I wrote a little while ago about the Dutch advice to the IPCC. It was quite an interesting document as it appeared quite critical of the IPCC. In particular it suggested that the IPCC should, more explicitly, address natural climate change, rather than focusing on human-induced climate change. This was lept on by some skeptics as an indication of a major problem with the IPCC. Collin Maessen, however, wrote to the Dutch Meteorological Institue who reponded and basically said that the were simply recommending that the IPCC should more explicitly address natural climate change and make it part of their remit. Nothing particularly major then.
However, the issue of natural variability is often raised by those who are skeptical of human-induced climate change. They claim that it appears to be ignored and that climate scientists (and the IPCC) have not explained why what we’re experiencing today cannot simply be due to some kind of natural variability. Now, I don’t believe that scientists are ignoring it, but the skeptics may have a point. It does appear as though the issue of natural variability has not been addressed – in the public debate at least – particularly clearly and it would seem sensible to address it more explicitly in future. I do, however, think that there may be a perfectly good scientific reason why it has typically not been addressed particularly clearly in the public realm. The reason is that it really can’t explain (or be the reason for) what we’re currently undergoing.
Let me see if I can explain why. Bear in mind that I’m a physicist and not a climate scientist. Also, these are just some of my musings so I may well overlook something, or make some silly mistake. The first thing to realise is that the Earth gets most (virtually all) of its energy from the Sun. The surface temperature will tend towards a value at which we are radiating as much energy back into space as we receive from the Sun. We’ll call this the equilibrium temperature. There are 3 primary factors that determine what this equilibrium value is. One is the amount of energy we intercept from the Sun (the Total Solar Irradiance, TSI); another is our albedo (what fraction of this energy we reflect directly back into space); the third is the composition of our atmosphere. Why is our atmosphere important? Well, the energy we get from the Sun is primarily in the visible band and our atmosphere is largely transparent at these wavelengths. Apart from the fraction reflected, the rest passes through our atmosphere to be absorbed by the surface (and some by the atmosphere). The Earth is much cooler than the Sun and so re-emits at longer wavelengths. Our atmosphere is not transparent at these wavelengths and so much of this is absorbed and essentially trapped by the atmospheric greenhouse gases. This causes the surface temperature to rise until it reaches a value at which as much is escaping into space as we receive. There are a few other factors which Tom Curtis explains very clearly in this comment, but these other factors aren’t particularly important.
That’s the first thing to realise. Another thing to realise is that the heat content of the atmosphere and surface is quite low. The atmosphere has a mass of 5 x 1018 kg. The mass of the surface is tricky to determine (as far as I can tell) but is quite low. Let’s assume, for simplicity, that the total mass of the atmosphere and surface is 1019 kg. The specific heat capacity is 1000 J kg-1 K-1. This means that to increase the average temperature of the surface and atmosphere by 1oC (1 K) would require 1022 J. The rate at which the surface loses energy (J s-1) is
L = 4 π R2E σ T4,
where RE is the radius of the Earth, σ is the Stefan-Boltzmann constant, and T is the average surface temperature. If T = 290 K, the surface is losing energy at a rate of 2.04486 x 1017 J s-1.
Let’s consider a hypothetical situation where some natural event increases the average temperature of the surface and atmosphere by 0.1oC. This will increase the heat content of the atmosphere and surface by 1021J. Since this changes the surface temperature so that T = 290.1, the rate at which the surface loses energy is now 2.04768 x 1017 J s-1, 2.821 x 1014 J s-1 faster than when T = 290 K. This means that it will lose this excess energy (1021 J) within about a month. According to my calculation, this doesn’t really depend on how big the perturbation is. Even if the temperature is increased by 0.5oC, it will still only take about a month for the temperature to return to its equilibrium value. The same is true if the temperature were to drop slightly. It would return to equilibrium very quickly.
