Global temperature is affected by a lot of things, not just humans and their greenhouse gases. We know what some of those things are, and we can even estimate their impact over the last 40 years or so (since 1979 let’s say). Then we can subtract that effect from temperature data, to estimate how hot Earth would have been without those “other things.”
What other things, you wonder? Known factors include the El Niño Southern Oscillation (ENSO), volcanic eruptions, and variations in the output of the sun; published research outlines the method I used (although I’ve tweaked it a bit). Let’s take, for instance, the global temperature from NASA (monthly averages from January 1979 through December 2019):
The red line shows the linear trend (estimated by least squares regression), temperature rising at an average rage of 0.0188 (± 0.0031) °C/yr. But temperature itself (the thin black line) doesn’t just follow the trend, it fluctuates up and down. Quite a lot.
According to the analysis, those other factors were responsible for quite a bit of the fluctuations, influencing temperature in this way:
I’ve included a trend line here, too, but this one slopes downward, very slightly, at a rate of -0.0003 (± 0.0021) °C/yr. That’s -2% of the overall trend since 1979. That leaves the part of the trend that is probably man-made, at 102% of the observed trend.
When we do subtract the “other factors” estimate from the observed temperature data, we get “adjusted” data that better reflects the man-made part of climate change, now that many of the natural fluctuation factors are removed:
It shows a lot less fluctuation than the raw data, and promises us a better estimate of humanity’s influence on the trend, heating up at a rate of 0.0192 (± 0.0016) °C/yr. The part of global warming not due to those “other things” (which is, for the most part, the human influence) is a tiny bit bigger even than we’ve seen. Do note, however, that when it comes to trend the difference is not “statistically significant.”
It’s interesting to look at what the individual factors themselves have been up to. Here’s the estimated strength of each, with the solar influence shown in blue, the volcanic effect in brown, and the impact of ENSO in red:
The solar influence isn’t very big, and its trend is truly tiny, only about -4% of the observed trend. It’s not because the sun can’t affect Earth’s climate, it’s just that the sun’s output hasn’t changed that much.
The ENSO impact goes up and down, but its overall trend during this time span has been downward, at an estimated rate of -0.0008 (± 0.0007) °C/yr. That’s -8% of the observed temperature trend during this time span.
The volcanic effect is dominated by two large eruptions, El Chicon in the early 1980s and Mt. Pinatubo in the early 1990s. Because they cooled the Earth noticeably, the trend in the volcanic impact is upward, at +0.0020 (± 0.0018) °C/yr. That’s 10% of the observed trend.
The bottom line is that yes, those “other things” do affect trend estimates, with volcanoes causing extra heating while ENSO caused extra cooling (with a little help from solar), but the combined effect is just about zero.
We can repeat the exercise with, say, the lower-troposphere temperature from UAH (Univ. of Alabama at Huntsville). I’m not fond of that data set, but let’s see what it has to say. By the same analysis, since 1979 ENSO decreased the trend, changing it by -19%, solar by a further -8%, while volcanoes increased the trend by 27%. Again, the net result is about zero.
It turns out that the “other things” don’t show any trend long-term, and in the not-too-long term (since 1979) have trended oppositely to each other, with a net effect of just about zero. The trend we’ve got now isn’t because of those things.
I have the impression that Roy Spencer wants to blame global warming (since 1979) on volcanoes.
These volcanic effects on the post-1979 warming trend should always be kept in mind when discussing the post-1979 temperature trends.
What he does is compute the trend contribution from volcanoes (alone). He doesn’t use surface temperature, or lower-troposphere temperature, he uses ocean temperature, not from observations but from his climate model. I confess I don’t have much confidence in his calculations, but he ascribes 40% of global warming to volcanoes.
I think his implication is “that only leaves 60% due to mankind, tops.” I think he’s confused.
Thanks to all who kindly donated to the blog. If you’d like to help, please visit the donation link below.
This blog is made possible by readers like you; join others by donating at My Wee Dragon.
Open Mind 
In case folks are curious: below are some more recent sources doing analyses removing the effect of non-anthropogenic factors, to show the dominant anthropogenic trend. They thus further undermine the point Spencer aims for. The first couple of sources look at the satellite-based bulk tropospheric analyses like UAH, while the rest look at surface analyses (I’m not a huge fan of the last two papers listed):
10.22499/3.6701.001 , 10.1002/wcc.511 ;
10.1175/JCLI-D-18-0555.1 , 10.1126/sciadv.aao5297 , 10.1029/2018JD028776 , 10.3390/cli6030064, 10.1002/2017GL072908
I said this above:
“(I’m not a huge fan of the last two papers listed):”
Frank Bosse, I know you’re reading this, since you lied about what I said there, as part of your comments on the blog article you wrote for Judith Curry on climate sensitivity.
