The whale shark wears a map of the stars

The first thing you notice, when an eight-metre whale shark slides out of the blue beneath you, is the pattern. Pale spots and lines, scattered across slate-grey skin like someone tipped a jar of stars over the animal's back. It looks decorative. It is, in fact, a name.

Every whale shark on Earth carries a spot pattern that is unique to it — as individual as a human fingerprint, and concentrated in the region just behind the gills and above the pectoral fin. No two are alike. The pattern is stable for life. Which means that, in principle, if you photograph the same patch of skin on the same shark a decade apart and a thousand miles apart, you can prove it's the same animal.

The hard part was never the biology. It was the matching.

A problem astronomers had already solved

Picture the job: tens of thousands of photographs of spotted flanks, taken by divers at wildly different angles, distances and light levels, and the question is this shark already in the catalogue, or is it new? Done by eye, it's hopeless at scale.

But "match this scatter of bright points against a library of other scatters of bright points" is a problem another field had been chewing on for decades — astronomy. To line up two photographs of the night sky, you have to recognise the same constellation of stars across images taken by different instruments, at different rotations, at different scales.

In 1986, a Princeton physicist named Edward J. Groth wrote an algorithm to do exactly that, in service of mapping star fields for the Hubble Space Telescope. The Groth algorithm builds triangles between every possible triplet of stars in an image, then compares those triangles' geometry against the triangles in another image. Triangles are clever: their shape survives rotation, scaling and translation, so two photos of the same star field will share the same set of triangles even if one is tilted or zoomed.

A whale shark's flank, it turns out, is just a star field that swims.

The crossover

The leap was made by an unlikely trio. Brad Norman, a marine biologist in Western Australia and founder of the conservation nonprofit ECOCEAN, had spent years identifying the whale sharks of Ningaloo Reef the slow way — photographing each animal and matching its spots by hand. He understood that the spots were a fingerprint. He needed a faster way to read them.

He found it through Jason Holmberg, a software programmer, who in turn brought in Dr. Zaven Arzoumanian — an astrophysicist then working with the Universities Space Research Association at NASA's Goddard Space Flight Center. Arzoumanian recognised Groth's star-matching algorithm as the right tool. Holmberg and Arzoumanian adapted it for living animals, adding rotation correction and contrast enhancement to cope with murky water, and compensating for the fact that a shark is a curved, three-dimensional thing rarely photographed in perfect profile — unlike a flat, obliging sky.

The result was published in 2005 in the Journal of Applied Ecology (Arzoumanian, Holmberg & Norman) under a title that says it all: An astronomical pattern-matching algorithm for computer-aided identification of whale sharks. The telescope had been pointed at the sea.

A library anyone can add to

What that software made possible was a global, crowdsourced catalogue. The ECOCEAN Whale Shark Photo-identification Library — which grew into today's Wildbook for Whale Sharks — lets anyone who has swum with a whale shark upload their photographs. The algorithm checks each new flank against every shark already on file and either returns a match or enrols a new individual. A snorkeller's holiday snap becomes a data point. By the time NASA itself wrote up the spin-off, the library held images of more than 1,600 individual sharks; it has grown enormously since.

It is one of the purest examples of citizen science there is: a tourist with a waterproof camera, an algorithm built for the Hubble, and an endangered animal that finally became countable.

Which brings us home, to the Maldives

If you want to see this system at work, there are few better places on the planet than the Maldives — and one place in particular.

Off the western rim of South Ari Atoll, around the islands of Maamigili and Dhigurah, there is a population of whale sharks that simply doesn't leave. The Maldives Whale Shark Research Programme (MWSRP) has studied them for well over a decade, and its conclusion is remarkable: there is no seasonality in the number of individual sharks here. This is one of the only places on Earth where whale sharks can be encountered year-round, rather than during a brief migratory window.

It was MWSRP's work on these core habitats that, in 2009, provided the baseline data for the creation of the South Ari Marine Protected Area (SAMPA) — the largest protected area in the Maldives. And the same photo-ID method born from the Groth algorithm underpins their catalogue, the Big Fish Network: as of early 2022, 726 individually identified whale sharks, drawn from over 8,000 encounters logged by 185 contributors. One shark, catalogued as WS071, has been recorded more than 380 times across 14 years — the same animal, the same map of spots, recognised again and again.

The population is overwhelmingly juvenile males, which is part of why South Ari matters so much scientifically: it appears to be a nursery, a waystation in lives we still barely understand.

So the next time you're hovering at the surface in the Maldives and a whale shark drifts past beneath your fins, look at the spots. You're not looking at decoration. You're looking at a name, written in a script that an astronomer taught us to read — and, very possibly, at an individual that researchers have known for years.

Meet the species, and learn where and when to find it, in our field guide to the whale shark in the Maldives.