So, why is this relevant? Well there’s a few things we can conclude. Since the mid-1800s the surface temperature has increased by about 1oC. This can’t simply be a slow recovery towards equilibrium because as I’ve shown above, that should happen quickly. It can’t take a century. It also can’t be a series of natural events that just happen to have warmed the atmosphere and surface because these would have to have happened almost every week in order for the energy from the previous event not to have been lost into space before the next event occurs. For example, it can’t be due to ocean cycles such as ENSO events. These can indeed bring energy from the ocean to the surface where it can heat the atmosphere and land, but they only occur every few years and there are both heating (El Nino) and cooling (La Nina) phases. Furthermore, if it were due to ENSO cycles we would expect the ocean heat content to be dropping as the energy were transferred from the oceans to the atmosphere. Instead, it is rising.
So, we can conclude that the change in surface temperature since the mid-1800s has to have been accompanied by a corresponding change in our equilibrium temperature. The equilibrium temperature must be about 1oC higher today than it was in the mid-1800s. Are there any natural processes that could do this? Well, one obvious candidate is the Sun. It clearly influences our climate and plays a key role in setting our equilibrium temperature. The TSI does indeed vary. There is an 11-year solar cycle, but this shouldn’t produce a century-long increase in surface temperatures. Furthermore, it’s only associated with a small change in surface temperature. There are much longer cycles and indeed the TSI did rise during the first half of this century. However, this too was quite a small change (and so should only have produced a small change in surface temperature) and it’s been falling since about 1970. The Sun alone can’t explain what we’re currently undergoing.
What about volcanoes? They do indeed influence our equilibrium temperature, but tend to have a cooling effect. They eject aerosol particles into the atmosphere which slightly increases the albedo, directly reflecting more sunlight back into space. Also, these aerosol particles tend to precipitate out and so only influence the equilibrium temperature for a few years.
What about water vapour? It is indeed a greenhouse gas and so if the water vapour concentration has increased it would act to increase our equilibrium temperature. One problem with this is that the amount of water vapour that the atmosphere can hold depends on the atmospheric temperature. Any excess should precipitate out quickly. Also if some natural event could increase the atmospheric water vapour concentration so as to produce a 1oC surface warming in a century, it would be difficult to explain why our past climate appears to be so stable.
Could the albedo have changed? There seems to be no evidence for this. This would probably require that there was a change in the area of polar ice and snow cover and this hasn’t been observed. Maybe more correctly, it didn’t appear to start changing significantly until about 1960 and the changes we’re seeing now are almost certainly anthropogenic. Also, if it isn’t anthropogenic, what would be the driver. Ice doesn’t just simply decide to melt.
Clouds? Clouds can play a role in changing our equilibrium temperature, but can both heat and cool. They can influence the albedo and reflect more incoming radiation (cooling) but also can trap more outgoing long-wavelength radiation (heating). I will accept that the role of clouds is quite uncertain, but it does seem unlikely that they could explain the surface warming seen since the mid-1800s.
One could argue that maybe they’re all linked. Maybe a small increase in TSI can slightly increase atmospheric temperatures. This can allow more water vapour into the atmosphere which further increases temperature. This could lead to more clouds and to the melting of polar ice and snow which then changes our albedo. Maybe, but there are some fairly fundamental problems with this. TSI is thought to be about 0.25 Wm-2 greater today than it was in the mid-1800s. An increase of 1oC in surface temperatures means that the surface is radiating about 5.5 Wm-2 more today than it was in the mid-1800s. These linked effects would therefore have to essentially amplify a small change in TSI so as to produce a large change in surface flux. The amplification (or feedback) factor would need to be about 20. This is extremely high and, again, would make it hard to explain why our past climate appears to have been so stable. If it’s so sensitive to small changes in solar flux, surely we would have detected this in the paleoclimatological record.
Anyway, this has all got rather long. As I said at the beginning, these are just some of my musing and I am no expert at climate science. I do think that addressing natural variability more openly is a good thing. I also understand, I think, why this may not have happened in the past – it can’t really explain what we’re currently experiencing. I hope this post makes some kind of sense and – as usual – happy to take comments from anyone, but in particular from those who can correct any of my mistakes/misunderstandings.