Control engineers would call this a runaway controller.

When the response to a change is to increase the change, rather than decrease it (which is what, for example, thermostats normally do), change happens very quickly, and uncontrollably. One of the issues with global warming is whether or not the feedback is positive, or negative. That is, does the warming result in even more warming, resulting in...or do things happen at higher temperatures that result in cooling?

One potential positive feedback might be that if glaciers and ice caps melt, the albedo of the planet decreases, which means that less energy is reflected back into space, which could result in further warming. In the other direction, if we are headed into a new glacial period despite the greenhouse effects (perhaps because solar activity dominates all else), then increasing snow cover makes things colder because more solar energy is rereflected, thus causing more cooling, and an accelerating glacial advance.

On the negative feedback side, though, it could be that more warming results in more clouds, which might in turn have the effect of cooling things off.

I suspect that the reality will be a combination of positive and negative feedback mechanisms, and it's hard to know what the overall effect will be, though ultimately, it will be negative, but perhaps at a significantly higher (or lower) temperature. I'd be very surprised if the seas end up either boiling, or freezing solid.

The data from the period around the appearance of plants in the fossil records tells us a couple of things. 1) An atmosphere containing no free oxygen, lots of methane, water, nitrogen, and carbon dioxide is _STILL_ not a place where the seas are boiling. 2) You aren't going to make an environment that is more strongly influenced by global warming that _that_ even if you burned every ounce of coal and all the hydrocarbons materials.

This makes it closed-loop-stable.

Just knowing that it _is_ closed loop stable tells us how hysterical we need to be.

I believe solar output has increased substantially since then. We may no longer be closed loop stable.

Ok, it appears that the Sun was several billion years ago about 75% as bright as it is today. That's far too large a difference to assume closed loop stability IMHO. But it's obvious that the Earth wouldn't achieve Venus level global warming without a factor of four increase in solar output over current levels IMHO.

This is known as "the paradox of the faint early sun." That is, the sun has brightened substantially over geological time, but the earth's temperature has stayed within relatively narrow limits.

Over the same time, the CO2 content of the atmosphere has declined. If you assume that the decline of CO2 has balanced the brightening of the sun, you can calculate the dependence of global temperature on CO2. It's about 0.4C/doubling of CO2, or an order of magnitude less than current global warming models. BZZZZZZZZT un-PC result YOU MUST FORGET WHAT YOU JUST READ

(See e.g. http://oldfraser.lexi.net/publications/books/g_warming/real_world.html
scroll down to "The greenhouse effect from an increased concentration of atmospheric CO2 over geologic time")

Thank you Karl Hallowell, I'd forgotten that aspect. Irritating thing, facts. ;)

From Bob Hawkin's link:
"which is believed to have contained much more CO2 than it does today (Hart 1978; Holland 1984; Wigley and Brimblecombe 1981; Walker 1985)."

Does anyone know what percentage 'much more' is? Venus is 96%...