The linear predictions and feedback loops aren't speculation. It is known, at least to an order of magnitude, the amount of sequestered carbon in boglands and c

The linear predictions and feedback loops aren't speculation. It is known, at least to an order of magnitude, the amount of sequestered carbon in boglands and clathrate deposits. Same thing with trees in places vulnerable to wildfire. We know the amount of ice in places vulnerable to climate change, and we have a solid understanding of the albedo effects of Arctic sea ice. The correlation between CO2 and atmospheric temperatures isn't just empericial, but can be well modeled with today's forecasting supercomputers. The predictions that were made 40 years ago are in line with what we are observing today, and the ensemble modeling system we use to look at different climate scenarios shows the ample world of possibilities. None of those things indicate that there's any reason why there should be a _better_ case than the already conservative predictions, and all point towards those feedback loops being substantially more dramatic. For the record, the "extrapolation" isn't unbounded; there are limits to atmospheric temperature determined by atmospheric radiation, ocean albedo, the effects of temperature on cloud formations, etc. It's just that the limit is high enough that there will be literally no ecosystem able to adapt and our capacity to grow food is out of the question at the point.