It doesn't ask the question "why did cloud cover go down?" Perhaps they are both anthropogenic.
It does more than that. It predicts that for every 1% decrease in low clouds, the temperature goes up 0.11 C. They also used all available data (25 years) of cloud cover.
The correlation is very high. Figure 3 shows this.
> It doesn't ask the question "why did cloud cover go down?" Perhaps they are both anthropogenic.
Perhaps, but it's entirely unproven that co2 affects low cloud generation.
I get that dumping a large amount of CO2 in the atmosphere and hoping nothing will change seems scary. But so is converting a large portion of land into farmland and cities, drilling into tectonic plates for geothermic energy and oil (which seems to correlate with earthquakes forming nearby), etc. And keep in mind it's a trace gas. We're talking about parts-per-million here.
With 8 billion people on the planet, we are going to be making changes to the environment regardless. It's important to understand what we are doing, and not blindly try to stop co2 production, when something else could be much more problematic.
The problem with this line of thinking is that, when you do something to a system, usually lots of properties of the system change. So you can always say "Look, properties A and B changed at the same time right after we did X! So maybe A caused B, not X."
What sort of evidence are we talking about? We don't have a another exact copy of Earth so we can't perform a controlled experiment by somehow inducing low clouds to form in one and not in the other. But if you need this sort of proof then how can you turn around and justify the belief that CO2 is definitely causing climate change?
Even NASA would disagree with you (https://visibleearth.nasa.gov/view.php?id=54219):
Clouds play a complex role in the Earth's radiation budget. Low Clouds reflect much of the sunlight that falls on them, but have little Effect on the emitted energy. Thus, low clouds act to cool the Current climate.
(Note the paper is specifically referring to low clouds)
> The problem with this line of thinking is that, when you do something to a system, usually lots of properties of the system change. So you can always say "Look, properties A and B changed at the same time right after we did X! So maybe A caused B, not X."
Sure, but also, when you make changes X1, X2, X3, ... Xn, you can't just arbitrarily say that X1 must be the cause and X2 .. Xn are not.