09:46:54	 From  Agathe Toumoulin : Hello !
09:53:23	 From  madeleinevickers : worked for me
09:53:29	 From  Kasia_K_Sliwinska : yes:-)
09:53:31	 From  Yurui Zhang : yes
09:53:51	 From  Xiangyu Li : yes
09:58:09	 From  Dan Lunt : Hi everyone!  We will start in a couple of minutes!
10:14:22	 From  Kerim Nisancioglu : Great to see you all - the PMIP community feels so much closer. Thanks for hosting Dan!
10:16:22	 From  Dan Lunt : Do use the chat if you'd like to ask any questions or post comments...
10:23:05	 From  Dan Lunt : Could we use a similar approach for CO2 estimates?
10:33:48	 From  Petra Langebroek : Maybe related to Dan’s question; what CO2 values did you use for your ECS calculations?
10:36:33	 From  Appy Sluijs : How different are the answers relative to Margot Cramwinckels paper using an approach just trying to optimally validate a fully couple climate simulation?
10:37:26	 From  Anna Nele Meckler : I might have misunderstood but you said that jackknifing out methods did not change the means, but the dataset without TEX was 2.5 °C colder in GMST... can you explain?
10:38:17	 From  Kate Littler : Q. How much influence does our uneven spatial coverage of proxy data have for the GMST estimates (and hence ECS)? e.g. if the SW Pacific data is truly  “weird” can that skew the global estimates, and can we correct for it?
10:39:10	 From  Anna Nele Meckler : ok, thanks
10:40:17	 From  Kate Littler : ok thanks (I love the SW Pacific, don’t shoot me!)
10:40:24	 From  Appy Sluijs : Thanks Gordon!
10:41:07	 From  Chris Hollis : SW Pacific thinks that could be worth trying :-)
10:41:36	 From  Gordon Inglis : For more information on the GMST methods, see here: https://cp.copernicus.org/preprints/cp-2019-167/
10:42:08	 From  David Evans : Dcomb has no SW Pacific!..
10:42:23	 From  Laia Alegret : Sure Kate! Just about to submit a paper incl. benthic d18O across the EECO (mainly ETM3) from SW Pacific Site U1510…
10:46:27	 From  gavin : only CESM is obviously non-linear on this plot, no?
10:47:13	 From  gavin : And I assume, CESM=CESM1 (CAM5)?
10:47:31	 From  David Hutchinson : @gavin I’d say most of the models tend to have somewhat increasing climate sensitivity as they get warmer
10:47:45	 From  David Hutchinson : Just that CESM sticks out a lot at 9x
10:48:23	 From  Manoj Joshi : One of the models seemed very nonlinear in its polar amplification vs global warming- was that GFDL?
10:48:33	 From  David Hutchinson : yes
10:53:17	 From  Manoj Joshi : I guess some emissivity change comes from changes in water vapour transport, so it’s sort of a transport term?
11:01:35	 From  Kate Littler : -Sounds great, Laia! :)
11:01:52	 From  gavin : how confident are we that models at 6x or 9xCO2 are behaving well? (Where is the tropopause? changes in WV impacting surface pressure?)
11:01:54	 From  Yurui Zhang : What would be reasons of that GMST and polar amplifications are contributed from quite different components
11:01:59	 From  Jean-Baptiste Ladant : Could the good performance of NorESM in your last slide be related to the different land-sea mask that you mention they used? 
11:02:00	 From  Appy Sluijs : It seems that all models are sort of consistent with the current state of data. Would it be fair to say we just need much more data and better uncertainty?
11:02:05	 From  Manoj Joshi : Why does GFDL do that?
11:02:38	 From  Seb Steinig : I will talk a bit about the nonlinearity in GFDL in the afternoon. SPOILER: seems like a nonlinear SW cloud feedback influencing the polar amplification
11:03:00	 From  Manoj Joshi : Aha ok- thanks
11:05:35	 From  Alan Kennedy-Asser : I’m surprised how much impact the removal of the AIS has on GMST, given the amount we saw it reduced GMST at the EOT. Any thoughts Dan?
11:11:16	 From  Dan Lunt : Alan - the partitioning is not perfect - for example the 1x Eocene minus 1x pi is not only ice sheet removal, but also vegetation; and the emissivity/albedo terms also include seaice and snow and cloud feedbacks. 
