CO2 and CERES

Guest Post by Willis Eschenbach

The Intergovernmental Panel on Climate Change, the bureaucratic agency which appropriated the role of arbiter of things climatic, has advanced a theory for the lack of warming since the turn of the century, viz:

The observed reduction in warming trend over the period 1998–2012 as compared to the period 1951–2012, is due in roughly equal measure to a cooling contribution from internal variability and a reduced trend in radiative forcing (medium confidence). The reduced trend in radiative forcing is primarily due to volcanic eruptions and the downward phase of the current solar cycle. However, there is low confidence in quantifying the role of changes in radiative forcing in causing this reduced warming trend.

So I thought I’d look at the CERES dataset, and see what it has to say. I started with the surface temperature question. CERES contains a calculated surface dataset that covers twelve years. But in the process, I got surprised by the results of a calculation that for some reason I’d never done before. You know how the IPCC says that if the CO2 doubles, the earth will warm up by 3°C? Here was the question that somehow I’d never asked myself … how many watts/m2 will the surface downwelling radiation (longwave + shortwave) have to increase by, if the surface temperature rises by 3°C? 

Now, you’d think that you could just use the Stefan-Boltzmann equation to figure out how many more upwelling watts would be represented by a global surface temperature rise of 3°C. Even that number was a surprise to me … 16.8 watts per square metre. 

Figure 1. Blue line shows the anomaly in total downwelling surface radiation, longwave plus shortwave, in the CERES dataset, March 2000 to September 2012. Red line shows the trend in the downwelling radiation, which is 0.01 W/m2 per decade. Gray area shows the 95% confidence interval of the trend. Black line shows the expected effect of the increase in CO2 over the period, calculated at 21 W/m2 per doubling. CO2 data are from NOAA. Trend of the expected CO2 change in total downwelling surface radiation is 1.6 W/m2 per decade. CO2 data from NOAA. 

But as they say on TV, wait, there’s more. The problem is, the surface loses energy in three ways—as radiation, as sensible heat, and as the latent heat of evapotranspiration. The energy loss from the surface by radiation (per CERES) is ~ 400 watts per square metre (W/m2), and the loss by sensible and latent heat is ~ 100 W/m2, or a quarter of the radiation loss.

Now, the sensible and latent heat loss is a parasitic loss, which means a loss in a heat engine that costs efficiency. And as any engineer can testify, parasitic losses are proportional to temperature, and as the operating temperatures rise, parasitic losses rise faster and faster. In addition, the 100 W/m2 is the global average, but these losses are disproportionately centered at the hot end of the system. At that end, they are rising as some power factor of the increasing temperature.

But let’s be real generous, and ignore all that. For the purpose of this analysis, we’ll swallow the whopper that a 3° temperature rise wouldn’t drive evaporation through the roof, and we’ll assume that the parasitic sensible and latent heat losses from the surface stay at a quarter of the radiation losses.

This means, of course, that instead of the increase of 16.8 W/m2 in downwelling radiation that we calculated above, we need 25% more downwelling radiation to account for the parasitic losses from the surface. (As I said, the true percentage of parasitic losses would be more than that, likely much more, but we’ll use a quarter for purposes of conservative estimation.)

And what that means is that if the IPCC claim of three degrees of global warming per doubling of CO2 is true, when the top-of-atmosphere radiation goes up by a doubling of CO2, an additional TOA 3.7 watts per metre squared, the surface downwelling radiation needs to go up by no less than 21 W/m2 per doubling. And although I was surprised by the size of the number, to me was very good news, because it meant that if it were there, it should be large enough to be quite visible in the CERES data. So I took a look … and Figure 1 above shows what I found.

The red line shows the trend over the ~ 13 years of the record … which is 0.01 W/m2 per decade, statistically no different from zero. 

The black line, on the other hand, is the change in downwelling radiation expected from the change in CO2 from 2000 to 2012, calculated at 21 W/m2 per doubling of CO2. As you might imagine because of its steady increase, there is little difference between the CO2 data and the CO2 trendline, so I’ve left it off. For the same reason, there is virtually no error in the trend in downwelling radiation expected from CO2. The result is an expected increase in downwelling surface radiation of no less than 1.6 ± 0.007 W/m2 per decade. Over the period of the CERES data, it totals almost 2 W/m2, which in terms of the precision of the individual CERES datasets should certainly be visible.

So … does Figure 1 falsify the CO2 hypothesis? Not yet, we’ve got a ways to go, but it is an interesting finding. First, we need to look at the two explanations postulated by the good folks at the IPCC that I quoted at the head of the post—volcanoes and solar variations. And the amount that we are looking to explain is a missing increase of 1.6 W/m2 per decade.

