How Do Scientists Spot Hidden Impact Craters From Satellite Images?

A circle on a satellite map is only a hint, never proof. To confirm an impact crater, scientists have to find rock that a shock wave crushed, and they trust two clues above all: shatter cones, which are fans of grooved cones you can spot by eye in the field, and quartz grains carrying sets of microscopic parallel lines. All but one of the 191 impact structures confirmed as of 2023 were sealed by one of those two clues, which is why roughly nine suspicious circles out of ten never make the list.

Picture this. A man in Quebec is planning a camping trip. He is dragging his cursor around Google Maps, hunting for somewhere decent to pitch a tent.

Instead he finds a hole in the world. About 15 km (9 miles) across, as he sized it up then. Nearly a perfect ring, around a lake called Marsal, roughly 100 km north of the village of Magpie. The real figure would turn out to be 25 km (15.5 miles), but establishing that took two years and a float plane.

Joël Lapointe is an amateur astronomer, so he knew the shape was odd. He was not out looking for asteroids. He was looking for a campsite.

That was 2024. In October 2025, four scientists waded ashore from a float plane to find out whether he was right. So here is the real question. How do you tell an ancient asteroid scar from an ordinary round lake? And why does the answer barely involve the satellite photo?

How Do Impact Craters Form?

Throw a stone into wet sand. You get a pit with a raised lip and a spray of sand around it. That pit is not a hole you dug. It is the shape the sand took while it was moving.

Now speed the stone up. The rock that dug Meteor Crater in Arizona arrived at 12 to 20 km/s. That is 43,000 to 72,000 km/h (27,000 to 45,000 mph). The ground has no time to move out of the way.

So the energy all goes in at once. What spreads out is a shock wave, a pressure pulse that travels faster than sound through solid rock. For a few seconds, hard bedrock flows like syrup.

Near the point of impact, rock melts or turns straight to vapor. Farther out it survives, but not unchanged. It gets crushed and rearranged at a pressure it will never see again.

Here is the one idea the rest of this article rests on. An impact does not just dig a hole. It stamps a fingerprint into the rock that lives. Weather and ice erase the hole. They do not erase the fingerprint.

It also settles a small mystery. Craters come out round even when the rock arrives at a slant, because the shock wave spreads as a ball. Hence the Moon's neat circles rather than gouges.

What Is The Difference Between A Crater And An Impact Crater?

This sounds like word games. It is not. The distinction is the reason so many circles get rejected.

Western University runs Impact Earth, the working database of confirmed impacts. An impact crater, in its wording, is a dip in the ground with no confirmed shock damage. A hypervelocity impact crater is one where the shock damage has been found and checked. In the 2023 update it listed 191 of the second kind and only 12 of the first.

One more wrinkle. Most of Earth's craters are so worn down that "crater" flatters them. Meteor Crater still looks like a bowl. A 390-million-year-old scar looks like some lakes and a few hills. The word geologists reach for is "structure."

How Do You Identify An Impact Crater From Satellite Images?

Now the fun part, the part anyone with a browser can do. From orbit, a suspect looks like this.

It is close to circular. It has a raised rim, or a ring of hills where a rim used to be. Streams run out from the middle like spokes on a wheel. And the rock inside is a different age from the rock outside.

Lapointe's pit had the ring. Around Marsal Lake sat a set of small mountains about 8 km across. He emailed his way to Pierre Rochette, a French geophysicist in Aix-en-Provence. Rochette read the shape of the land and told CBC it pointed to an impact.

And now the catch. Earth is covered in circles. Volcanic calderas, sinkholes, salt domes, glacial hollows and plain coincidence are all round. Round is not rare. Round is the default.

Gordon Osinski, the planetary geologist behind Impact Earth, gets a steady stream of hopeful emails. His verdict on Google Earth circles: "nine times out of 10 they're not."

So yes, a crater can be a lake, and often is. Which is what makes lakes such reliable false alarms.

What Does It Take To Prove A Crater Came From Space?

The satellite picture nominates. The rock convicts. Two things do the convicting, and a third one corroborates.

1. Shatter cones. Picture a bundle of ice-cream cones frozen into stone, with fine lines running down each cone. Osinski and Ludovic Ferrière reviewed them in 2016. Two of their points matter here. Shatter cones start forming at about 2 GPa, roughly 20,000 times the air pressure sitting on you right now. And they are the only shock feature you can see by eye on a lump of rock.

They also turn up almost nowhere else. The short list: impact craters, their debris, meteorites, and craters blasted out by big explosions. The review calls them unequivocal evidence of an impact.

2. Shocked quartz. Push past about 8 to 10 GPa and quartz grains grow sets of microscopic parallel lines inside them, called planar deformation features. Those need a lab, not a good eye.

3. Impact melt rock. Rock that was liquid and then froze. Osinski's team found cliffs of it in Quebec. In a crater this old he had expected none, and called the examples spectacular. But melt rock does not convict on its own, because volcanoes melt rock too. Impact Earth's rule is narrower than that: to be listed, a structure has to show confirmed shock damage in the target rock.

At Uhackatik, Jérôme Gattacceca found shatter cones on the second day. The samples dated the impact to 390 million years ago, in the Devonian period.

The tally shows how strict this is. As of the 2023 update, Earth had 191 confirmed hypervelocity impact structures, and Uhackatik now makes 192. All but one of the 191 were confirmed by shatter cones, shocked quartz, or both. The exception is Pantasma in Nicaragua, where the volcanic target rock held too little quartz and the rainforest outcrops were too weathered to read shatter cones from.