11:16:15	 From  Pierre Sepulchre : Seb, does this mean  we can open the netcdf files remotely in scripts (i.e. without downloading them locally?)
11:19:48	 From  joepvandijk : Thanks Seb, this was super useful for a proxy person like me :)
11:19:52	 From  Laia Alegret : Thank you for such a clear set of instructions!
11:21:02	 From  Chris Hollis : Terrific. Shouldn't we include the proxy data here too?
11:21:03	 From  Pierre Sepulchre : OK, thanks a lot for this very clear presentation !
11:21:06	 From  Petra Langebroek : How do we deal with acknowledging the original modelling groups/ scientists who performed the simulations?
11:23:05	 From  Petra Langebroek : Sounds like a good procedure, Dan!
11:33:26	 From  Dan Lunt : Jean-Baptiste: Partly, yes.  However, it still has the best RMS when calculated on the other reference frame, so my feeling is that the reference frame is not crucial. 
11:34:59	 From  Dan Lunt : Gavin - My understadning is that NCAR have done a lot of work to ensure that their radiation scheme is robust up to very high CO2...not sure for other models.  I know that the radiation scheme of HadCM3 is very iffy above 4x CO2.  On another note, many of the models crashed when run at 4x or above....
11:53:17	 From  Kate Littler : Q. I thought all the models were using the same land-sea mask/ bathymetry… but it looks like the straights/gateways (e.g., Tasman) have different geometries from model to model… have I misunderstood?
11:57:05	 From  Tina vdF : Question: what proxy data do you plan to use for your next stage?
11:57:27	 From  Ulrike Baranowski : Thank you that was super interessting! - You see a lot of salinity variation within the Atlantic compared to Pacific and Indian. I might have missed this but why do you think this higher latitudinal salinity variety is caused in the Atlantic?
11:58:16	 From  Kate Littler : Ah ok, thank you.
11:59:17	 From  Tina vdF : Great talk Yurui.
12:00:18	 From  Isabel Sauermilch : Is there an attempt to run all DeepMIP models in the same resolution in the future for better comparison? (related to Kate’s question)
12:00:22	 From  David Hutchinson : Just to follow up on the gateway issues between models: In the ocean models, there needs to be at least 2 grid cells width for a strait to be dynamically open. (If there’s only 1 grid cell wide, there is no velocity allowed.) This means in practice, every group has to make manual adjustments to narrow straits, which means the final land-sea mask is always different.
12:00:23	 From  Ulrike Baranowski : Thank you very much
12:02:08	 From  Kate Littler : Thank you Yurui and David for clarifying RE Straits.
12:02:33	 From  Dan Lunt : Isabel - a good idea!  But it would be very challenging because the resolution of many models is hard-wired into the model, and/or only one resolution has been properly tuned.
12:03:46	 From  Manoj Joshi : Could parameterise throughflow in small channels like the straits of Gibraltar (but we have observations for this though…)
12:07:09	 From  valeria luciani : Sorry but I must leave the meeting. All the best to everybody, Valeria
12:07:40	 From  Isabel Sauermilch : Thank you, Dan! It’s too bad, especially in the ocean resolution seems to matter quite a bit. Perhaps it might be worth discussing to manually adjust (as David mentioned) the straight’s geometry to the same widths and depths?
12:08:36	 From  Dan Lunt : Isabel - Yes, good idea!
12:19:41	 From  hbrinkhuis : EECO Arctic kicks of the Azolla phase in the Arctic, ~100% fresh water surface, then later mid Eocene alleged sea ice diatoms.. And cf Appy
12:21:28	 From  Pierre Sepulchre : Yannick looked really young on this picture xD
12:21:30	 From  Margaret Collinson : Agreed from palaeobotanical perspective no evidence for sea ice on Arctic in Early Eocene
12:37:32	 From  Tom Dunkley Jones : Just wondering about the configuration of the Tasman Gateway in the models and what “age” this might represent? i.e. you’ve got a sensitivity study on the DC but the Tasman Gateway is presumably also in play ...
