I have tattoos of a rooster and a pig on my feet. They’re meant to protect me from sinking. I have a nautical star on my forearm, so I can always find my way home. I wear red pants at boat shows and lectures. I have a passion for the traditions of the sea.

Celestial navigation tops them, with its blend of romantic art and practical science. Since I first read Bernard Moitessier’s book The Long Way, long before ever going offshore myself, I’ve wanted to cross an ocean using only sun and stars as my guide.

In the spring of 2017, sailing north from the BVIs to Bermuda with the ARC Europe fleet, we raised the stakes – we’d sail the route on our Swan 48 Isbjörn navigating entirely by celestial means. We wanted to see if we could do it.

I first learned celestial navigation ten years ago from John Kretschmer at a workshop he hosted at his home in Fort Lauderdale. John is the reason I pursued a career on the ocean. He’s well known to most sailors in America and made history in 1984 when he sailed a Contessa 32 called Gigi from New York to San Francisco the ‘wrong way’ round Cape Horn, an adventure that is immortalised in his book Cape Horn to Starboard. The very day that Gigi rounded the Horn, 25 January 1984, was the day I was born.

During the weekend workshop I got to practise taking morning sun sights on the beach with the old Freiberger sextant that John had used to navigate around the Horn on that famous voyage.

John described celestial navigation in romantic terms, explaining it in a way that made it as inspiring as it was understandable. Here was someone who spoke my language, the language of the great sailing romantics like Moitessier and Sterling Hayden. John made celestial bigger than just navigating for, after all, the likelihood of a modern day sailor actually needing celestial is effectively nil.

Article continues below…

Time is everything

“Has the boat motion really settled down a lot or am I just feeling better?” Tom, one of our crew, asked on the second morning of the passage north from Tortola.

He and Cheryl had the watch and were at the helm while the crew was gathered in the cockpit for the day’s noon sight. I led the process while eating a bag of corn chips in an effort to stave off the early-passage mal de mer. Thane had the sextant and Mike was note-taker and timekeeper.

“Is it the 8th? What’s today?” asked Cheryl. “It’s the 7th today, isn’t it? Or no, it is the 8th,” I replied, not so confidently.

Normally on an ocean passage the days really don’t matter. Not so when you’re using celestial navigation. A four-second error on the time you took the sight equates to a one-mile mistake in determining the sun’s geographic position. Time is everything.

Isbjörn had departed Tortola with the ARC Europe fleet and we’d initially sailed west down Sir Francis Drake channel, rounding Jost van Dyke to starboard and pointing the bow for Bermuda. The boat galloped north at first, carrying the easterly trades on a rhythmic swell under hazy skies. Our dead-reckoning plot was easy to keep track of as Isbjörn beam-reached up the rhumb line, full sail flying, at eight knots.

At just shy of 1,000 miles, the passage to Bermuda is long enough to find your sea legs, but short enough to forego that 5 o’clock cocktail without regret. The Gosling’s Family Reserve in Bermuda is worth waiting for anyway.

But the Trades faltered sooner than we all wanted them to. Through the winter in the Caribbean, Mia and I had got so accustomed to sailing in 20 knots of breeze with small sails that it felt rather odd when we first sailed into an area off the coast of northern Florida more affected by continental weather than the tradewinds and lost the breeze for the first time in months. A weak cold front passed overhead and suddenly Isbjörn was on port tack.

Secret GPS positions

We had to eliminate the nearly-impossible-to-avoid GPS inputs while still maintaining some semblance of safety. The old Garmin chartplotter’s GPS antenna had given up the ghost, so we didn’t have to worry about that, or the VHF, which was integrated to it.

We had an AIS app on the iPad that allowed us to see targets around us and their CPAs, streamed wirelessly from the built-in Vesper XB8000 transceiver, but that would hide our own position. We had a paper passage chart, bound copies of the Nautical Almanac and the Sight Reduction Tables for Air Navigation.

Mia would keep a secret GPS record in a separate logbook in case of emergency. Ironically, friends and family following the rally from afar would know our position more accurately than we would through our YB tracker.

Thane had signed up for the passage in spite of the celestial navigation part of it, not because of it. He was an experienced offshore sailor, having sailed across the Atlantic westabout, double-handed with his wife, Brenda, on their Bavaria 37.

“Holy smokes, this is so cool!” he exclaimed the first time he managed to grab an evening twilight star sight.