Their first explanation was solar. Since the downwelling surface radiation has not increased as expected, perhaps there’s been a decrease in the incoming TOA solar radiation. This would offset a warming from CO2. Here’s that data:

Figure 2. Trend in TOA Solar Radiation, 2000-2012. Red line shows trend, a decrease of – 0.15 W/m2 per decade.

So the IPCC is right about the solar. And from having to explain 1.6 W/m2, we’ve explained 0.15 W/m2 of it which leaves 1.45 W/m2 of missing warming.

Next, volcanoes. The IPCC says that the effect of volcanoes over the period was to cut down the amount of sunshine hitting the surface, reducing the total downwelling radiation.

The reduced trend in radiative forcing is primarily due to volcanic eruptions …

Here are the anomalies in that regard:

Figure 3. Action of volcanoes in reducing surface solar radiation. This measures the anomaly in downwelling solar at the surface minus the anomaly in downwelling solar at the TOA. The trend in the transmission is a warming of +0.34 W/m2 per decade.

Bad news for the IPCC hypothesis. Rather than volcanoes counteracting the expected warming and decreasing the atmospheric transmission of sunshine over the period of record, we had a trend of increasing amounts of sunlight making it to the surface. The trend of this increase was 0.34 W/m2 per decade. Kinda blows holes in their theory about volcanoes, but all we can do is follow the data …

And as a result, instead of having to explain a missing warming of 1.6 – 0.15 = 1.45 W/m2 per decade, we now have to add the 0.34 W/m2 to the missing warming, and that gets us up to 1.8 W/m2 in missing warming. So rather than explaining things, overall the IPCC explanation just makes things worse …

Anyhow, that’s how it goes to date. If the IPCC theory about 3°C surface warming from a doubling of CO2 is true, we need to either a) come up with something else in the CERES data to explain the missing CO2 warming of 1.6 W/m2 per decade, b) back off on the IPCC climate sensitivity by a factor of about ten … or my perennial favorite, toss out the idea of “climate sensitivity” entirely and recognize that at equilibrium, temperature isn’t a simple function of TOA forcings because the climate system has emergent phenomena which respond and react to counteract the TOA changes.

The big problem that I see for the hypothesis that GHGs rule the temperature is that over the period of the CERES data, we should have seen a shift of almost two watts in the downwelling total radiation … but I find no such thing in the dataset. So I throw this question out to the climate science community at large.

Where in the CERES data is the missing warming? There is no trend (0.01 W/m2 per decade) in the surface downwelling radiation. The IPCC says that over the period, CO2 should have increased the downwelling surface radiation by ~ 2 W/m2. SO … if the IPCC hypothesis is correct, what is countering the expected increase of ~ 2 W/m2 in the downwelling surface radiation due to the increase in CO2 over the 2000-2012 time period?

Solar explains perhaps 10% of it, but the volcanoes push it the other way … so why can’t I find the two watts per square metre of expected CO2 warming in the CERES dataset?

w.

NOTES

USUAL REQUEST: If you disagree with something that I or someone else said, please QUOTE THE EXACT WORDS YOU DISAGREE WITH. Then, and only then, let us know what you disagree with. I can defend my own words. I cannot defend your interpretation of my words.

DATA AND CODE: I’ve put the data and code used to produce the graphs and calculations online. There are three code files: CERES Setup.R, CERES Functions.R, and the code for this post, CO2 and CERES.R. In addition, there are two datafiles, one for the CERES TOA files, and the other for the CERES surface files, entitled CERES 13 year  (230 Mbytes), and CERES 13 year surface (112 Mbytes). I think that the data is turnkey, just pull up the CO

All of them need to be in the same folder, because the CO2 and CERES.R file calls the setup file, which loads the data files and the function file. If you’ve downloaded the CERES 13 year file, it is unchanged, no need to reload. Open the CERES Setup.R file to see the names of all of the datafiles loaded, and open the CERES Functions.R file for functions and constants.

And as Steven Mosher recommended to me, use RStudio as your portal into R, much the best I’ve found.

CERES Data: The top-of atmosphere CERES data is measured by the satellites. On the other hand, the CERES surface data is calculated from the TOA CERES data, plus data from the MODIS and GOES satellites. The calculated surface data is energy balanced, meaning that the surface flows sum up to the TOA flows.

I’ve run my own version of ground truthing on the CERES surface data by comparing it to the surface temperature data I was using previously. Differences were small overall, and both sets shows the same small details and fluctuations.

Is this how I’d like to do the analysis? Not at all. I’d rather that everything were measured … but this is the best we have, and the various climate scientists involved have used all of the available observational data from a variety of satellites to determine the various values, and have ground truthed the surface data in a variety of ways. So until we have better data, the CERES datasets are the closest we have to actual measurements … and as near as I can tell they show no sign of the claimed 2 W/m2 increase in downwelling radiation that we are assured is going on over the period of record.