12:39:00	 From  Michael Henehan : Hello! Nice talk.. Linked to Tom’s question, you mentioned the fact that you don’t get North Atlantic Deepwater because your Arctic gateways are open.. How sure are we that the gateways are that open at this time? Is this a settled question? Or could the Fram Strait and Barents Sea have been more restricted than your continental configuration prescribes?
12:39:55	 From  Manoj Joshi : Decadal/centennial variability of the ACC seems much larger in D2500 than D1000- what’s the reason for that? I missed that.
12:40:41	 From  Tina vdF : Really nice talk. Question: Are the results sensitive to the pCO2 chosen at 1120 ppm?
12:41:37	 From  Michael Henehan : Thanks!
12:41:55	 From  Carlos Jaramillo : Q. If Antartica is covered with forest, will that make a difference in the land temperature?
12:42:16	 From  Tom Dunkley Jones : Hi Michael - think you’re right; interesting potential coupled opening of both the DC and the GSR gateway - in the early middle Eocene it might make it hard to diagnose what’s driving changes in circulation / temp / salinity in the N. Atlantic? ie. wether DC or Arctic gateways…
12:42:53	 From  hbrinkhuis : NB: quite interesting that apparently, some people think the timing of opening and subsequent history of the Tasman Gateway is not well constrained…
12:44:17	 From  Tina vdF : Thanks.
12:45:26	 From  Yannick Donnadieu : yeah, thanks to all speakers !
12:48:44	 From  Dan Lunt : Thanks to all speakers...we will reconvene at 1:20pm UK time, i.e. in ~30 minutes from now.
12:56:45	 From  Agathe Toumoulin : @hbrinkhuis, sorry, my answer was a bit messy. Owing our deep TG, simulations with a shallow DP are probably a bit idealized. But they help understanding the contribution of the DP depth to these the different points. Since some studies describe the earliest opening of the TG around 49Ma and a deepening around 35 Ma, we may say that our results reflect at least the latest Eocene.
13:34:08	 From  Dan Lunt : Please use the chat if you have questions!
13:35:54	 From  Chris Brierley : Given the reduced land extent at the Eocene, is it worth comparing with future monsoon changes over a present-day ocean region?
13:36:11	 From  Daniel Breecker : Why are the models drier during Eocene at 1x CO2
13:36:12	 From  Pierre Sepulchre : Thanks Charlie for this nice presentation. Do you know what methods is used to retrieve precipitation from proxy data ?
13:36:20	 From  Gordon Inglis : Nice talk Charlie! For the multi-model mean, there is a data-model mismatch with MAP - is this resolved when you use JJA or DJF?
13:36:22	 From  Rich Pancost : Very nice!
13:36:39	 From  Alex Farnsworth : To what extent do you think we can believe the paleo-topography in the region? Especially considering this can modify monsoon strength and put models in or out of the proxy envelope.
13:40:44	 From  Pierre Sepulchre : @Alex unfortunately I think the topography over eastern and central Africa is indeed too high in these boundary conditions.
13:41:23	 From  Daniel Breecker : Thanks Charlie. Nice talk!
13:43:38	 From  Alex Farnsworth : Great! Thanks @Charlie and @Pierre
13:44:10	 From  Pierre Sepulchre : Thanks Charlie!
13:58:47	 From  Michiel Baatsen : I’d be very interested to see how these changes occur in Summer vs. Winter. You mentioned this in the outlook, but do you have some first results on this already?
13:59:06	 From  Bridget Wade : Nice talk, can you comment on how your model results compare with Paleogene dust records (I realize that the dust records are very limited).
14:02:53	 From  Daniel Breecker : Thanks Xiangyu!
14:07:49	 From  Dan Lunt : Bridget - dust records can be challenging to interpret as wind strength/direction - soil moisture and vegetation and precip can also play a key role...cf. for the LGM. 
14:09:03	 From  Charlie Williams : It would be interesting to compare side-by-side your comparisons to CMAP with my comparisons to TAMSAT, to see if the biases relative to 2 sets of observations are the same.
14:13:33	 From  Michiel Baatsen : You could try to compare the precipitation climatologies with Era5 reanalysis as well, to check if this is indeed an observational bias. The older (Era40) reanalysis also shows 2-4mm/day over the Southern Ocean, which seems to agree better with what you find for the PI controls.