The sun had only just sunk beneath the horizon to port. The western sky was painted an array of pinks, yellows and oranges, while overhead blue faded to black as night approached to starboard. If you looked hard enough, you could just make out the evening’s first stars. We were in that ethereal slice in time photographers call the magic hour and navigators call civil twilight.

Thane had used the ‘no scope, two eyes open’ approach on that first star sight that Moitessier had used on Joshua. ‘I felt that I was becoming an expert in taking star sights since I discovered that it can be done without the telescope, keeping both eyes open,’ Bernard Moitessier wrote in his book Cape Horn: The Logical Route.

‘In this way, a star can be brought down to the horizon because the latter can be seen quite clearly with both eyes open. It is impossible to do this properly while looking through the telescope where the horizon always looks hopelessly blurred. In my innocence, I thought I was the first to discover this method…’

During our one-day crash course in Tortola, I’d described to the crew this method in theory. With one sight that evening, on the rolling deck of a boat at sea where the accuracy of his sight had real-life consequences, Thane had instantly and enthusiastically bridged the gap to celestial in practice, experiencing the same joy of discovery that Moitessier had uncovered and written about some 50 years earlier. ‘Even the best navigators are not quite sure where they’re going until they get there, and then they’re still not sure!’

Breadcrumbs in the wood

Traditionally, navigation was about keeping a detailed record of where you’d been in order to plot a course to where you’d like to go. Hansel and Gretel knew how to navigate – the breadcrumbs-in-the-forest trick was the fairytale version of dead reckoning.

Navigation was rooted in superstition. Never did a sailor tempt fate by arrogantly declaring they were sailing ‘to’ a faraway port; it was always ‘towards’. This thinking contained equal doses of humility and flexibility that the modern navigator ignores at their peril.

Teaching celestial navigation in a modern context, then, involves filtering fundamental concepts through a particular lens. Take latitude, for example. It’s derived by taking a north-south cross-section of the earth and extending lines from the centre outwards, like spokes on a bicycle wheel.

Where those spokes intersect the surface of the earth creates a given line of latitude, which is drawn on the earth’s surface around the world horizontally. The degrees between lines of latitude on the surface are actually the angle between those bicycle spokes.

Nautical miles on the surface of the earth, then, correspond to those angles. Everyone knows that one minute of latitude is equal to one nautical mile, and that 60 of these make one degree of latitude. But have you ever stopped to think how far a nautical mile is on the moon? Or on Jupiter?

A nautical mile on another planet is still derived in exactly the same way, but it’s the body’s circumference that determines the actual geographic distance of it on the surface of that body. A statute, or land mile, is contrived. A nautical mile is an elegant expression of geometry.

Dive a little deeper. The distance on the surface of the earth from 0° to 231⁄2° North, for example, is 60×23.5 or 1,410 nautical miles. It’s also 1,410 nautical miles from the moon’s equator to 231⁄2° north on the moon, but the distance as measured in feet or metres is much shorter because the moon isn’t nearly as big.

That 23 1⁄2° North, by the way, is the Tropic of Cancer. The Tropic of Capricorn, conversely, lies at 23 1⁄2° South. Those aren’t just made-up boundaries: the geographic tropics are de fined naturally by the limits of the movement north and south of the sun’s declination throughout the year as it traces a sine curve from season to season, due to the tilt of the earth.

The other half of the sun’s geographic position (GP) – longitude, or Greenwich Hour Angle (GHA) in celestial parlance – is directly convertible with time and changes by the second. The sun’s GP travels westabout through 360°, right around the earth, in 24 hours, or 15° per hour.

Logically, then, I can predict the sun’s GHA in my head if I know the time in Greenwich, 1400 UT, for example, would put the sun about 030°. GHA, unlike longitude, is measured through 360°; the sun can never travel east, after all.

In simplified terms, when we take a sextant altitude of the sun we’re creating a right angle triangle between it, the earth’s surface at the GP, and ourselves. Grade school geometry tells us that the two angles in a right-angled triangle must equal 90°.

So, the complement to the altitude projects an angle from the sun onto the surface of the earth which, just like in the latitude example above, can be converted to nautical miles. After accounting for the sun’s declination north or south, depending on the season, this is precisely how we get our latitude from a noon sight.

A single sextant sight produces a giant circle of position, with the complement to our sextant altitude describing the radius of the circle, the GP at its centre. If we had a large enough chart, and an accurate way to take a compass bearing towards the GP, you could plot this using the simplest of fixes, bearing and range, to pinpoint a position on that circle. Alas, we have neither.

