What if glaciers melted?

High on cold mountains, snow piles up and turns into ice. That ice is a glacier.

What if all the glaciers melted? Where would the water go?

A glacier is a huge mass of ice and snow that slowly flows over land, like a frozen river. There are more than 200,000 of them, high in mountains and near the poles.

When ice on land melts, the water doesn't stay put. It runs downhill into rivers and, in the end, into the sea. So how much would the sea rise if every mountain glacier melted? In this lab you can melt them all in a model and measure the rise.

Glaciers are accumulations of ice and snow that flow over land. Excluding the Greenland and Antarctic ice sheets, the world's glaciers number about 215,000 and hold a volume of 158 ± 41 thousand km³ (Farinotti et al., 2019).

Land ice adds water to the ocean when it melts; floating ice barely does. This lab asks how far the sea would rise if all of that glacier ice melted, and how the simple volume answer compares with a careful scientific estimate.

Glaciers grow when lots of snow falls. They shrink when lots of ice melts.

Many people use glacier water to drink and to grow food.

A glacier grows where more snow falls each year than melts, and the extra ice slowly flows downhill. Glacial ice covers about 10 percent of Earth's land, counting the giant ice sheets of Greenland and Antarctica.

Glaciers matter to people. In dry places near mountains, their meltwater feeds rivers in late summer, when snow has already melted. Farmers in some valleys have used glacier water for their crops for hundreds of years.

NSIDC reports that glacial ice covers about 10 percent of Earth's land, and that glaciers and ice caps, though only about 0.5 percent of land ice, contributed more to sea-level rise during the last century than the ice sheets.

Glacier meltwater sustains rivers: in Central Asia's Amu Darya, glacier melt is about 8 percent of annual flow but 27 percent in late summer. The IPCC (AR6 WG1 SPM, 2021) assesses that almost all the world's glaciers have been retreating at the same time since the 1950s and that human influence is very likely the main driver since the 1990s.

In our pretend lab, Olly melts all the mountain glaciers at once.

The water flows into the sea. The sea goes up a little, a bit more than the length of a school ruler.

Choose 'all mountain glaciers'. That is about 158 thousand cubic kilometers of ice. In the model, its meltwater spreads over the whole ocean and raises the sea by about 40 centimeters. That is only about 5 percent as much ice as Greenland holds.

Now try melting the same amount of floating ice instead. The sea barely moves: about 1 centimeter. Floating ice is already pushing aside almost its own weight of seawater.

With 158 thousand km³ of ice at 917 kg/m³ melting into fresh water over a fixed 361.8 million km² ocean, the model gives Δh = 0.40 m, 5.3 percent of Greenland's volume. The same volume of floating ice adds only 10.9 mm, the small extra volume of fresh meltwater over denser seawater.

The model shows how much, not how fast. Its 0.40 m is a simple volume answer. Farinotti et al. give 0.32 ± 0.08 m once the roughly 15 percent of glacier ice already below sea level is subtracted: that ice sits in space the ocean already fills, so melting it adds little. Taking 15 percent off the model's answer gives about 0.34 m; the rest of the gap comes from their more detailed calculation, which this model does not reproduce.

Ice on land is like ice cream on a plate next to a bath. When it melts, it drips into the bath.

Ice floating in the sea is like an ice cube in a glass. It is already in the water.

Ice resting on land is not in the ocean yet. When it melts, all that water is new to the sea, so the sea gets higher.

Ice already floating in the sea is different. It already pushes aside about as much water as it will make when it melts, like an ice cube in a glass of water. That is why melting floating ice barely changes the level, and why scientists watch the ice on land most closely.

Land ice: Δh = V·(ρ_ice/ρ_fresh)/A_ocean. Floating ice already displaces its own mass of seawater; melting adds only V·ρ_ice·(1/ρ_fresh − 1/ρ_sea), about 2.5 percent of its ice volume, because fresh meltwater is less dense than seawater.

The model's limits: an evenly rising ocean of fixed area, no ice grounded below sea level, and no information about timing. Real sea level also varies from place to place, because of gravity, land motion and currents.

Scientists watch glaciers from the ground and from space.

In one dry valley, a clever engineer even built little glaciers to store water for farmers!

Scientists measure glaciers carefully, from the ground and with satellites. A set of reference glaciers that scientists have tracked for decades lost ice about five times faster in 2010–2018 than in the 1980s. Glacier melt is part of why the sea has risen 21 to 24 centimeters since 1880.

People are finding ways to adapt. In Ladakh, India, engineer Chewang Norphel built his first 'artificial glacier' in 1987: water is slowed and frozen in winter, then melts in spring just when farmers need it.

NOAA Climate.gov reports that decadal average loss from the World Glacier Monitoring Service's reference glaciers grew from the equivalent of 171 mm of water in the 1980s to 850 mm in 2010–2018, and that global sea level has risen 21–24 cm since 1880. The IPCC (AR6 WG1 SPM, A.4.3) attributes 22 percent of 1971–2018 sea-level rise to glaciers.

Adaptation is local: NSIDC describes Chewang Norphel's low-cost artificial glaciers in Ladakh, first built in 1987, which store winter water as ice at lower, earlier-melting altitudes.