Here's something to keep you amused while I compose my summary of the AGU. It's from one of the presentations I attended.
The left hand side shows global mean surface temperature from an 800 year integration of a well-known climate model (click on the pic for a larger version). The puzzle is to identify what the colour changes represent, and what caused the climate change. The right hand pic (from the IPCC AR4) is simply to provide some context on the vertical scale, it has no direct relevance to the puzzle - note the horizontal scales aren't the same either.
Those who actually know the answer are invited to not answer too soon, so others can have the fun of trying to work it out.
The left hand side shows global mean surface temperature from an 800 year integration of a well-known climate model (click on the pic for a larger version). The puzzle is to identify what the colour changes represent, and what caused the climate change. The right hand pic (from the IPCC AR4) is simply to provide some context on the vertical scale, it has no direct relevance to the puzzle - note the horizontal scales aren't the same either.
Those who actually know the answer are invited to not answer too soon, so others can have the fun of trying to work it out.
9 comments:
some changes in parametrisation somewhere in the model? ( re-compiled during the simulation )
With a hat tip to Andy Dessler, this is clearly a simulation of the effect of hoards of flying monkeys with mirrors.
My guess is that it is an attempt to simulate the younger dryas, and the temperature change refers to somewhere not too distal to the North Atlantic.
Hey, don't give up so easily! So far, ICE is closest (but not close enough to be called correct).
The closed down the run and went out for beers, recording the last step. Then they restarted it, but got smashed by limited precision?
A change in software version (compiler, routines..) along a control simulation
in fact i was thinking about something dealing with convection or cloudiness parametrisation (i feel i'm moving away here...)
Well, I think we can call viento the winner - congratulations. The story is only at best 2nd hand to me, but AIUI the colour changes indicate when the integration switched hardware (and thus presumably the model was recompiled). There were no deliberate changes in model parameters or forcing though.
This reminds me of the old joke:
Patient: Doctor! It hurts when I do this!
Doctor: Don't do that.
Post a Comment