So what did convict Pantasma? Zircon. It is a stubborn little crystal that outlasts the minerals around it, and a shock wave rearranges its insides in a way that stays on the record. In Pantasma's impact glass, researchers found the trace of a high-pressure form of zircon that had since reverted, alongside a chromium isotope ratio that only comes from space. The work was led by Pierre Rochette, the same geophysicist Lapointe emailed about his ring in Quebec five years later. Rochette's first Uhackatik samples contained zircon too. That was enough to justify the float plane. It was not enough to close the case.

None of this is comfortable work. The float plane dropped the team 50 m offshore, so they waded in carrying their gear. Osinski has worked on six continents, and called this "one of the hardest field expeditions I've ever done." The terrain was rough. The bugs, by his account, were worse.

How Many Impact Craters Exist On Earth?

Fewer than you would guess. Osinski puts the count of known craters worldwide at about 200, which is the 191 with proven shock damage plus a dozen more that look right but have never been proven. Canada holds 31, and nearly a third of those sit in Quebec. At 25 km, Uhackatik is one of the biggest found in years. Most new finds land in the 5 to 10 km range.

For scale, go to the famous one. Meteor Crater in Arizona is 1,300 m (0.8 mile) wide and 174 m (570 feet) deep. It is only about 50,000 years old, and the iron rock that dug it was roughly 50 m across. It is the poster child because it is young and sits in a desert, where nothing has filled it in.

Which is the reason the global count is so low. Earth erodes, floods, buries and recycles its own surface. The Moon does none of that, so it keeps every scar it has ever taken, and the far side keeps even more.

How Do Scientists Find Craters That Are Completely Buried?

Sometimes there is nothing to see at all. Then you stop looking and start measuring.

Impact rock has the wrong density and the wrong magnetism for its surroundings. Both leak upward. A gravity map or a magnetic map can pick out a ring no camera will ever see.

That is how the dinosaur-killer turned up. Chicxulub sits under about a kilometer of sediment on Mexico's Yucatán Peninsula. Oil-survey crews mapped a huge circular anomaly there. The geophysicists Glen Penfield and Antonio Camargo worked out what it was. A 180 km ring, drawn from gravity and magnetic data rather than from a photograph. The oil industry went looking for oil and found the end of the Cretaceous.

Greenland offers the same trick with ice. Under Hiawatha Glacier lies a 31 km bowl, buried beneath up to a kilometer of ice. Researchers spotted it in radar maps of the bedrock, some of it flown by NASA's Operation IceBridge. Then they sifted the meltwater sediment washing out from under the glacier. It held shocked quartz, plus nickel, cobalt, chromium and gold, the signature of an iron rock from space.

Note what did not change. The eye was optional. The rock was not.

How Do Scientists Track A Meteorite That Is Still Falling?

Everything so far has been forensics on a 390-million-year-old crime scene. Catching a rock on the way down is a different job, with different tools.

NASA runs the All Sky Fireball Network, 15 cameras spread across the United States. Each one stares up all night for meteors brighter than Venus. The point is not romance. NASA models how much rock is flying around near Earth, so spacecraft can be built to survive it.

The trick is having more than one camera. Two stations in different states see the same fireball against different stars. From that pair of views you can pin the path in three dimensions. Run it backward and you get the orbit the rock was on.

Australia built the biggest version of this. The Desert Fireball Network runs about 50 cameras across two states. It has led searchers to at least eight Australian meteorites, among them Murrili and Dingle Dell. The same cameras now go up worldwide, as the Global Fireball Observatory.

Where Does A Meteorite Land? The Physics Of Dark Flight

Here is where it stops being astronomy and turns into weather forecasting.

A meteor glows because the air ahead of it is crushed and heated. Once it slows enough, the glow shuts off. For the Murrili fall in Australia, that happened around 18 km up. Everything after that is dark flight: an invisible rock at terminal velocity, for a couple of minutes.

Predicting where it lands needs four inputs. The mass. The shape. How much drag that shape makes. And the wind at every height on the way down.

Wind is the troublemaker. On its own it can move the answer 200 to 300 m, and high-altitude jet streams do most of the shoving. So teams do not calculate one path. They run the fall thousands of times with small changes to the numbers. The scatter of endpoints becomes the search map.

For Murrili that map came down to about 0.2 km². The rock turned up roughly 100 m along the fall line from the predicted point, and 40 m to one side. A fine result. It still leaves you walking a patch the size of 28 soccer fields, hunting one dark rock among a great many dark rocks.

One last detail explains why searchers walk in long lines. A rock that breaks up makes pieces of many sizes, and big pieces punch farther through the air. So the fragments land strung along the flight path in a long ellipse, called a strewn field. Heavy at the far end, light at the near end.

So How Do Scientists Spot Hidden Impact Craters From Satellite Images?

Strictly speaking, they do not. Satellite images do one job, and it is worth doing. They hand scientists a list of suspects. Everything after that happens on the ground, under a hand lens, or down a microscope. The photo nominates. The rock decides.

Which is why a camper with a cursor could matter at all. Lapointe never claimed to have proved anything. He supplied a hunch good enough to be worth a float plane. Osinski has spent a career filtering these tips. He was frank about the odds: "there could always still be that one that will surprise you. This is the first for me."

The crater is now called Uhackatik (also spelled Uhaachatik), a name chosen with the Innu Council of Ekuanitshit. Osinski will present the work at the Meteoritical Society's yearly meeting in Germany.

Lapointe's own summary is hard to beat. "It's not every day that an ordinary citizen finds a 390-million-year-old crater," he wrote. Then the useful part. "I encourage everyone to not ignore intuition or an observation, even if it isn't part of your field of expertise."

Sit with the timing for a second. Something hit northern Quebec while the first forests on Earth were still working out how to be trees. The scar sat there through every ice age since. And for twenty years it has been sitting on a free map, in public, at a zoom level anyone could reach. It was waiting for somebody to go looking for a campsite.