Another “escalator“

Guest post by Jan Zeman

Some CAGW proponents argument against the recent stall in the global warming trends with this graph called “escalator”. Source: www.skepticalscience.com

But one of my favorite “escalators” is this one:

Source: http://www.woodfortrees.org

The beginnings and the ends of the global sea surface temperature (SST) trends – the colored lines – are the time centered solar minima and maxima – quite clearly follow the solar signal – except the last: a quite apparently downward(!) slope from the minimum to the current maximum period of the solar cycle SC24 – e.g. until the most recent Hadsst2gl data available.

…and some say the trends shorter than 30 years don’t tell anything about the climate and its drivers… 

It quite consistently looks like the sea surface temperature anomaly trends in the last ~half of the century more or less follow the rises and drops of the solar activity during the solar cycles, except the last trend since the beginning of the SC24, where the temperature trend goes down, although the solar cycle was on the rise – but it appears to agree with the really considerable descent of the solar activity since the peak of the SC22 and especially after the peak of the SC23, only with a minor lag.

I would like to note that the heat capacity of just the upper ~3.2 meters of the ocean water out of the several kilometers deep ocean is the equivalent of the whole atmosphere’s heat capacity, so the global sea surface temperature anomaly looks like it is even better indicator of the solar activity’s influence on the Earth surface heat budget and temperatures than the global air surface temperature anomaly.

Let’s have look at the trends for the same periods using HADCRUT4GL data for comparison:

Source: www.woodfortrees.org

Again the global surface air temperature trends’ direction more or less follow the solar cycles up or down, up until the peak of the SC23. After which there is an anomaly – first the trend goes up while the solar activity descends and then it goes down while the solar activity rises. Which I propose could be attributed to a transient lag in the periods when the solar activity trends abruptly change as in our case after the SC22 peak and especially after the SC23 peak. (The SSN averages are in SC21 81.16, in SC22 80.63, in SC23 53.92, and in the SC24 at its peak period is so far 34.36 and it will yet fall significantly.)

All real thermodynamic systems, especially those involving significant latent heat exchanges – as in our case with the ice melting and evaporation (both from the sea surface and land) – have some thermal inertia. The question is only how big its effect is on the surface temperature anomalies.

Let’s yet check the same periods with the GISTEMP data:

Source: http://www.woodfortrees.org

We can see quite a similar pattern as with the HADCRUT4GL data.

…some say the sun does not have major influence on the surface temperatures (– sometimes they say at least since ~mid 20th century – which seems to me a bit like a contradiction: Sometimes influences, sometimes not? Such a hot giant as our sun, delivering most if not practically all the heat to the Earth’s surface?)

I don’t think so. The solar activity measured in sunspot number obviously correlates well with the TSI and it correlated quite well with the surface temperature anomalies throughout most of the record up until the end of the 1970’s too. We can see it prima facie:

Source: http://www.woodfortrees.org

The only question in my opinion is how fast the solar activity influences the global surface temperature anomalies when the solar activity trends relatively abruptly change (– as in the last two solar cycles) and transient phenomena take place.

The visual comparison of the trend graphs (- the above SSN v. SST, HADCRUT4GL and GISTEMP) also seems to provide a clue that the changes of solar activity could influence the sea surface temperature anomaly a bit faster than it influences the surface air temperature anomaly. Which is what one might expect (anti-intuitively): In my opinion it is caused by the fact that the epipelagic zone (the “sunlight zone” below the ocean surface up to ~200m depth) of the sea has more than 50 times higher heat capacity than whole the atmosphere. Therefore it always traps much more solar radiance converting it to heat than the atmosphere*.

This massive reservoir of sea surface heat** moreover mostly stays on the top, because most of the ocean surface water has lower density than the water below. The waters are mixed by wind and waves only to quite shallow depths. The heat gets into the depths of the ocean mainly by the thermohaline circulation, and it takes quite a long period of time for them to get the heat into the ocean depths. Some estimate this is taking hundreds to thousands of years (see slide 29 here). Otherwise the heat from the ocean’s surface propagates into deeper ocean layers by thermal conduction. Liquid water however does not have very high thermal conductivity, so it also takes considerable time to change the temperature equilibrium state this way, when the long-term solar irradiance/heat input trends and also the possible inducted cloudiness/albedo trends change (as proposed by H. Svensmark and others). So there quite likely can be lags of the surface temperature anomalies trends behind the solar activity trend changes. The question in my opinion is just how long the lags are.

The Occam’s razor principle says “that among competing hypotheses, the one that makes the fewest assumptions should be selected” or in other words: “simpler hypotheses about nature are more likely to be true”.

The average total solar irradiance per time descended quite sharply during the SC22 and SC23 with the pace of ~0.4W/m² (SC22) and ~0.7W/m² (SC23) per solar cycle*** and quite apparently continues to further significantly descend in the SC24. Similar it is with the sunspot number, which looks to touch the Dalton minimum levels****. Do you really think this will not have a significant impact on the surface temperatures in the future?

You decide.

* this underlines the fact that the sea surface water has also higher average temperature (the global average sea surface temperature is about 290 Kelvin) than the global average surface air temperature (~287 Kelvin) and is much higher than the average temperature of the atmosphere (254.3 Kelvin is the blackbody temperature of the Earth’s atmosphere which well agrees with the average temperature obtained by the standard atmosphere model). But is also good to note, that the constructs of the global average temperatures and their anomalies respectively have big uncertainties (estimated as high as ±0.46 degrees Celsius), that it poses serious question how significant the warming trend last hundred years of like ~0.72 (HADCRUT4GL) or ~0.77 (GISTEMP) degrees Celsius per century really is. But this is not the topic of this my article.

** continuously and distinctively heating the surface air wherever its temperature is lower and cooling it wherever its temperature is higher, while the water also evaporates from the surface, mainly due to direct heating effect of the solar irradiance on the water surface’s skin able to “knock out” the water molecules into the air.

*** just for illustration see the trends here – note: the PMOD values must be corrected according to this TIM/PMOD correlation, so in reality the SC23 trend (green) is up to ~0.05W/m² per solar cycle less steep then the graph shows.

**** especially if we use the sunspot number correction proposed by L. Svalgaard – see the slide 8 here)