Last Millennium, CCSM4: restart and forcing files

[born@climate_unibe_ch]

Dear all:  I would like to set up a series of sensitivity runs to decompose the effects of solar and volcanic forcing and compare the results with the CCSM4 transient simulation of the last millennium (b40.lm850-1850.1deg.001). For this I need:1) three sets of restart files from the end of the 850 AD control run that was used to initialize the LM simulation (I suppose). I need three to start a mini-ensemble.2) the full set of forcing files used in the LM simulation so I can use them component-wise in my runs.3) some information on the model version that was used for the transient LM run.Thank you very much,  Andreas

 

[nanr]

Hi Andreas - Thank you for your interest in the LM CCSM4 simulation.  I will compile the files I think you needand let you know when they are ready.
Nan Rosenbloom

 

[born@climate_unibe_ch]

Thank you! Let me know if you need more information. I hope it is not too much trouble!  Andreas

 

[nanr]

Hi Andreas - I have a few questions:1)  where are you running the model?2) what version of CESM are you planning to use. The CESM release version most similar to what we used for the LastMillennium simulations is CESM1.0.5.3)  The NCAR LM simulation used the B_1850-2000_CN compset.  Do you have access to that compset?4)  Do you have access to the NCAR SVN data repository (https://svn-ccsm-inputdata.cgd.ucar.edu/)?  I list the forcing files used for the LM simulation on the webpage listed below.  Click on the filename to download the file directly from the SVN data repository.   If you have registered for CESM releases, you can use your username and password.  Or you can use the guest account- http://www2.cesm.ucar.edu/working_groups/Paleo/docs/special-projects/last-millennium-ccsm4
5)  I will retrieve three restart files and let you know when they are available for download. Best - Nan Rosenbloom, PWG liaison 

 

[born@climate_unibe_ch]

Hi Nan,  thank you very much for the data. There was a problem with the ozone file which appears to be incorrectly linked. The answers to your questions (plus some follow-up questions):ad 1) I am running on a CRAY XE6 (Monte Rosa) at CSCS here in Switzerland.ad 2) I have an older version, 1.0.1, which I was planning to use. I suppose the differences between 1.0.1 and 1.0.5 are not huge, or did something change that is important for th LM run? In any case it is CAM4, correct?ad 3) I can create and configure the B_1850-2000_CN compset. However, since I was planning to decompose the solar and volcanic forcing in sensitivity experiments, I would prefer to run the model without the land biosphere part, i.e. compset B_1850-2000 without CN. The data you sent should be compatible and consistent with that, right?ad 4) I downloaded all the files except for the ozone (see above).I will start playing around with the files you have sent so I know what I need to do once I use the proper restart files. Could you point me to a description of the 850 AD control run? It would be good to know how long it was run and what boundary conditions were used.Best,  Andreas

 

[nanr]

Hi Andreas -I answer your questions below, but first I have a question about the restart files you want from the LM control simulation.  Do you have a preference for how I choose these restarts?  Should I take the last three restarts?  Or stagger them by 5 years, or some other increment?  Unfortunately, I can't locate the diagnostics that were run on the control simulation, so I can't point you to a timeseries analysis that might indicate one time period over another.1)  I fixed the ozone file link; thanks for letting me know about the problem.2)  cesm1.0.1 should be fine.  cesm1.0.5 is the version that was ported to our new supercomputing center, but it should be comptible with the version you are using. 
CAM4 is correct.3)  All the files, including the LM B_1850-2000CN restart files, should work fine with B_1850_2000. 4)  I added a description of the LM control simulation to the LM webpage (https://www2.cesm.ucar.edu/working_groups/Paleo/docs/special-projects/last-millennium-ccsm4).  I also added a figure that graphically shows the relationship between the 1850 PI control, the LM control/spinup, and the Last Millennium simulation. Best regards - Nan 

 

[born@climate_unibe_ch]

Hi Nan,  once more thank you very much for your help. The figure is excellent and I now have almost everything I need to start my experiments. I had a problem opening the description of the LM run because it asks for a login and password which apparently are different from this forum´s and the SVN.Concerning the restarts, I might not need the 850 AD run after all, I am sorry. I have been working intensively with the files you sent today and yesterday and from what I understood I might just use the B_1850 compset and modify it to my needs. So I would use a 1850 AD restart which anyway is quite similar to 850 except the land use. The 1850 restart from the SVN is from model year 350, I think. Do you have a more recent one? I think the 1deg control experiment (b40.1850.track1.1deg.006) was run for 1300 years. One restart would be enough in this case because I could easily create some more myself.Best,  Andreas

 

[nanr]

Andreas - I am a bit confused about what you are looking for in your restart.  The LM control branched from year 301 of the 1850 control, but then continued independently for 100 yrs with the LM forcing datasets (e.g., landcover, CO2, etc), before launching the LM simulation.  So it will be quite distinct from the PI control.  But maybe I don't understand what you need for your ensemble.

 

[born@climate_unibe_ch]

Hi Nan,  sorry for the confusion. I understand that landcover and GHGs are slightly different in 850 but that should not matter for my analysis. I have looked at the transient LM run in some depth so that the only purpose of the new decomposition runs is to answer one very precise question.In the meantime since I started this post and after seeing the forcing files you sent, I decided to prescribe an idealized solar minimum resembling the Maunder minimum and the volcanic forcing will be taken from the series of eruptions after 1250. I am confident that these changes have a stronger signal on the oceanic mechanism I am looking at than changes in landcover or anything else, so taking the 1850 control is OK and would even have the advantage the the ocean is closer to equilibrium, hopefully.