In February 2023 a new study was published measuring what is colloquially called the “greenhouse effect” even though the actual mechanics involved are quite unlike covering a growing space with a thin exterior glass wall. The paper (journal article here, non paywalled version here, h/t NoTricksZone) used satellite data covering 1983 to 2020, during which we all know that greenhouse gases went way up in the atmosphere. And we all know (h/t Al Gore) that they “trap heat” and roast the planet. But how exactly do they “trap heat”? Well, they’re transparent to visible light coming in, hitting the Earth’s surface and warming it, but they then absorb long wave or infrared radiation coming back from the warmer ground and re-radiate it in every direction including, for some of it, back downwards, which further warms the surface. (We have a backgrounder in which a physicist explains the process in detail here.) So it stands to reason that as greenhouse gas levels go up, since they are supposedly the control knob on the climate, the amount of downward or “downwelling” longwave radiation should be going up too. Should. According to the theory. But according to the data, ain’t. It’s going down. And some of us still care more about what the facts are than what the models say they ought to be.

The data series is the red line in this graph (LWdn is terrestrial longwave downwelling radiation, SWdn is solar shortwave downwelling radiation):

To assemble it the authors made use of newly-available high precision measurements from a group of satellite monitoring systems, which they combined with existing models and data on cloud cover, atmospheric chemistry and other climate-related processes. And the end result shows that from 1983 to 2008 downward longwave radiation, instead of rising as “greenhouse gases” did, trended downwards. After that it flattened out a bit. And the puzzles don’t end there.

As the authors explain, their method takes account of rising CO2 and methane (CH4) levels which on their own should cause the line to slope up. But it goes down. And not only that, but the new data system also yields estimates of air temperature and “skin” temperature on the Earth’s surface, and it went down over this interval too, even though the usual thermometer data sets said it went up.

“The [satellite-based] FH calculations (and previous versions) account for increasing CO2 and CH4 abundances, which should produce an increase in LWdn, all other things being equal; but as Fig 3 (lower panel) shows, the near-surface air temperature (Ta) and skin temperatures (Ts) from ISCCP-H used in FH are generally decreasing.”

Now here are the temperature readings:

So why is the “greenhouse effect” getting weaker even if greenhouse gases are going up? It’s because greenhouse gases are only part of the story. Changes in water vapour and cloud cover also count in the total greenhouse effect and they have gone down. So this graph shows the changes in cloud amounts globally over the period:

Cloud cover has gone down, which on its own has a major (positive) effect on warming because the net effect of clouds is to reflect solar radiation back into space before it can hit the ground and warm it. But CO2 didn’t make it go down. In the real world, as opposed to the artificial one inside simple climate models, greenhouse gases aren’t even the control knob on the total greenhouse effect, let alone the climate.

Check out the entry at NoTricksZone for further discussion including a list of earlier papers showing the same thing.