Over geologic time the atmospheric pressure has been much higher than today. Likely 2-3x as much. With CO2 being 10X or likely 100X todays. There is a shit ton of carbon and Oxygen locked up in the rocks. It boggles the mind.
Currently, solar energy fluctuates up and down 6% over each orbit - with northern winter getting more and northern summer getting the least. Just by orbital mechanics. Orbital variation and the wobble in Earth's rotation on its axis and the precession of same is poorly understood and there is little in the way of proxy data to model it either. The big planets and Venus have a lot to say about it and since the NICE work came out, there is a lot of uncertainty now.
There is a lot of evidence of solar and interstellar dusty transiting the Earth's orbit. Some days see up to 30% less visible/thermal radiation due to it. Again, we just don't know.
The fact that the ice age glaciers melted in just a few thousand years each time and that heat fluctuations due solely to Milankovitch cycles only add up to 20% of the required heat flux means we just don't know.
I spent the summer up at Glacier NP and hiked up to several recent terminal moraines from the little ice age. They are huge relatively speaking. Many have buried ice from then. The fact that these glaciers were 500-1000 feet thick and began to push down to the hanging valleys is fucking scary. No one has an explanation of why they began to grow or why it got "warmer." The younger dryas saw temps drop 10 degrees F in a few years with a return to full ice age conditions in two decades. MUCH colder than the little ice age. No one knows why. again, fucking scary. And consider the Alps. or the Arctic, both were mostly ice free 6000 years ago. Now, tons of ice. Much of it came back fast.
The last few months we've looked at solar irradiance at the surface for a very large resilience project and it looks like the NREL data is junk for large parts of the year across many locations in the US. Abut 30% of the days look bad. It was based on satellite measurements. I'm afraid to go look at sat data for other things - it can't be much better as the methods are the same.
