don't have to do the figuring by hand, your position is a probability bell curve with the high point in the center of the triangle. If you want to assume perfection and take that center point that's fine by me. However, in the real world, no such assumption is made except by fools.

Real navigators are aware at all times what error is intrinsic in the process and generally laugh at precise points of position. We, who live in a GPS world, expect accuracy measured in terms of six feet, two meters or so. Celestial navigation deals with accuracies measured in error ranges of dozens of miles.

The three lines on your celestial nav charts are position lines. Each line says "I'm somewhere on this line." They should join at a single point. However, due to the intrinsic inaccuracies of the measuring process, they never do. You end up with a triangle, each line saying "you're somewhere on this line." Surely you're not going to argue that they are all accurate, for then you would have to be three places at once!

The correct answer to "where am I" is "probably somewhere in this triangle." The midpoint is just as bad a position as somewhere on one of the lines marking the border of the triangle. Any real celestial navigator knows this. This is not a new chapter, but one two hundred years old.

Quote:

Originally Posted by ETsd4 (Post 1787343)

The aim was to find the incenter of the triangle. It is the point of intersection of the triangle's three angle bisectors.

OK. I couldn't tell from the picture there.

I believe there are methods for refining the initial LOP's, but when I was using Cel Nav, I didn't get that far into it. RR probably knows all of this stuff.

Quote:

Celestial navigation isn't for sissies..... And it isn't for perfectionists because they go stark raving mad.

You can say that again.

After doing this for a week or so, I felt like I'd been subjected to a 12 hour depth charging. :dead:

Quote:

Originally Posted by TorpX (Post 1787812)

OK. I couldn't tell from the picture there.

I believe there are methods for refining the initial LOP's, but when I was using Cel Nav, I didn't get that far into it. RR probably knows all of this stuff.

You can say that again.

After doing this for a week or so, I felt like I'd been subjected to a 12 hour depth charging. :dead:

Doing it for real on a real small craft multiplies the uncertainties tremendously!