[redthrowaway]

Simplest, but not optimal. Ideally you'd want to settle on a mean that:

a) Maximizes arable land, b) Minimizes climate-based natural disasters such as hurricanes, and c) Is as near to self-sustaining as possible. That is, the global mean should not lead to significant growing or shrinking of ice caps and glaciers, nor significantly alter carbon uptake by the oceans. While some degree of human control could make up for heterostatic forces, a swing too far in either direction would quickly exceed our ability to compensate.

[curious_man]

> While some degree of human control could make up for heterostatic forces, a swing too far in either direction would quickly exceed our ability to compensate.

Exactly. That's why I'm suggesting that the "not worse" solution could be to artificially heat up the earth to the actual temperature for every latitude. Politics aside, a different temperature would probably disturb the current equilibrium and potentially cause some side-effects worse than the suspected mini ice-age.

[wunderfool]

do people really believe that rapid climate change will reveal some new mass of desirable farm land? maybe in 5000 years as the soils adapt to new climates, but not fast enough to feed a hungry planet.

there are some lands on the periphery of existing arable zones that might become desirable in the short term, but certainly there will be a net loss worldwide...local economies have grown around existing arable zones, you have to transplant all that too

[redthrowaway]

That's what I was getting at. Too big a swing could, for example, hasten the expansion of the Gobi Desert, destroying millions of acres of arable land and endangering the food supply for a billion people.