<a href="http://www.physics.mcgill.ca/~gang/eprints/eprintLovejoy/esubmissions/Climate.sensitivity.20.2.12.pdf" rel="nofollow">Another recent paper on sensitivity</a>,
with one calculation having it as an increasing scaling function of time scale, with feedback factor varying between 0.3 and 20. Temperature series also show 1/f scaling behaviour, not generally reproduced by GCMs. As far as I know, there are no satisfying explanations, either for the scaling in nature or the lack of it in GCMs. Anyone know different? Also well-known (?) that such scaling can be modelled by a handful of AR1 processes, with power related hierarchically. Anyone have any ideas about the physics behind such a model?
↧