14:16:59	 From  Charlie Williams : I realise I didn't answer the question by @DanielBreecker very well, "Why are the models drier during Eocene at 1x CO2", but something @NatalieJBurls just said gave me an idea.  Obviously the 1x is the same as the PI, so the only changes here are related to the land-sea mask, topography, veg etc.  So might it be that these changes are the dominant driver (causing a weaker West African monsoon relative to PI) under the same CO2, but that when we increase the CO2 that becomes the dominant driver and enhances the monsoon?
14:17:43	 From  David Hutchinson : It seems like the amplification of (P-E) is pretty clear over the oceans, but not over the land. It might be interesting to take a zonal mean of each domain separately?
14:18:09	 From  Daniel Breecker : Thanks Charlie, yes I was wondering if there was one particularly important boundary condition
14:19:18	 From  Yurui Zhang : @ David, it is seems that Runoff makes differences between atm and ocean
14:21:11	 From  Michiel Baatsen : Question on the moisture transport: this is mean(v)*mean(q) right? In that case it is not the actual mean transport as the eddies also contribute to the mean through v’*q’
14:22:12	 From  Jiang Zhu : Nice talk! Any comments on the single ITCZ in CESM9x? Double ITCZ is gone.
14:24:07	 From  Michiel Baatsen : Most models actually have the VT and VQ means, which include both transports
14:24:48	 From  Michiel Baatsen : Then indeed you can get V’Q’ from mean(VQ)-mean(V)*mean(Q)
14:25:09	 From  Michiel Baatsen : Actually: mean(V’Q’) in that case
14:25:32	 From  James Zachos : Good morning all
14:26:10	 From  joepvandijk : :D
14:26:17	 From  Michiel Baatsen : I know all about Dutch accents ;)
14:26:34	 From  Appy Sluijs : Sorry Gordon, for the interruption!
14:26:41	 From  Henk Brinkhuis : I know nothing about Dutch accents
14:39:11	 From  Joost Frieling : Q. re the compilation; any progress / ideas how to increase coverage in dry regions? Seems that the ‘average’ site in the compilation (probably in part related to preservation?) is wet with >1000 mm/yr
14:40:03	 From  Gavin Foster : when considering data model comparison for hydrography it might be worth screening out/splitting the data into areas with active topography through the Cenozoic vs. more stable regions
14:42:02	 From  Daniel Breecker : @Joost - I’ve been thinking of looking at dD of hydrated volcanic glasses. Biggest problem is preservation of Eocene glass.
14:42:19	 From  Pierre Sepulchre : Thanks Gordon !
14:43:20	 From  Appy Sluijs : I guess ultimately, one of the most relevant aspect is seasonality of precipitation rather than MAP.
14:43:25	 From  joepvandijk : @Gordon, siderite forms in wet regions too :)
14:43:27	 From  Appy Sluijs : How shall we proceed on that?
14:44:48	 From  joepvandijk : If we decipher the evaporative enrichment in calcite d18O, possibly with D17O analysis, we may be able to provide hydrological reconstructions in drier regions.
14:45:26	 From  David Hutchinson : I can test it
14:45:55	 From  Agatha De Boer : Need to go now. Thanks for all the nice talks!
14:52:13	 From  James Zachos : Rather than just comparing mean annual, am more curious about seasonal variations (differences) in P-E. The marine observations are likely seasonally biased. And marine biota don’t care as much if its wetter or dryer.
14:53:09	 From  Dan Lunt : We are on a Break...returning in 30 minutes form now, i.e. 3:20pm UK time.
15:15:16	 From  Dan Lunt : We are on a Break...returning in 5 minutes form now, i.e. 3:20pm UK time.
15:20:49	 From  Ulrike Baranowski : works
15:20:51	 From  Gregory Tourte : all good!
15:22:09	 From  Dan Lunt : After 1 minute I will skip to the Results so that we have a 20-minute talk.
15:27:19	 From  Dan Lunt : If you have questions about Polina's talk, which is currently playing from a recording while she is on fieldwork, you can email her here: polynom7@gmail.com
15:42:17	 From  Dan Lunt : If you have questions about Polina's talk, you can email her here: polynom7@gmail.com
15:55:46	 From  Frederic Fluteau : Need to go now. Thank you for these nice talks
16:01:21	 From  Pierre Sepulchre : I am wondering if, for present-day, atm moisture can be back-tracked over Antarctica to know where is the main evaporative source. What about diagnosing this in deepMIP simulations ?
