No, The Carbon Sinks Aren’t Sinking

Guest Post by Willis Eschenbach (@WEschenbach on eX-Twitter)

The usual font of misinformation, the Guardian, has an article claiming the following:

In 2023, the hottest year ever recorded, preliminary findings by an international team of researchers show the amount of carbon absorbed by land has temporarily collapsed. The final result was that forest, plants and soil – as a net category – absorbed almost no carbon.

Hmmm, sez I as my bad number detector starts ringing … sounds fishy. Something on the order of 55% of emitted CO2 is sequestered and 45% remains airborne. Of this, some 75% is sequestered on land. So if land sequestration has “collapsed”, we should see an immediate jump in airborne CO2 of about 75% * 55% ≈ 41%. But I don’t recall seeing that in the data … hmmm.

And indeed, this jump in atmospheric CO2 is what the “preliminary findings” paper linked above uses as the basis of their claims about the carbon sinks, viz (emphasis mine):

In 2023, the CO2 growth rate was 3.37 ± 0.11 ppm at Mauna Loa, 86% above the previous year, and hitting a record high since observations began in 1958, while global fossil fuel CO2 emissions only increased by 0.6 ± 0.5%. This implies an unprecedented weakening of land and ocean sinks, and raises the question of where and why this reduction happened.

…

The CO2 growth rate in the decade 2013-2022 has averaged at 2.42 ± 0.08 ppm yr -1 . In 2023, it increased to a record high value of 3.37 ± 0.11 ppm yr -1 at the Mauna Loa station (MLO).

Whoa, biggest jump in the record, which “implies an unprecedented weakening of land and ocean sinks“! EVERYONE PANIC!!

So as is my habit, I went to the Mauna Loa site and got the CO2 data. It’s monthly, so I converted it to annual data. Here is the result, showing the growth rate for a given year as the average value for the previous year subtracted from the average value for the given year.

Figure 1. Annual change in atmospheric CO2

Hmmm, sez I once again… the change from 2022 to 2023 looks perfectly average, and at 2.5 ppmv, it’s certainly not the 3.37 ppmv they claim. Why would that be? They can’t be calculating the difference in annual averages.

A closer look at what the NOAA folks are doing when they calculate what they call the “annual mean rate of growth” explains the mystery.

The annual mean rate of growth of CO₂ in a given year is the difference in concentration between the end of December and the start of January of that year. If used as an average for the globe, it would represent the sum of all CO₂ added to, and removed from, the atmosphere during the year by human activities and by natural processes.

There is a small amount of month-to-month variability in the CO₂ concentration that may be caused by anomalies of the winds or weather systems arriving at Mauna Loa. This variability would not be representative of the underlying trend for the northern hemisphere which Mauna Loa is intended to represent.

Therefore, we finalize our estimate for the annual mean growth rate of the previous year in March, by using the average of the most recent November-February months, corrected for the average seasonal cycle, as the trend value for January 1. Our estimate for the annual mean growth rate (based on the Mauna Loa data) is obtained by subtracting the same four-month average centered on the previous January 1.

Now, that all sounds perfectly legit, but doesn’t explain the large 2023 value. So I looked at the monthly data. Figure 2 below shows the more recent of the four-month November-February periods that they are averaging.

Figure 2. Year over year difference, monthly CO2 data. Blue bars highlight the periods from November to February used in their calculations

Following their instructions, I can replicate their calculations exactly. And in Figure 2, we can see why they’ve gotten such a high number—by chance, the data centered on Jan 1, 2024 occurred during a peak time of annual CO2 rise, and the data centered on Jan 1, 2023 occurred during a low time.

What difference does that make? Well, they’ve taken a four-month average starting in November, and compared that to the same period of the previous year … but they could just as easily have started their four-month average in January or June and done the same calculation of annual change.

But here’s the thing. If they’d chanced to start their four-month average in January rather than November, instead of the difference from 2022 to 2023 being the largest jump in the record, it would only be the 24th largest out of the 64 years. If they’d started in February, 11th largest. And if they’d started in June, it would be the 7th largest jump in the record.

And of course, this means that their claim that the carbon sink “is failing” is a totally falsified artifact of their calculation method. It depends entirely on the random choice of the month they’ve used to start their four-month average, and if they’d started in January, they’d have been forced to conclude that there’s nothing unusual going on with the carbon sink in the slightest.

Having been caught by confirmation bias more than once, I can only have compassion for the authors. They were looking for a climate disaster, they thought they’d found one, and unfortunately, because that confirmed their preconceptions, they didn’t look any deeper.

However … they’re still totally wrong. The data doesn’t show any change in the carbon sinks. It’s just an artifact of the odd way that they are calculating the “annual” change.

My very best to everyone, I’m going outside to see the moon.

w.

As Usual: When you comment, I ask that you quote the exact words you’re discussing … it avoids endless misunderstandings.