The Earth is a Fridge

No, I’m not a teapot. I’m serious. The way the climate system works is that, over a year, there is a net gain of heat in low latitudes and a net loss at high latitudes. Heat is transported from more tropical regions and radiated away at the poles.

Now, I’ve been mulling over the mystery of why Northern Hemisphere warming (as measured by the mean surface temperature) appears to have slowed over the past decade or so. I suggested a while back that, in view of the rapid industrialisation of China in particular, perhaps renewed global dimming has a role to play.

I recently felt some encouragement to persist from Sue Solomon’s comments in the Guardian recently that:

“…there are climate scientists round the world who are trying very hard to understand and to explain to people openly and honestly what has happened over the last decade.”

And so they should.

Realclimate was a little sniffy about the Guardian’s reporting of the science aspect, with a curious exchange at comment 47, but the (tentative) conclusion seems to be that Solomon’s findings relate to some kind of poorly understood feedback mechanism rather than a climate driver (i.e. an external effect on the climate system).

Back to the story. As I said at the start, the Earth is a giant fridge.

Now, it has suddenly occurred to me that the efficiency of the fridge could be different when the whole system is in a warmer (or cooler) state. If this effect is significant you’d therefore expect periods of more and less rapid warming as the Earth’s ability to radiate away heat changed.

Cutting to the chase, it seems to me that sea ice cover reduces the ability of the planet to radiate heat away; more to the point, loss of sea ice increases its ability to radiate heat away. Ice is a good insulator.

What’s been happening up in the Arctic is that “multiyear” ice has disappeared rapidly over recent years.

Now, if some relatively warm water ends up under some ice that’s already there, at best it can slowly cool to around -2C (when it is in equilibrium with the ice) – because of insulation the ice will not get much thicker. But if, come winter, the sea is not already covered in a layer of ice, the water can cool relatively more and can turn to ice and lose a lot more energy in doing so. Simples. [Actually, it’s not: what may be critical is the amount of surface water that, as it cools, becomes more dense and sinks, allowing heat to be lost from a greater volume of water than at a lower initial surface temperature. The amount of “ventilation” of the water column (by wind) may also be an important factor in determining how much heat can be lost before the insulating ice layer is formed at the surface. Furthermore, Wikipedia notes the process of “brine rejection” whereby water just under the freezing layer becomes more dense (because ice doesn’t incorporate salt) and sinks may also be important – obviously the amount of brine rejection depends on how much freezing occurs each year.].

What I’m suggesting is that the Earth’s refrigeration mechanism will be more efficient the less – in extent and thickness – sea ice there is at the start of winter. This doesn’t mean the planet will start cooling, of course, but it could slow the warming.

I thought I should do a rough calculation to see how much energy it takes to melt the Arctic sea ice each year. The interesting Stoat blog links to some data showing that very roughly 10 million km2 of ice freeze and melt each year.

I’ve seen the nature documentaries, so let’s guess that this ice is on average 1 metre thick.

To melt this ice alone takes 10^7 (the area) *10^6 (to metres cubed) *10^3 (to litres ~= kg) *334*10^3 J (latent heat of fusion of water) = ~3.34*10^21J.

I also happen to know that doubling CO2 will lead to a forcing of around 4W/m2 over the whole planet. 1W/m2 is therefore quite a significant number. How much is 1W/m2 over 1 hemisphere over a year?

The area of the Earth’s surface is ~500m km2, so 1W/m2 of the northern hemisphere is, over 1 year, 250*10^6 *10^6 (converting to m2) *365*24*3600 (a year’s worth of seconds = ~30*10^6) = ~7.5*10^21J.

So, just freezing the Arctic sea ice every year, never mind cooling the water or ice down implies that the Earth radiates away heat equivalent to a continuous forcing of around 0.4W/m2 of the entire surface of the northern hemisphere.

In fact, if we assume the water has to be cooled down as well, that 0.4W/m2 becomes a little bigger (the specific heat of water is around 4J/g/C – i.e. 4J heats 1g by 1C).

Of course, the extra heat loss in winter while the water is cooling and freezing when the ice extent is low needs to be weighed against the extra heat gain in summer by the albedo change due to the absent ice sheet. Looking at it another way, when there’s no permanent sea-ice, the albedo-feedback-assisted summer melting and winter freezing exactly cancel out. Obviously. My point, though, is that there is a circulation and the Arctic cools water that ends up flowing back south as a cold deep current (so it’s the 4J/g/C released when water cools rather than the 334J/g when it freezes that’s important). This mechanism is cut off by the insulating effect of a layer of sea ice. A corollary is therefore that improved Arctic fridge efficiency should strengthen the thermal oceanic circulation. In total, over a year, once it’s warm enough for the sea-ice to disappear in summer, more cold water should sink and flow south than before, thereby allowing more warm surface water to drift north.

There could be an optimum Arctic cooling efficiency when it’s still cold enough for the ice to freeze by the end of the winter (to reduce heat uptake during the early summer) but warm enough to mostly thaw by the end of summer.

In conclusion, I present, in the hope of encouraging progress towards an explanation of the lack of 21st century warming in the northern hemisphere, and to supplement the Renewed Global Dimming Hypothesis, the possibly even more tentative Strengthened Earth Refrigeration Mechanism Hypothesis.

I should repeat what I may term the Warming Warning, that is, that, if underlying warming is being masked, or postponed, by either of these mechanisms and/or others, we could be in for a real shock in later decades.