Naval Academy reinstates celestial navigation | Hacker News Reader

Naval Academy reinstates celestial navigation

navytimes.com

80 pointsaoldoni35 comments

I was in the last class to have the old celestial navigation class. It's not a difficult thing to do. You take a measurement with a sextant then consult a fifty pound book full of tables, get the two closest numbers to what you want and interpolate.

There are various techniques based on time of day and the stars you can see, but if you're not doing it regularly you're going to have to look up the procedure anyway.

And although you might not always be able to rely on GPS, inertial navigation, dead reckoning, and visual navigation using landmarks, celestial navigation has a huge drawback.

It's defeated by cloud cover and fog.

Not to be overly pedantic, but if you have an accurate timepiece you can get your longitude from averaging sunrise and sunset times. It's not particularly accurate, if you can't see the sun, but could be useful if you where otherwise lost at sea and under total cloud cover for a long period of time.

Do you mean latitude? I remember the days being longer in the summer when I lived at higher latitudes, but I'm having trouble seeing how you could calculate longitude by looking at sunrise/sunset.

Sunrise and sunset together will give you local noon (or local midnight). Local noon plus an accurate timepiece set to a known location will give you a time offset. Knowing your time offset is equivalent to knowing your longitude (which was the whole purpose behind the Harrison chronographs). You won't know it to a tremendous accuracy, since you can't get the exact moment of sunrise or sunset, just a vague idea, your position is likely to change between sunrise and sunset, and dead reckoning on a featureless sea under a featureless sky can only be so accurate. But it's knowing the difference between local time and a reference time that gets you longitude.

No, it would be the longitude. The average of sunrise and sunset will be local noon, and from that you get the longitude. To get the latitude you would need to see where the Sun rises and sets or see the stars, neither of which you could do very accurately under heavy cloud cover.

Thanks stan_rogers and splat, using local noon plus a clock makes total sense after you described how it works. And I just read a fun Wikipedia node on the history of longitude for good measure too.

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

There is an excellent book about "the Longitude problem" that details John Harrison and the invention of high precision chronometers: http://www.amazon.com/Longitude-Genius-Greatest-Scientific-P... Great short book for any engineer.

I'd rather give each sailor a course in the kind of math and programming needed to create the tables.

The Naval Academy absolutely does train thousands of naval officers in this kind of math and programming. For the most part, they will apply those skills to more difficult problems, but if those tables had to be recreated, you would have no difficulty finding an Academy officer who could do it.

The Academy doesn't train the "sailors", though, just an elite corps of officers, and you don't need to have everyone on board making his own book, just have a few specially trained people (on land, presumably) who could if they needed to. They Academy makes sure many such people exist.

https://en.wikipedia.org/wiki/Sunstone_%28medieval%29

Celestial Navigation was a part of USAF navigator training, but was retired about the same time as the USN - and it should remain that way. The USN's reinstatement of CN is a signal that the Navy does not understand it's own technology and threats(cyber attacks). And just common sense says a ship operating off CN will not be able to bring kinetic effects to the fight. The article mentions CN oriented a disabled Apollo 12? Gimme a break, what saved the mission was the quick wit of the engineering staff... and a crew that understood its own technology.

> It's defeated by cloud cover and fog.

Only on Earth we used it as one of the navigation systems to get astronauts on the moon.

Three hours of training? Either celestial navigation is much easier than i think it is or this is exactly the sort of compromise mothers should warn their children about. I'd rather sail into a GPS outage with a one in ten chance of having a navigator trained 30 hours on board than with a 100% chance of having one trained three. And it gets better when you consider crews larger than one. Sure, we all know that decision making isn't easy, but that should not be an excuse to get away with a compromise like that.

the problem persists because like any skill use it or lose it

Celestial navigation is way easier than you think it is. We have accurate clocks now, and you can just take the measurements and feed it into the computer.

The point of this training today, I would expect, is that computers might not always be there.

We might have computers, but GPS signals can be spoofed:

https://en.wikipedia.org/wiki/Spoofing_attack#GPS_Spoofing

Is there a reason why the signals aren't digitally signed? I understand that GPS and GLONASS was deployed over 30 years ago, but surely Galileo, with its first satellite launched 4 years ago can have this?

Spoofing of GPS signals isn't generally the problem the US military is most concerned about. It's availability.

The assumption is that in any significant event, US space assets will be unavailable either momentarily or for some extended period. Chinese Anti-SAT weapons are one indicator that this is a likely avenue of attack for any weaker actor seeking to degrade our capabilities.

So, celestial navigation is a crude but workable substitute for GPS but it has the distinct advantage of being ancient and nearly impossible to impede unless you can change the weather or the positions of the heavenly objects.

Here's a paper that proposes GPS authentication based on signing and statistics about the signals (lots of math): http://radionavlab.ae.utexas.edu/images/stories/files/papers...