So, in a nutshell, modern celestial using the Sight Reduction Tables for Air Navigation (Pub. 249 in the US), allows us to compare the sextant reading from our unknown location at a known moment in time, with a sextant reading from a known location that’s somewhere in our neck of the woods, called the ‘assumed position’, and plot the difference on a chart, producing a single line of position that just so happens to be a tangent to that larger circle of position… Deep breath!

In reality, none of this is important to the modern GPS navigator. But – and here’s why I love teaching celestial navigation so much – these Eureka moments about geography and geometry and the basic understanding the fundamentals of celestial makes everyone a better navigator, whether you actually ever pickup a sextant or not.

The ocean felt deserted. There were no other boats to be seen, and no more flying fish. No dolphins. Nothing but the routine.

I don’t stand a watch on our Isbjörn passages, instead maintaining a more traditional captain’s role, overseeing the big picture and forever on-call should the crew need me on deck. Again, I’m modelling Moitessier.

He wrote once that when the weather is nice and things are going well, the captain can sleep for 36 hours if he wants. On the other hand, when the weather is bad, and stress high, the captain must remain at the helm indefinitely.

When things are good, I’ll often take half of Mia’s nighttime watch. There’s something about being alone in the cockpit at night. It’s precisely why I go ocean sailing.

Sunrise and moonset

I relieved Mia pre-dawn at 0400 and settled in for my two hours outside while the crew slept. Firmly into the mid-latitudes, and after another clearing frontal passage, the sky had lost all its Caribbean moisture and haze, replaced by a clarity in the air rarely seen ashore.

The glimmer in the east came early that morning. In opposition, the full moon casually and simultaneously sank lower on the horizon. I couldn’t decide where to focus my attention; I wanted to witness that first glimpse of the sun piercing the eastern horizon, but didn’t want to miss Mr Moon dipping ever lower in the west.

Isbjörn sailed on a northerly zephyr and an oily sea, forcing me to concentrate on the helm in order to maintain her momentum, but distracting me from that beautiful sunrise and moonset. It was very fine light-air sailing, but there were troubles with celestial. Where were we?

We’d forgotten to account for the apparent altitude when taking the noon sight the day before, a correction to the sextant angle that’s applied to account for the refraction of the sun’s ray’s in the atmosphere. The log read 581 miles sailed since leaving Tortola when I wrote in the logbook on the morning of 10 May, our fourth day at sea. It had been overcast the day before, so difficult to take any sun sights, and the ones we did get were off.

To boot, we’d gone 12 hours overnight, sailing well east of the rhumb line, close-hauled on a light northerly, which didn’t allow us to lay the course.

Non-sailors assume celestial is about navigating by the stars, at night. It’s not, of course – star sights do the job, but you need a visible horizon, which only happens at dusk and dawn. So it’s down to Mr Sun, who guides you most of the way, and on cloudy days Mr Sun is hard to find. You’re always sailing blind at night.

No matter. At 0300 on the morning of 12 May, just before the dawn of our sixth day at sea, Gibb’s Hill Light on the south-west corner of Bermuda hove into view right where we expected it to. The log read 838 miles sailed.

Accurate enough

Celestial navigation had gotten Isbjörn to Bermuda, legitimately, and with a crew of amateur sailors, two of whom had only just learned the methods literally the day before departure. I’d always wondered if we could do it, and now I know.

It’s certainly not a practical, efficient means, by anyone’s reckoning. They say that ‘close only counts in horseshoes and hand grenades’. And in celestial navigation.

The interesting part is that, without a GPS, we never really knew how accurate our sights were, and we still don’t. In the end Gibb’s Hill Light appeared where we expected it to. Our sextant sights, DR plots and LOP reductions were accurate enough to get us there successfully.

Nobody cared whether our individual LOPs throughout the trip were within two miles of our GPS position or ten, and the crew enjoyed stargazing at night, quickly forgetting the chartplotter gazing we’re all so used to.

Not unlike Heisenberg’s famous principle, perhaps the most profound irony of modern navigation is that the closer we get to perfect GPS accuracy, the farther we get from ever knowing where we truly are.

About the author

Andy Schell and his wife, Mia Karlsson, sail 10,000 miles per year on their S&S Swan 48 Isbjörn, taking paying crew on ocean passages in the Atlantic, Arctic and worldwide. Andy also hosts the On the Wind sailing podcast on his website (59-north.com) featuring interviews with well-known sailors from around the world.