16:06:32	 From  Tina vdF : Brilliant last sentence!
16:06:42	 From  Pierre Sepulchre : Thanks !
16:30:13	 From  igor : Maybe I missed this but why you didn't include COSMOS in the analyses?
16:30:38	 From  joepvandijk : Hey Seb, great work! I probably missed it, why would MHT decrease from 4x to 6x PI CO2?
16:30:56	 From  Jiang Zhu : We haven’t figured out the clear-sky radiation fluxes.
16:33:41	 From  joepvandijk : Do you know why latent heat goes down?
16:34:05	 From  Matthew Huber : Interesting study.  I’m skeptical of APRP.  Caballero has shown that 50% of CRF is due to the very high temporal resolution transient cloud fraction, not the mean cloud amounts.  I understand that you are empirically accounting for that effect, but that effect should be a function of climate state (baroclinicity) and model.
16:35:56	 From  Jiang Zhu : Hi Matt, I compared APRP with full PRP in CESM1.2 and the difference is in general less than 15%.
16:36:42	 From  Jiang Zhu : I guess the difference might be greater over mid-latitudes?
16:40:30	 From  Matthew Huber : Thanks for reminding me of that.  I did like your analysis of that.  For a given model and a given gross climate state I think this is a constrainable number, and as you say in those simulations the impact is in the midlatitudes (although it could also be relevant when there is thin sea ice at high latitudes). But, it would be useful to see what happens for example with specified (lower gradient) SSTs to see if that zone shifts or has different properties.  It is also very likely to be model dependent given different cloud overlap assumptions and cloud physics. So I guess for me to have confidence in a broad application of the method I’d be happier to see a model-by-model and climate state by climate state validation study.
16:41:52	 From  Jiang Zhu : Sure. I am happy to compile some results for you!
16:41:54	 From  Matthew Huber : This could be done in a 10 year fixed SST run, say 3 SST distributions, for all models, high temporal resolution saves for all cloud and radiative parameters.
16:43:24	 From  Dan Lunt : We have some simulations just like this that were carried out by Greg Tourte with HadAM3 and a high-resolution version of HadAM3.  he will be in touch!!
16:44:55	 From  Matthew Huber : cool.
16:49:11	 From  Paul Valdes : I've got to go now. Great set of talks. Thanks to everyone, especially Dan for organising it.
16:57:19	 From  Michael Henehan : I also have to go - thanks everyone!
17:04:10	 From  Natalie J Burls : Nice talk!
17:04:33	 From  Jiang Zhu : Nice talk! Height of Andes may also matter.
17:05:42	 From  Matthew Huber : What we did in Herold et al was a compromise between just taking initial CESM integrations and driving BIOME4 and using parameter choices that matched the existing paleobotanical data.
17:05:58	 From  Matthew Huber : Sensitivity studies would help.
17:06:44	 From  Appy Sluijs : Which game is your son playing Dan?
17:07:00	 From  joepvandijk : :’D
17:07:12	 From  Appy Sluijs : Looks pretty cool. Lets’ do that next session
17:07:30	 From  David Hutchinson : Have to go I’m afraid. Thanks so much for an excellent day of talks!
17:07:49	 From  Michiel Baatsen : I’ll try and bike very fast ;) Was hoping to have 30mins for that
17:07:56	 From  Dan Lunt : He is a Fortnite addict!!
17:08:10	 From  Appy Sluijs : I also have to leave I am afraid. Thanks al lot to all speakers and also to Dan et al for organizing and convening
17:08:29	 From  Appy Sluijs : Really enjoyed to be in touch again
17:09:02	 From  Michiel Baatsen : Thanks also for all the talks and Dan of course for organising, will try to reconnect but cannot stay after 7 (6 UK time)
17:09:32	 From  Joost Frieling : Thanks everyone - great session - very cool science :)
17:11:59	 From  Dan Lunt : We are just on a Break...back in 5 minutes from now for the Discussion.
17:13:27	 From  Pierre Sepulchre : I have to leave, sorry. Thanks a lot Dan for organising this. And thanks everyone for the discussions !
