What if every snowflake were the same?

Catch a snowflake on your mitten. Look closely!

It has six arms. Every snowflake looks a bit different. What if they were all the same?

Look closely at snowflakes landing on a dark mitten and you'll see tiny six-armed stars, needles and flat plates. People often say that no two snowflakes are alike. But why should that be? And what would it take for every snowflake to be exactly the same?

This thought experiment follows a snowflake's journey through the clouds, finds out what decides its shape, and asks what we would lose in a world of identical flakes.

Snow crystals form when water vapor turns directly into ice. Their shapes are famously varied and yet each crystal is often strikingly symmetrical, which raises two questions at once: why are its arms alike, and why are different crystals not?

This lab uses the physics of crystal growth, chiefly from Caltech physicist Kenneth Libbrecht, to examine what an all-identical snowfall would require.

A snowflake starts as a tiny bit of ice in a cloud.

Water from the air sticks to it. It grows six arms as it falls!

A snow crystal begins in a cloud when water vapor, the invisible water in the air, turns straight into ice. It starts as a tiny six-sided plate. As more vapor freezes onto it, arms sprout from its six corners.

As the crystal tumbles through the clouds, the temperature and dampness around it keep changing, and each change makes the arms grow a little differently. Its final shape is a record of the exact path it took.

Libbrecht describes a stellar crystal starting as a small hexagonal plate, with branches sprouting from the six corners as it grows. Changing temperature and humidity along its fall alter the growth again and again, so the final form encodes the precise path through the cloud.

The arms are not linked by any special force. They grow independently, but all six share the same changing conditions at the same moments, so they grow in near synchrony. Libbrecht also notes that many real crystals are less symmetrical than the famous photographs suggest.

In our pretend world, every snowflake looks exactly the same.

For that, every flake would need the very same trip through the cloud!

Imagine a snowfall where every flake was identical. For that to happen, every crystal would need to start the same way and then meet exactly the same temperatures and dampness at exactly the same moments as it fell. In a real, swirling cloud, that doesn't happen.

But in a lab, it nearly can. Physicist Ken Libbrecht grows pairs of snow crystals side by side under carefully controlled conditions. Because they share the same changes, they grow into nearly the same shape. He calls them identical-twin snowflakes.

An identical snowfall would require every crystal to experience the same sequence of temperature and supersaturation at the same times: effectively one path for all. Real clouds mix and swirl, so paths diverge.

Libbrecht's 'identical-twin' crystals show the principle in reverse. Seed crystals grown near each other under tightly controlled temperature, humidity and air flow develop nearly the same shape. They are very similar but not precisely identical, and when their branches grow close they compete for vapor and diverge.

Snowflakes change shape with the cold.

A little cold makes flat plates. A bit colder makes long columns. Even colder makes big, flat stars!

Scientists found that snowflake shapes depend on two things: temperature and dampness. Thin, flat plates grow near −2 °C, slender columns near −6 °C, and big thin plates near −15 °C. In damp air crystals grow fast and form fancy branches; in drier air they stay simpler.

Japanese physicist Ukichiro Nakaya worked this out in the 1930s by growing snow crystals in a cold lab. Even today, scientists can't fully explain why ice grows as plates at some temperatures and columns at others.

The snow crystal morphology diagram, discovered by Ukichiro Nakaya and collaborators in the 1930s, maps habit against temperature and supersaturation: plates near −2 °C, columns near −6 °C, large thin plates near −15 °C, and either at colder temperatures; higher humidity gives faster, more branched growth. An updated version of the diagram described by Hokkaido University's Institute of Low Temperature Science shows the same alternation: plates from 0 to −4 °C, prisms to −10 °C, plates again to −22 °C, then prisms.

The branching instability explains complex structure and its humidity dependence, becoming essentially chaotic for fast-growing fernlike dendrites. Libbrecht writes that why habit alternates between plates and columns with temperature is still not fully understood; it depends on the molecular dynamics of ice surfaces.

Long ago, a farmer named Wilson Bentley took photos of snowflakes.

He took more than 5,000 pictures! Every one was different.

Wilson Bentley, a farmer from Jericho, Vermont, was a pioneer of photographing single snowflakes. He photographed his first one in 1885, when he was 19, using a camera attached to a microscope. Over 46 years he captured more than 5,000 snow crystal images. The saying that no two snowflakes are alike seems to have started with his photos.

Nakaya liked to say that snowflakes are like hieroglyphs, a kind of picture writing, from the clouds, because their shapes tell you about the clouds they fell through.

Bentley photographed his first snow crystal in 1885, aged 19, and recorded more than 5,000 on glass plates over 46 years; in the late 1920s he teamed with W. J. Humphreys of the U.S. Weather Bureau to write Snow Crystals, with more than 2,000 images. Libbrecht traces 'no two alike' to Bentley's work.

Nakaya likened snowflakes to hieroglyphs from the clouds: from a crystal's form you can infer conditions aloft. A capped column, for example, grew first near −6 °C and then in colder air. If all flakes were identical, that record would be lost.