Thread by @AlexSteffen

You should immediately mistrust any linear projection of positive climate impact, such as "Alaskan agriculture will benefit from warming temperatures."

(mini-rant)

[image:flickr.com/photos/baggis/…]

First, tho models are improving, there's always the chance that specific, long-term local impacts may work out differently than expected, for complex reasons.

"Our model indicates temperatures will rise" is different that knowing, for sure, that temperatures will rise steadily.

Second, variability is at least as important as trend. If nine years out of ten there are warmer conditions, but the other year there are unprecedented and catastrophic floods, the net result is not a gain for farmers.

The fewer factors you model, the more confidence you can project; but the real world teems with complex interactions, threshold effects and incomplete data.

The real world is becoming more unpredictable faster than our models are improving.

Even apparently beneficial trends, like warming spring and fall temperatures, may actually be disastrous, if they, for instance, bring new pests—or familiar pests get a head start before their predators arrive.

The list goes on, but here's one more: the way people respond to expectations.

The belief that Alaskan farmland is improving may draw speculative investors looking to leverage future productivity declines in bread baskets elsewhere.

Their likely unrealistic expectations can distort local land prices, pressure local farmers and create economic instability.

If you're really looking to understand comparative advantage in the planetary crisis, what you want to seek is relative stability in the face of expected changes and a growing commitment to ruggedization.

Old: These places will "win" because of climate changes, etc.

New: Places that show fewer signs of getting thrashed by change are undervalued.