Global warming accelerates to 0.3 degrees per decade

 From Open Mind (Tamino)

July 2023 was the hottest month in history, maybe even the hottest in the last 120,000 years. The months that followed broke monthly records by surprisingly large margins. It’s no wonder that 2023 has turned out to be the hottest year we’ve seen, and not just by a little.

Here are yearly average temperature anomalies for the whole planet, from 1950 through 2023, according to HadCRU, the Hadley Centre/Climate Research Unit in the U.K. (they haven’t yet reported their value for December 2023, so that year’s value is the January-through-November year-to-date average):

Each dot shows a year’s average, with the solid red line a smoothed version. The smooth shows is a good estimate of the background value, and reveals how climate has changed (temperature-wise), while the actual yearly values (black dots) fluctuate about that background value.

We know some of the things that contribute to those fluctuations. Volcanic eruptions, for instance, can fill the atmosphere with reflective aerosols which take years to settle out of the air, all the while cooling the planet. The el Niño southern oscillation (ENSO) tracks changes in how the ocean and atmosphere exchange heat; in its el Niño phase the surface gets noticeably warmer, but in its opposite la Niña phase the world tends to be cooler. The sun, of course, is the ultimate source of heat for our climate, and when it shines hotter, Earth heats up, but when the sun cools (relative to itself, of course) the Earth follows. These factors can be accounted for, at least approximately, and removed from the temperature data to define adjusted data, which will (we hope) give us a clearer picture of the changes which are due to other things — like greenhouse gas emissions.

Such has been done in peer-reviewed scientific research, and I’ve updated the procedure to include more recent data as well as alternative methods. In my latest version, the estimated contribution of each factor to the HadCRU temperature series looks like this:

The influence of volcanic eruptions, in brown, shows the cooling from the three large volcanic eruptions during this time, especially the massive explosion of Mt. Pinatubo in the 1990s. Solar variations, in red, have had only a small influence on Earth’s temperature because the solar variations themselves are small. The ENSO, in blue, has sometimes warmed (the el Niño phase) and sometimes cooled (the la Niña phase).

Add them all together, we have an estimate of the total influence of these factors on global temperature:

 

We can use this to estimate what global temperature would have been without these factors; just subtract this estimate from the observed data to yield adjusted data. When I do so, then compute yearly averages to yeild adjusted yearly average temperature anomalies for the whole planet from 1950 through 2023 according to HadCRU, I get this:

 

Each dot shows a year’s average, with the solid red line showing a smoothed version.

I’m mainly interested in the rate at which temperature is rising, so I took the original (un-adjusted) data and analyzed it with my modified lowess smooth, which is what I used to generate the solid red lines (smoothed values) in the above graphs. I’ve custom-designed it for this kind of analysis, so it reports, for each moment of time, not only the estimated value and its uncertainty, but the estimated rate of change (just what I’m looking for) and its uncertainty.

When I apply it to the HadCRU data, and choose the “time scale” (a free parameter in the analysis) so it will mimic rates of change on a 20-year time scale, it yields this:

The solid blue line shows the estimated rate of warming, the light blue shading shows its uncertainty range (±2σ). There is an obvious change in the rate around the year 1970; before that the rate was near zero (within the uncertainty range), but afterward is assuredly positive. This agrees with what we see in the first graph, that the smoothed value changes very little (zero rate from 1950 to 1970), then rises steadily until the present (rate about +0.02 °C/yr until now).

There is also a rise in the main estimate (solid blue line) starting around 2010, which suggests that the rate of warming may have increased lately (i.e. recent acceleration), but the uncertainty range is wide enough that it still includes the value +0.02°C/yr which had been observed already for decades.

Conclusion: since 1950, the data show at least two different warming rates: near zero from 1950 until about 1970, then about +0.02 °C/yr until now. There is evidence of a recent increase, but the evidence is inconclusive.

What about the trend in the adjusted data, i.e. apart from the factors that make those incessant fluctuations? I can apply the same analysis and get this:

 

Much is essentially the same as with the original, raw data: there is undoubtedly a change in rate around 1970, and there is evidence of another change around 2010. But this time the uncertainty range is narrower, the uncertainties are a lot smaller, and the evidence for recent change is now conclusive.

Conclusion: since 1950 the adjusted data show at least three different warming rates: near zero from 1950 until about 1970, then about +0.02 °C/yr until around 2010, and about +0.027 °C/yr since. Not just the above analysis, but other statistical tests confirm that although the uncertainty in the current rate is considerable, we conclude with confidence that it’s faster than it was during the preceding decades. Global warming picks up speed.

That’s using the data from HadCRU, and the story is the same when using data from NASA (the GISTemp data from the Goddard Institute for Space Studies), from NOAA (the National Oceanic and Atmospheric Administration), from the Berkeley Earth surface temperature project, or the ERA5 data from Europe’s Copernicus Climate Service.

James Hansen and others published a new paper recently, claiming that not only will global warming, in the very near future, proceed faster than expected, it is already doing so — that the pace of global warming had accelerated. Temperature increase after 2010, it suggests, will be at 0.027°C/yr, 50% faster than the lazy 0.018°C/yr it had been rising for decades before that. As a result, we have less than a decade until we cross the much-discussed threshold of 1.5°C above pre-industrial, so any idea of keeping global warming below that limit is “deader than a doornail.”

 

[Read more here; Tamino lists some of the caveats of his analysis.]

We are still not doing enough to slash emissions.  Remember, even if emissions peak this year, the rise in temperatures is proportional to the level of emissions.  For temperatures to stop rising, we need to cut emissions to zero.   On top of which, even if CO2 emissions have peaked,  methane emissions have accelerated.  Methane is 82 times as potent a greenhouse gas (over 10 years) as CO2.