Test 2

This is a story without an image blah blah. 

I think this is a fair critique. Many current-generation climate models do indeed show a remarkably stable AMOC, even under high-emission scenarios. To my knowledge, there are several reasons why models might underestimate the risk. First, models have biases in the hydrological cycle. They tend to underestimate the freshwater input from Greenland ice sheet melt and from increased precipitation in the North Atlantic. Freshwater is precisely what destabilizes the AMOC by reducing surface density. Second, very related, many models do not fully couple the ocean to dynamic ice sheets. If Greenland melts faster than models currently assume, the freshwater input could be much larger than simulated. Third, most climate models do not have high enough resolution to properly resolve turbulent processes, like the mesoscale eddies, mixing that play a key role in AMOC dynamics.