Kate Marvel (NASA GISS): "Towards Robust Community Assessments of ECS"
Harry Mutton (Met Office): "Biases in Climate Model Global Warming Trends Related to Deficiencies in Southern Ocean Sea Ice Evolution Over Recent Decades"
00:28:14 Marcus Sarofim - EPA - FCLA Econ: What do the last couple of years Antarctic sea ice decrease mean for this research?
00:28:43 Marcus Sarofim - EPA - FCLA Econ: (oh, this is addressed by the last bullet)
00:29:30 Karsten Haustein (he/him): linked to the 1st question: do we understand what has caused the flip in recent year?
00:29:36 Tim Merlis: Was the change in lambda_net looks larger than the change surface albedo feedback? If so, what other feedbacks are changing?
00:31:29 Karsten Haustein (he/him): thanks Harry! Fair enough ;)
00:44:05 Kevin Bowman: Can we get Kate’s paper citation?
00:44:52 Thorsten Mauritsen: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-391/
00:46:27 Cristian Proistosescu: I think I know who Expert C is ;)
00:46:36 Ceppi, Paulo A: Reacted to "I think I know who E..." with 😂
00:46:41 Karsten Haustein (he/him): Reacted to "I think I know who E..." with 😂
00:47:03 Tim Osborn: Reacted to "https://egusphere.co..." with 👍
00:47:46 David Sexton (Met Office): One thing to watch for is when the evidence lies in the tails of the prior predictive distribution (use the prior to predict the distribution of the observable). If the evidence is in the tail, then often the Bayesian analysis can be overly sensitive.
00:48:53 Thorsten Mauritsen: Replying to "I think I know who E..."
I have no further comments 😅
00:50:47 Marcus Sarofim - EPA - FCLA Econ: Replying to "I think I know who E..."
https://davidkeith.earth/wp-content/uploads/2023/10/13.morgan.1995.subjectivejudgmentsbyclimate_experts.s.pdf - Figure 1, Expert #5.
00:50:48 David Sexton (Met Office): This can be diagnosed so it might be used to tell the expert their prior is not consistent with the evidence.
00:51:28 Cristian Proistosescu: Isn’t the choice of priors sometimes a statement of belief on the interpretative model? If we use a flat prior on lambda (instead of S), isn’t that the same as saying that we think the system operates through feedbacks, rather than through a bulk S?
Do you have any thoughts on that?
00:53:41 David Sexton (Met Office): Tricky to talk at moment
00:53:55 Tim Merlis: Replying to "I think I know who E..."
thank you for sharing that pdf!
00:54:49 Andrew Williams: Reacted to "https://davidkeith.e..." with 😳
00:55:42 Mark Webb: Agreed!
00:57:33 David Sexton (Met Office): See section 5.1 of Sexton et al 2012 Multivariate probabilistic projections using imperfect climate models part I: outline of methodology | Climate Dynamics