17:15:02	 From  Chris Hollis : Just tuned in again to catch the discussion. 4.15 am here!
17:19:51	 From  Xiangyu Li : Mid-latitude Asian climate pattern from DeepMIP simulations
17:20:05	 From  Rich Pancost : Woop!
17:21:12	 From  Matthew Huber : We are also happy to take papers or a special collection in AGU journals including Paleo.
17:24:25	 From  Yurui Zhang : for sure. I will do that
17:24:26	 From  Rich Pancost : If we're confident that we've spun the models up enough to do deep ocean... can we explore CCD issues through the Eocene??  There are rich datasets and could explore the interplay of CO2 change, weathering and circulation in a more sophisticated way than has been done in the past.  
17:25:51	 From  Charlie Williams : I don't think we EVER will for HadGEM3, Rich!
17:26:46	 From  Rich Pancost : Yes please on orbital impacts on hydrology, please!! 
17:26:52	 From  Rich Pancost : Sorry Charlie....
17:27:10	 From  Kate Littler : Agree with Rich, if this is possible with the models we have. We have so much data that is related to the carbon cycle (d13C, Ba/Ca, %CaCO3) and it would be really interesting to compare/contrast with the model outputs.
17:27:26	 From  zhongshi : I will carry out some the sensitivity experiments for topography or paleogeography. 
17:28:26	 From  Matthew Huber : To do the ocean carbonate right one needs full sediment models and the tend to be something coupled GCMs, including ones with biogeochemical cycles, do not have.
17:28:56	 From  Gavin Foster : but the circulation from a GCM could be s
17:29:06	 From  Gavin Foster : applied to GENIE, no?
17:29:08	 From  Yannick Donnadieu : We have a biogeochemistry model (PISCES) embedded in the IPSL model and we are still hoping to reach this goal, i.e. moving to the marine carbon cycle to study changes in PP, alkalinity and more generally the nutrient cycle
17:29:47	 From  Matthew Huber : yes, one can force GENIE or an emic with output from the various GCM experiments, which may be the most efficient approach.
17:30:56	 From  Yannick Donnadieu : both approach will be complementary but I would argue that ocean dynamics may be more trustable in NEMO :)
17:31:03	 From  Matthew Huber : yes, yannick I think your group is in a better position than most to do that.  Still the spins would be very, very long and the results model dependent.
17:32:47	 From  Yannick Donnadieu : yes, that’s right. Let’s see where we will be standing few months/years from now
17:32:48	 From  Rich Pancost : @ Yannick - nice!  Very keen to push into some critical biogeochemical questions. 
17:35:02	 From  Matthew Huber : I will be adding cGenie to my toolbox I think, so it may be possible for me (on some long time scales) to do (or collaborate with someone to do) simulations with that model using circulation forcing from these deepmip runs. that would enable a useful type of sensitivity study to complement others.
17:36:04	 From  Matthew Huber : it’s in the chat.
17:37:23	 From  Matthew Huber : In the Herold et al boundary conditions we included our old (Sewall et al) veg distribution interpolated to the deepmip paleogeography. I encourage everyone to do a sensitivity study with those, they are ready to go on the same grid (just do a slab or fixed SST run).
17:37:38	 From  Matthew Huber : It has a lot more rain forest in the tropics.
17:38:00	 From  Pierre Sepulchre : Agreed (with both Bette & Matt) !
17:38:41	 From  Pierre Sepulchre : I’d really like to run these sensitivity exp.
17:42:31	 From  Pierre Sepulchre : Just a quick comment : Last year we ran and published Eocene -only simulations with interactive chemistry in the strato. It’s quite expensive, but shows strong changes in atm temperature at high latitudes.. I don’t know if it’s something that should considered for future deepMIP runs. This time i really need to go, sorry !!
17:42:47	 From  Pierre Sepulchre : *atm-only simulations
17:44:22	 From  Matthew Huber : The Miocene is ripe for inclusion in DEEPMIP
17:44:54	 From  Anna Nele Meckler : it might be crucial to focus on a period with paleodata already available, as these take so long to generate. EOT?
17:45:02	 From  Gavin Foster : the Miocene is the new Pliocene
17:45:35	 From  gavin : It's all about Mi.
