Musings of the Hemispheres

I should really try to finish one blog post before I start another on a similar topic. My last (published) post noted that the North Atlantic Oscillation (NAO), an atmospheric phenomenon, is not a climate driver, rather it’s a measure of the state of the climate – incidentally, I’m pleased to discover this morning that Philip Eden at holds similar views. In a post that may or may not ever appear, I was going to note similar thoughts about the so-called Atlantic Multidecadal Oscillation or AMO.

It’s not my understanding of the oceanic circulation that great surges of current drive the climate. Rather the oceanic circulation is itself driven by changes in heat distribution at the surface. OK, there may be timelags and of course there’s the El Nino, but that’s about it. If the El Nino is driven by ocean currents (which I believe it is), these are, crucially, east-west, not north-south. The planet loses heat partly because heat moves (in water and air) from the tropics towards the poles where it is more easily radiated away (or used to melt ice). I suggest, therefore, that changes in oceanic circulation are primarily caused by changes in the absorption of heat at the surface. For example, if the planet is warming, you’d expect a general strengthening of oceanic circulation.

What’s piqued my renewed interest in this topic is James Hansen’s release of temperature data for 2009, available over at Realclimate. In particular, Hansen includes separate temperature graphs for the southern and northern hemispheres. I reproduce these here for convenience (the diagrams in my previous post were from Wikipedia, btw):

Fig 1: Annual temperatures
Fig 2: Running mean temperatures

Looking at the right hand graphs, comparing temperature changes in the hemispheres, we see that sometimes the northern hemisphere warms quicker than the southern hemisphere, whereas at other times the reverse is true. What would we expect, though? Well, there’s a lot more water in the southern hemisphere and a lot less land. We’d therefore expect the south to warm (and cool) slower than than the north (and, in the long-term, catch up when temperatures stabilise at a different level). And, indeed, this is what appears to happen most of the time: since the mid 1970s, the northern hemisphere has warmed much faster than the south; on the other hand, the cooling (clearest in Fig 1) caused by the Mt Pinatubo eruption in 1991 was most evident in the northern hemisphere.

But – there’s always a “but” – for significant periods of time (I consider the weather can affect annual data, but not decadal trends) – for example, from around 1950 to the mid 1970s – the southern hemisphere has actually warmed when the northern hemisphere has cooled. This requires explanation.

There are only two possibilities: either the one I’ve already dismissed, that large amounts of heat are, by some unexplained causal mechanism, transferred between the hemispheres, or, that there are factors causing the hemispheres to gain different amounts of heat at different times. Specifically, for several decades from around 1950, the southern hemisphere must have either gained heat, whilst the northern hemisphere lost it, or, more probably, two countervailing factors were involved: one causing a general warming and the other cooling, but disproportionately of the northern hemisphere.

We know that increased levels of greenhouse gases are tending to warm the planet. The inescapable conclusion is that another factor tends to cool the northern hemisphere more than the southern hemisphere. The argument that from the 1940s through to the 1970s this was “global dimming“, caused by sulphur dioxide and other pollution is highly persuasive. Most of this pollution is emitted in the northern hemisphere and doesn’t stay in the atmosphere long enough to spread evenly.

What’s happening now, though?

Well, what strikes me in Hansen’s graphs is the levelling off of warming over the last few years. There’s not yet really enough data to reach any sort of conclusion, but Hansen notes 2009 was the second warmest year on record. In fact, though, his data (Fig 1), suggest it was the warmest year in the southern hemisphere and around the 7th warmest in the north.

Given the rapid industrial development of China and India, it seems justifiable to hold a working hypothesis that we face renewed global dimming.

You would expect a layer (or layers) of reflective particles in the atmosphere to reflect a greater proportion of light from the sun in the winter than in the summer, so another way to test the hypothesis would be to examine seasonal rates of warming over the past century or so. The trouble is, seasonal temperatures are very much affected by poleward heat transport and weather patterns themselves dependent on whether the planet (or hemisphere) is warming or cooling, but nevertheless I’d expect average temperatures in continental interiors (with stable seasonal weather patterns and especially anticyclonic conditions in both winter and summer) at high latitudes to fall more in winter than in summer during periods when global dimming increases, i.e. when the rate of warming of the planet as a whole slows down.

Of course, we could also put a bit more effort into simply measuring the strength of sunlight at the top of the atmosphere (to account for variations in solar output) and at ground level in cloudless conditions (or controlling for cloud cover).

If the analysis that the climate is being affected by renewed global dimming is correct, it’s really bad news. What it implies is that, when the presently industrialising nations reduce their sulphur emissions (and assuming other countries don’t repeat the exercise), we could be in for another period of rapid warming (several tenths of a degree C per decade on average in the northern hemisphere), similar to that over the last quarter of the 20th century.

14/6/10: Minor correction, AMO stands for Atlantic Multidecadal Oscillation, not “Meridional”.