Can artificial snow save a melting glacier? Scientists tested cloud seeding on China’s Muz Taw Glacier and found a surprising change in ice loss

Scientists have tested an unusual idea for slowing glacier loss: make it snow.

In a 2018 experiment on China's Muz Taw Glacier, researchers used cloud seeding to encourage snowfall over the shrinking ice field. The goal was straightforward — cover the glacier with fresh, bright snow so it would reflect more sunlight and absorb less heat during the summer melt season.

The results were striking but limited. After the artificial snowfall, researchers recorded a 14% to 17% reduction in glacier mass loss compared with the preceding observation period. The researchers stressed that the experiment was preliminary and that much larger and longer studies would be needed before artificial snowfall could be considered a practical way to protect glaciers.


The experiment offers a glimpse into one of the more unusual approaches being explored as scientists grapple with accelerating glacier retreat.

The glacier at the centre of the experiment

The Muz Taw Glacier sits in the Sawir Mountains along the China-Kazakhstan border. In measurements from 2016, it covered about 3.13 square kilometres and stretched roughly 3.2 kilometres along its valley.

The glacier has been retreating for decades, with researchers noting continued and accelerating ice loss since 1959. That made it a useful location for testing whether additional snowfall could temporarily reduce melting.

The scientists focused on the glacier's summer ablation period, when ice loss is particularly pronounced.

Their thinking was based on a familiar property of snow: fresh snow is highly reflective. Bare or darker ice absorbs more incoming solar energy, while a clean snow surface sends a greater share of that radiation back into the atmosphere.

If enough fresh snow could be placed on the glacier, the researchers reasoned, it might both add some mass and temporarily make the surface less efficient at absorbing heat.

How scientists made the snow

The team did not simply manufacture snow and transport it onto the glacier.

Instead, they attempted to encourage clouds to produce additional precipitation.

Researchers used silver iodide, or AgI, as a cloud-seeding agent. Silver iodide particles can act as nuclei around which ice crystals form under suitable atmospheric conditions.

Fourteen remotely controlled, solar-powered AgI generators had been installed along rivers near the glacier by the local meteorological service. The scientists used weather radar to track clouds approaching the mountain area and determine when conditions were suitable for seeding.

When potentially suitable clouds were identified, the generators were activated. Valley winds then carried the particles upward toward the glacier.

Importantly, the process did not create water from nothing. It depended on clouds already containing sufficient moisture and appropriate atmospheric conditions for ice formation and precipitation.

Fresh snow changed the glacier's surface

The researchers carried out artificial precipitation experiments on August 19 and 22, 2018.

Automatic weather stations and measurements taken at different points on the glacier allowed the team to compare conditions before and after the snowfall.

One of the most important changes was in surface albedo — the amount of sunlight reflected by a surface.

Fresh snow increased the glacier's reflectivity, particularly across its middle and upper sections. At some higher locations, researchers recorded albedo values approaching 0.8.

That matters because a brighter surface generally absorbs less solar energy.

The effect is similar to the difference between wearing a dark shirt and a white shirt on a sunny day: the darker surface tends to absorb more radiation, while the lighter one reflects more.

For a glacier, even a temporary increase in reflectivity can influence how quickly snow and ice melt.

Ice loss dropped after the snowfall

The most notable result emerged when scientists compared two six-day periods.

From August 12 to 18, before the artificial snowfall, the estimated glacier-wide mass balance was about −61.4 millimetres of water equivalent.

From August 18 to 24, following the experiments, the figure improved to approximately −37.2 millimetres.

In other words, the glacier lost substantially less mass during the second period. The researchers calculated that average mass loss was between 32 and 41 millimetres lower, equivalent to a reduction of roughly 14% to 17% compared with the earlier period.

The result was notable because the accumulated positive temperature — a measure related to melting conditions — was actually higher during the second period.

The researchers attributed the reduced loss to the combination of additional snow and the resulting increase in surface reflectivity.

How much of the snow was actually artificial?

This is where the experiment becomes more complicated.

Scientists cannot simply assume that every snowflake falling after a cloud-seeding operation was produced by the silver iodide.

Natural precipitation was also occurring, meaning researchers had to estimate how much of the measured snowfall was associated with the intervention.

By comparing observations from a weather station on the glacier with measurements at a control location outside the seeding area, the researchers estimated that natural precipitation could account for as much as 21% of the snowfall recorded during the experiments.

Their analysis suggested that artificial snowfall may have accounted for at least 79% of the snow measured at the glacier's equilibrium line during the experiment.

That distinction is important because cloud seeding remains an area where separating the effects of human intervention from naturally occurring weather can be difficult.

Why the snow could slow melting

The researchers proposed a simple chain of effects.

First, cloud seeding encourages additional snowfall under suitable conditions.

The new snow then adds some mass to the glacier. More importantly, its bright surface increases the glacier's albedo.

That higher reflectivity means less solar radiation is absorbed by the surface. With less energy available to melt the snow and underlying ice, the rate of mass loss can fall.

The process can therefore create a short-term feedback: more snow means a brighter surface, a brighter surface reflects more sunlight, and less absorbed energy can mean slower melting.

The experiment's results were consistent with that mechanism.

This is not a solution to climate change

Despite the eye-catching results, the study does not show that cloud seeding can save glaciers from climate change.

The experiment covered a very small area and lasted only a short time. The researchers themselves described the work as a preliminary trial and called for further controlled experiments across larger areas and longer periods.

There is also a fundamental limitation: cloud seeding cannot work without suitable clouds.

If the atmosphere is dry and clear, there is no ready supply of moisture for the technique to convert into snowfall. The experiment therefore cannot simply be repeated whenever a glacier begins melting rapidly.

There are also broader questions about whether artificially changing precipitation in one location could affect precipitation patterns elsewhere.

A glimpse at an unusual form of climate intervention

The Muz Taw experiment is part of a wider search for ways to slow glacier loss.

Scientists and policymakers have explored a range of interventions, from physical coverings designed to reflect sunlight to proposals involving artificial snow. The idea behind these approaches is not to reverse global warming, but to temporarily reduce melting in particularly vulnerable locations.

For now, the evidence from Muz Taw is best viewed as an intriguing scientific experiment rather than a ready-made climate solution.

Still, the results show why researchers are investigating increasingly unconventional methods. A relatively small amount of fresh snow was enough to make the glacier surface considerably brighter and was followed by a measurable reduction in short-term mass loss.

The bigger question is whether that effect could be maintained over months or years — and whether it could ever be achieved on a scale large enough to make a meaningful difference.

Until those questions are answered, cloud seeding remains an experimental tool, not an escape route from a warming climate.