17:45:51	 From  Gordon Inglis : The Oligocene is the new Miocene. Just sayin'
17:46:04	 From  Kate Littler : Middle Palaeocene would be very interesting… but there’s not a huge amount of data to compare to. We have some stable isotope records, sparse TE records, and Michael is working on the pCO2 records… but we don’t have the coverage of the late Palaeocene/PETM etc.
17:46:36	 From  Kate Littler : This should spur us on to collect more data though!
17:46:37	 From  Matthew Huber : You’re right. We have an Oligocene model data paper in press. It will be of great interest to this community (and suggests a Deepmip linkage in a couple of years).
17:47:27	 From  Ulrich Salzmann : WE published in SciRep a data compilation for the EOT
17:47:46	 From  joepvandijk : I have to go people, thanks for this productive meeting. I am planning to expand terrestrial data into the Cretaceous and Permian/Triassic. Cheers!
17:47:47	 From  Kate Littler : EOT is a good target and is much more data-rich.
17:48:32	 From  igor : I have simulations from Late Cretaceous
17:48:37	 From  gavin : How about a filter on 1) the size of the observational signal, and 2) the plausibility of the forcing mechanism? 
17:48:51	 From  Chris Hollis : Our focus for the coming few years is the Miocene. Can you clarify if MioMIP can be included in DeepMIP?
17:50:11	 From  madeleinevickers : Is there anyone looking at more localised climates - e.g. is there the possibility of looking to see if there could have been localised cold outbreaks in the North Sea area
17:50:11	 From  madeleinevickers : ?
17:50:25	 From  madeleinevickers : (in the Early Eocene)
17:51:08	 From  madeleinevickers : (using models)
17:53:11	 From  Kate Littler : I guess if we’re expanding to other time periods we should reassess what questions we’re asking. So.. are we more interested in constraining background conditions (e.g., late Palaeocene) or transitions/events (PETM/EOT). These broader scientific questions should guide what we target.
18:02:14	 From  Gordon Inglis : BTW, when we use SST proxies ONLY to calculate EECO GMST, we are getting close to your value of 31'C (at least using Dan's method; Dsurf-2 in Inglis et al. 2020). But still bigger error bars that via DA
18:06:16	 From  Rich Pancost : Endorse isotopes!!!!
18:06:31	 From  Dan Lunt : Is it possible to run DA without the oxygen isotopes?  WHat happens when you use CESM without the modelled d18O.
18:08:03	 From  David Evans : It is an interesting paper but Mg/Ca biased from this kind of factor only in a very small number of cases - either v. low salinity or huge blooms to draw down [Ca]
18:11:11	 From  James Zachos : Again in considering observations (& biases) its the seasonal cycle of precipitation (dryer/wetter summers) that might be important for comparison. Especially for isotopes, O or D/H, and for picking new sites.
18:12:55	 From  Kate Littler : Q. How do we go about keeping our dataset compilations up to date, so that they incorporate new data as it is published? The Hollis et al., 2019 synthesis is great but it’s a snapshot of what we had in 2019. For example we published a new DeepMIP relevant dataset from the Indian Ocean (very under sampled) in EPSL recently, which would be really interesting to include in model-data comparisons going forward.
18:13:35	 From  Rich Pancost : Got to run - Thanks for the great talks and discussion and thanks to Dan for organising! 
18:14:31	 From  Kate Littler : Even something simple like a Google docs we could access and update...
18:14:50	 From  Gordon Inglis : I had looked into this with Matt, Chris, Kirsty etc, but still discussing options I think
18:16:04	 From  Gordon Inglis : Yeah, we have CSV for all DeepMIP data (average values only - not the time series). Will be in the SI for the Climate of the Past paper :) 
18:16:28	 From  Gavin Foster : There has (understandably) been a flurry of modelling activity recently, data keeps on coming though slowly and surely.  How much more modelling is expected? when is the next data-model comparison planned?
18:16:37	 From  Gordon Inglis : Agreed!
18:16:56	 From  Kate Littler : We have butted up against the same thing with plioVAR. Might be worth discussing with Erin McClymont to see what solution that group is coming up with. I think an online-database?
