Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

Huge sea-ice thickness

Hi Nan, I was controlling the sea-ice thickness of the MIS 6 simulation. It is huge, up to 50 m after 100 years of simulation.I guess that the climate is so cold, that all precip is turned into snowfall and accumulates on top of the sea-ice.The problem is that as far as I understand, snow cover is turned into ice according to Achimede's law and it seemsand infinite process. Is there any parameter to control that? Did you had a similar problem for LGM run? Flo
 

nanr

Member
Hi Flo - first, a few questions:  1)  What version of cesm are you using?  2) is the effect wide spread or can you tell whether it's a few isolated points?nan
 
Hi Nan, I am using CESM 1.0.5 and the effect is wide spread. The ice thickness gradually increases along the experimenteverywhere in Arctic. In CICE user guide they said that the fact the snow cover on top of sea-ice is turning into ice via Archimede's law can in some cases produces a very thickice cover...So I was wondering if I had to increase the ice melting under the sea-ice cover to compensate? Flo
 

ottobli

New Member
Flo,
Our CCSM4 LGM simulation generally had annual sea ice thickness in the Arctic less than 10 meters. A very few grid points in the Beaufort Sea along the coasts of Alaska and Yukon had annual thicknesses of 30-40 meters.
Bette
 
Hi Bette, Hi Nan, I see.. i was wondering if it could be due to different albedo values in CICE namelist associated with a particularly low ice transport in the Arctic.For CESM1.0.5 versione, the default values in CICE namelist are:  albicei                = 0.45 albicev                = 0.75 albsnowi               = 0.73 albsnowv               = 0.98 What are the values that you used in your LGM simulation? Flo
 
Dear Bette and Nan, I carefully checked sea-ice thickness and in central Arctic it reaches 21 m and 26 m along the coasts, and as for you, I have small areas reaching more than 30 min the Beaufort Sea, along the Alaska and Yukon.I decreased the albedo values just to see what was the effect, but except in the North Atlanticwere the sea-ice variability is larger, in central Arctic, ice thickness will not move much from its actual values. Actually, I don't think that this is so wrong because the Eurasian MIS6 ice sheet is much higher and larger thanthat of LGM. It confines ciruclation in Central Arctic, generating large negative temperatures. Since all the othercomponent variables are fine, I guess that CICE reacted to the strong cold topography in input. Maybe the ocean bellow is also toocold and heat fluxes are small. SIC values seem quite stable now. I have performed 100 years of simulations. I will let it run until 400 years and I will see. what do you think? Flo
 
hi, Flo,I also use CESM1 to simulate LGM, and come across the ice thickness problem. I wonder if you have checked the thickness of land-ice, if yes, could you tell me which variety stand for it? And is the thickness of either land-ice or sea-ice as one of your initialization to run up the model?Expected your replyTaily
 
Hi Taily, I am not sure to have understood your question. I have never simulated LGM so far. It was another glaciation. But I solved the problem thanks to David Bailey.It is a problem with land-ice runoff to the ocean. Please read this post:https://bb.cgd.ucar.edu/huge-arctic-sea-ice-thickness Flo
 
Hi flo,Thanks for your reply and suggestion. Have you ever checked the variable of the land-ice thickness? In fact, I have found a variable in the land model, namely GC_ICE1, but I still can't be sure if it has some relations with the land-ice thickness. Do you have any idea?Thanks again.Taily
 
Hi Taily, the variable attributed to GLC are not really active during your runtime since for the moment the ice sheet model is not completely coupledto the other components of CESM. However, when you specify some ice-covered areas, the model divides in vertical layers your ice sheetsbecause CISM has some vertical layers, as the other component of CESM. So if you look at this variable, you might have n levels containinga portion of your ice sheets from the base to the top. But as I told you, it is not active during runtime for the moment. So it might not be thecause of your problem. Flo
 
Top