Paper from Los Alamos with a receiver-only implementation that would use:

- abnormally high signal strength

- abnormally regular transmissions

- a secondary time source to double-check time (e.g. NTP on a smartphone)

- dead-reckoning based on accelerometers/gyroscopes/compass to double-check position

http://lewisperdue.com/DieByWire/GPS-Vulnerability-LosAlamos...

GPS signals are weak. You simply jam them.

If the people on a modern naval ship don't have access to some kind of basic computer, then they have bigger problems than navigation.

A total blackout has happened to a US Navy Ship in recent memory: https://en.wikipedia.org/wiki/USS_Yorktown_(CG-48)#Smart_shi...

And when that happened, they had a bigger problem than navigation.

GPS depends on satellites to stay in the right place, intact, and powered, as well as the ability to receive authentic RF signals from those satellites.

Celestial navigation requires (the computational equivalent of) a graphing calculator.

This topic, though a different article, received a fair bit of discussion a month ago:

https://news.ycombinator.com/item?id=10385244

Why not instead a "modern" digital sextant consisting of a digital camera and software. Still enables navigation by the stars (or sun) as a backup to GPS without having to spend thousands of man-hours training cadets to do it manually. And while not as accurate as GPS certainly, it would most likely be better than or at least equal to the best human navigators.

You'd harden the system against EMP which is probably a standard requirement of military electronics anyway (I'd hope...)

Even if you don't encounter an enemy attack that disables GPS (which is not difficult), you (meaning anyone concerned with national security) should still be prepared for navigation by alternate methods when you consider how often equipment malfunctions can cause self-jamming of GPS. It's happened to me on systems that I support.

Interesting. I've been reading the Aubrey-Maturin series by Patrick O'Brian, and besides looking up what in the world spotted-dog, calipash, staysails, the various and sundry varieties of vessels, or a Magellan jacket are, there's a fair amount of old-time celestial navigation throughout - the constantly documented daily ritual aboard ship is punctuated by the daily observation of noon, and comments on the nutation(?) of the planets... It's pretty amazing how well people were able to fix their positions with such primitive instruments.

I can pick out the Big Dipper at night, and figure out if I'm going east or west by the sun during the day, but that's pretty much the extent of it. I've never been out in the open sea, where I imagine you can actually see the stars well - the day-glo murk from the omnipresent streetlights blots out most of them if you are anywhere near civilization.

I must be the last person ever who survived thanx to sextant in 1990:

I had $50 plastic sextant "Ebco". I made a program in my pocket computer (Atari Portfolio) which automatically showed your position on a map (A line perpendicular to the sun, ie with several measurements you could get a total fix). Realized however that sextant is useless if there is no horizon visible in direction of the sun. I needed "artificial horizon" but it was $1000. Too expensive. Then I saw the movie about Nansen crossing the Greenland. All he had was a bottle of mercury. He simply measured the angle between the sun and its reflection. I felt soo stupid.

Then I started paddling from Vancouver city to the west in 1990. I did not understand them tides. I did not know you can get pretabulated tables. I was paddling against tides most of the time. Somewhere between Kelsey Bay and Telegraph Cove there was 2 weeks period of fog and rain. I totally lost it. I was running out of food. I decided to turn back to Kelsey. But then the setting sun peaked out and I was able get my longitude. I was 3 kilometers from Telegraph cove. Paddling back would have been a suicide.

About 80 km between those places. But I was doing it against tides averaging less than 10 km per day.

BTW. George Vancouver was also stuck for weeks at the same stretch of Johnstone Strait. The devilish tide current appeared to be totally random and followed none of the god-given rules and laws.

https://www.flickr.com/photos/timonoko/9513772987/in/album-7...

We will be seeing more of this kind of stuff as the world gets more complicated. Obviously, there are not enough people who know how to troubleshoot computer networks, debug poorly written legacy code and simply maintain the complex realtime systems that are used to automate buildings, boats and planes. I can imagine pilots will continue to be required on planes for this reason even though planes largely can fly themselves. Same thing with reading topo maps - a skill that all infantry will continue to need even though they have gps. In a sense, this is insurance needed to protect us from "the ghosts of Luddites'.

There's a reason certain ICBMs, like Trinity, still use celestial navigation. GPS can be disabled - but the stars aren't going anywhere.

If anyone has read "Ghost Fleet", this seems appropriate. However, I tend to agree with usrusr about the seemingly shallow depth of the course.

http://www.amazon.com/Ghost-Fleet-Novel-Next-World/dp/054414...

> The same techniques guided ancient Polynesians in the open Pacific and led Sir Ernest Shackleton to remote Antarctica, then oriented astronauts when the Apollo 12 was disabled by lightning, the techniques of celestial navigation.

You'd think the proof reader would at least catch an issue in the very first sentence.

Very cool.

Back in the 1970s I owned a sextant that I kept on my sailboat but I don't remember how to use one anymore. GPS is great but old skills like celestial navigation and old fashioned dead reckoning get forgotten.

Reply on news.ycombinator.com