“Atomsk, you state on other place that you are a “huge fan” also of this paper”
I also understand that Judith Curry removes criticisms that point out your lies, in order to protect you. I guess that’s what’s to expected from the likes of you denialists.
The rise in Roy’s UHA graph https://www.drroyspencer.com/latest-global-temperatures/ seems to correlate well with his increasing desperation to ascribe the cause of global warming to anything but CO2.
There are other things like brightening. Pollution has been reduced drasticly. S a greater part of available soalr energy on top of the atmosphere reaches the surface. My own research shows that brightening is stronger then the greenhouse effect. Brightening is antropogenetic as well. Should this not be mentioned in the list of causes of AGW?
“My own research shows ….”
It doesn’t show anything until you show the world detailed calculations. How about starting here?
Tamino
You wrote above:
I have the impression that Roy Spencer wants to blame global warming (since 1979) on volcanoes.
I think you misunderstood what Roy Spencer meant.
Look at this graph:
What he means isn’t that volcanoes contributed to warming during the sat era. He means that their cooling contributed to an artificial elevation of UAH’s trend, what in turn means that the current trend is due to the volcano eruptions, rather than to Mankind.
And that is, as you know, the most preferred drug for some of his blog visitors, and above all for the vast majority of Anthony Watts’ WUWT readers and commenters (such Spencer threads are automatically reblogged at WUWT). They all love him for that.
A few lines further, he writes, in the same vein:
It is interesting to see the “true” warming effects of the 1982-83 and 1991-1993 El Nino episodes, which were masked by the eruptions. The peak model temperatures in those events were only 0.1 C below the record-setting 1997-98 El Nino, and 0.2 C below the 2015-16 El Nino.
From his point of view, that is of course correct. We clearly can see, for El Chichon and Pinatubo, the two stratospheric highs and tropospheric lows in the UAH LS and LT records:
https://drive.google.com/file/d/1OO6HpUOvk_N_tC2fUt8wzDDvMzhYM8C_/view
Best regards
Jean-Pierre
@bindidon,
And, naturally, Dr Spencer insists that to be “objective”, it is legitimate to completely ignore the strong priors we have on such observations from ab initio physical theory, confirming laboratory experiments, and confirming successfully engineered systems.
Right.
I believe that the IPCC says that 110% of warming is human caused. In other words, we should be cooling now if not for human interference in the climate.
There is a localised indicator that the lack of volcanic action is causing global warming is a false premise.
The Giant Sequoias of North America are under attack from a beetle which presumably has always affected them but apparently has now become lethal. Over the trees 3000 year lives they would have lived through similar low volcanic activity periods and other naturally occurring stresses. However it is apparently only in the present that something else is combining with the natural background of stress to cause them lethal interactions with pests.
https://www.theguardian.com/environment/2020/jan/18/this-is-not-how-sequoias-die-its-supposed-to-stand-for-another-500-years-aoe
https://soundcloud.com/user-880618252/dr-christy-brigham-sees-lazarus-for-the-first-time.
Surely the neutral observer must start realising that localised stressed environment impacts are happening on a world wide scale that are beyond the scope of those environments to cope.
Thanks for your continued blogging Tamino, Always interesting to come and see how staistics should be analysed.
@Dave North-Lewis,
For various exogenous reasons, I’ve just dived into a self-assigned study of the Imbens-Rubin textbook Causal Inference for Statistics, Social, and Biomedical Sciences (Cambridge, 2015) after having taken Dr Beth Ann Griffin‘s short course “Estimating propensity scores and inverse probability of treatment weights when drawing causal inferences” last November through the Boston Chapter of the American Statistical Association.
Imbens-Rubin considers a comprehensive swath of causal inference, but the deal with things like propensity scores is it offers a means of establishing causation from observational rather than experimentally arranged datasets.
My point is that some day I hope someone will take one applying these methods to the various purported causes of climate disruption and demonstrate, as Professor Richard Alley says, Carbon Dioxide is “the biggest control knob” of Earth’s climate.
We kind of know it is now, but these causal inference methods make it nearly as strong as if the experimental design was set up and done.
Is the apparent recent acceleration in temperature significant? GISTEMP trend is 0.024ºC/yr over 2003-2019. Breakpoint?
[Response: See my latest blog post here:
.]