Albedo is the fraction of incoming sunlight that a surface reflects. When bright snow or ice is replaced by darker water, the Arctic absorbs more solar energy. That change closes a positive feedback loop.

Albedo and absorbed energy

Albedo ranges from 0 to 1:

• 0 means all incoming sunlight is absorbed;
• 1 means all incoming sunlight is reflected.

Typical Arctic values from the lectures are:

Under the same July input of 458 W/m² at 75°N, the corresponding absorbed energy is approximately:

Open ocean absorbs about six times as much energy as fresh snow under the same sunlight. Water can reflect more light at very low solar angles, but it remains much darker than snow or ice in the comparison used here.

The feedback loop

The ice–albedo feedback proceeds in four steps:

1. the Arctic warms;
2. some sea ice melts;
3. darker ocean is exposed, lowering surface albedo;
4. the surface absorbs more sunlight, causing additional warming and melt.

The ice–albedo feedback

• Fresh snow: 0.85
• Bare sea ice: 0.60
• Melt pond: 0.25
• Open ocean: 0.06

This is a positive feedback because it reinforces the initial change. It can operate in either direction: cooling can preserve snow and ice, increase albedo, reduce absorption, and cause further cooling.

Positive does not mean beneficial, and it does not mean that temperature must always rise. It means self-reinforcing.

How large can the effect be?

For a June-to-August comparison at 75°N, the lecture estimates that open water can retain about 1,800 MJ/m² more energy than ice over the summer.

Using about 306 MJ/m² to melt one vertical metre of ice,

$$\frac{1{,}800\ \text{MJ/m}^2}{306\ \text{MJ/m}^2\text{ per m}} \approx 5.9\ \text{m}.$$

The result is an upper limit, not a prediction. The 1,800 MJ/m² estimate is based on energy at the top of the atmosphere; clouds and the atmosphere remove a substantial share before it reaches the surface. Some absorbed energy also warms water or is lost through other fluxes.

The calculation still shows that the darkening effect is large enough to outweigh the latent heat absorbed during melting.

Feedback gain

A reinforcing loop multiplies the original response. In the simplified examples from class:

• if each pass through the loop returns half of the previous warming, the total response approaches twice the initial response;
• if each pass returns three quarters, the total response approaches four times the initial response.

A stronger feedback also magnifies errors in the initial state, which contributes to uncertainty in Arctic projections.

Why the feedback does not run without limit

Several processes oppose or interrupt it:

• winter darkness: no sunlight means no albedo effect for months;
• outgoing radiation: a warmer surface emits more infrared energy;
• winter freezing: exposed water loses heat quickly and can refreeze.

The feedback is strong, but it operates within a seasonal system containing stabilizing processes.

Evidence for the mechanism

A causal explanation should produce observations that can be checked. The lectures identify three fingerprints:

1. satellite observations show declining Arctic surface albedo as ice retreats;
2. buoys and ships measure upper-ocean warming in newly ice-free areas;
3. Arctic amplification is strongest in autumn and early winter, after the ocean releases stored summer heat.

These independent observations support the feedback mechanism rather than merely making it a plausible story.

Seasonal timing

The strongest sunlight and the largest open-water area occur at different times:

• solar input peaks around 21 June;
• open water peaks near the September sea-ice minimum;
• atmospheric warming from released ocean heat is strongest in October and November.

The ocean therefore acts as a heat reservoir:

1. summer sunlight is absorbed by open water;
2. the upper ocean stores the energy;
3. the energy is released to the atmosphere in autumn;
4. freeze-up is delayed;
5. the ice-growth season becomes shorter;
6. thinner ice enters the next melt season.

The atmospheric response can therefore peak after the sunlight has declined.

Melt onset and melt ponds

The date when snow begins to melt matters because it determines when the surface starts absorbing more energy. An earlier melt onset adds days or weeks during the period of strongest sunlight.

Melt ponds accelerate the darkening:

• bare sea ice has an albedo near 0.60;
• a ponded surface can have an albedo near 0.25.

Ponds form before the ice edge retreats, so they can strengthen the feedback while the ice cover still appears continuous. July pond coverage can therefore help indicate how much ice may remain in September.

Where the feedback is strongest

The effect is largest where the surface is changing:

• the central pack remains ice-covered, so little surface switching occurs;
• the open Atlantic is already dark, so no additional switch remains;
• the retreating ice edge repeatedly changes from ice to water.

The same logic applies to melting snow on land and to Greenland's ice surface. Greenland differs because its ice rests on land, so melt contributes to sea-level rise. Melting floating sea ice does not directly raise sea level.

It is one mechanism among several

Surface albedo is the most visible Arctic feedback, but not necessarily the largest. §1 summarizes the broader set of mechanisms behind Arctic amplification. The precise ranking remains a research question, so ice–albedo should not be treated as the sole cause.

Main chain

A small warming melts some ice. The exposed ocean absorbs much more sunlight than the original ice, stores the heat through summer, and releases it in autumn. That delays freeze-up and leaves thinner ice for the following year, reinforcing the original change.