18:17:34	 From  Jessica E Tierney : PlioVar just uploaded average values to Pangaea as far as I know
18:17:58	 From  Kate Littler : There were efforts to have a “living database” for plioVAR but I’m not sure if that is still moving forward.
18:18:04	 From  Jessica E Tierney : gotcha
18:18:22	 From  Kate Littler : George Swan was taking the lead for PlioVAR on this aspect.
18:18:50	 From  Kate Littler : Swann
18:22:42	 From  Kate Littler : Got to go now. Many thanks to everyone for their contributions and to Dan for organising!
18:22:55	 From  Matthew Huber : I have 4x daily saves from all my Eocene runs. You can feel free to analyze those.
18:23:39	 From  Chris Hollis : Heiko Palike had proposed to build a DeepMIP database, that was supposed to be linked to the Hollis paper, but he got too busy I think. I wondered if a live data repository could be hosted at Bristol along with the model data.
18:24:08	 From  Agathe Toumoulin : I also have to go! Thank you for this nice day and for offering me to present my results!
18:24:26	 From  Gordon Inglis : I have to head folks. Thanks for all the great talks, and thanks to Dan for organising! I will keep looking into the DeepMIP database and how it might evolve. 
18:24:39	 From  Gordon Inglis : Feel free to get in touch if you have any ideas!
18:24:59	 From  Matthew Huber : I can host a repository for proxy datasets and model output at Purdue at no cost, but the issue is having people who will curate it and also the interface for access.
18:25:43	 From  madeleinevickers : Thanks everyone, got to go now!
18:27:51	 From  Natalie J Burls : Have to run, thanks for all the great talks and discussion, cheers
18:28:00	 From  David Evans : I also have to head - thank you for the talks and for organising Dan!
18:28:16	 From  Tina vdF : I am a big fan of spreading a hole day meeting into 3 x 2 hour meetings over three days.
18:28:34	 From  Charlie Williams : Me too!
18:29:18	 From  Xiangyu Li : Thanks for having a nice day with you. Thank Dan for organizing the meeting. See you, everyone.
18:29:38	 From  Ying Cui : Thank you all for a great meeting! I have to run but would love to hear more. Enjoy the afternoon/evening!
18:29:46	 From  Matthew Huber : Will you be doing online assimilation?
18:30:19	 From  Matthew Huber : you can run them on my machines.
18:30:27	 From  Gavin Foster : How about transient model runs? PETM etc. any in the pipeline?
18:30:29	 From  Matthew Huber : DART?
18:30:48	 From  Chris Hollis : This has been great Dan, but I agree that shorter sessions would be better for crossing time zones.
18:31:26	 From  Tina vdF : Thanks a lot Dan and everybody who presented and contributed to discussion! I am off now.
18:31:32	 From  jw2u12 : This has been fabulous; thanks everyone!
18:31:40	 From  Kasia_K_Sliwinska : I have to go soon too. Thank you for such an interesting meeting! And thank you for organizing Dan.
18:31:46	 From  Charlie Williams : I have to get going, sorry. Enjoy the social. Thanks to everyone for an interesting day, and great talks, and thanks to Dan for organising. Lots of work to do, so I'd better crack on over the coming weeks!
18:31:52	 From  Vittoria Lauretano : Thanks for the great talks and for the interesting meeting!
18:32:02	 From  Margaret Collinson : Sorry have to leave now. I am really glad to have had the opportunity to attend the meeting. Thanks Dan and everyone  involved in planning
18:32:17	 From  Ulrike Baranowski : Thank you very much for the meeting it has been super interesting.
18:32:35	 From  zhongshi : Thanks for the Meeting.
18:32:45	 From  Anta Sarr : Thanks everyone & Dan for the meeting !
18:33:20	 From  Yurui Zhang : It was  a great meeting. Thanks all and thank Dan~~
18:33:33	 From  Chris Hollis : Still a bit too drowsy for the breakouts so heading off, but will catch up on recorded sessions later. Thanks Dan
18:33:38	 From  Dan Lunt : Anyone who would like to stay for a social....come back in 8 minutes (X:40) and we will assign some random breakout groups.....bring a beer if you like and if it is an appropriate (or inappropriate!) time zone!!
18:33:38	 From  Loure102 : Thank you all for the great talks and discussions.