Sea ice albedo change

[deepashree_dutta@student_unsw_edu_au]

Hi all,

Can I change sea ice albedo in FAMIP comset by adding the following in user_nl_cice:
albicev = 0.07
albsnowv = 0.07
shortwave = 'default'
In case it works do I need to change near IR albedo too?

Thanks a lot!
Deepashree

 

[dbailey]

This depends on what you are trying to do. If the incoming shortwave from the atmosphere is too low then you would want to decrease these albedos and vice versa. These values are extremely low for visible albedos. Are you sure you want to do this? Also, why are you using the older CCSM3 shortwave ('default') instead of the delta-Eddington ('dEdd')?

 

[duvivier]

We have a short FAQ about how we recommend tuning sea ice albedo posted: https://bb.cgd.ucar.edu/cesm/threads/faq-albedo-radiation-tuning-in-cesm-cice.2649/

 

[deepashree_dutta@student_unsw_edu_au]

Thank you so much!

I'm running a polar amplified experiment in which I've removed the sea ice cover (by making ice_cov = 0 in AMIP SST input file). But this increases surface temperature and reduces surface albedo at the same time.
I want to separate the impacts of sea ice albedo reduction only (without surface warming) and I thought the easiest way to do this is by changing albicev and albsnowv in user_nl_cice while keeping the ice cover same as AMIP.
I'm a little bit confused if it'll work.

In case it works, should I also change albedo_type to 'constant'?
Thank you for correcting - shortwave = 'dEdd'.

This depends on what you are trying to do. If the incoming shortwave from the atmosphere is too low then you would want to decrease these albedos and vice versa. These values are extremely low for visible albedos. Are you sure you want to do this? Also, why are you using the older CCSM3 shortwave ('default') instead of the delta-Eddington ('dEdd')?

 

[dbailey]

Actually, if you are using shortwave = 'dEdd', then the albicev and albsnowv will not do anything. For this type of experiment I would definitely change shortwave to 'default' and lower all of the albedo values to the same as the ocean (0.06). Now, keep in mind that the concentration is read in for an F compset and there will still be ice. So, that means that snow can still accumulate on the sea ice. However, if all of the albedos are 0.06, then it means the snow and ice are "transparent" and look like ocean. I haven't used albedo_type = 'constant' before. However, the issue is that it will set the albedos to internal values:

real (kind=dbl_kind), parameter :: &
warmice = 0.68_dbl_kind, &
coldice = 0.70_dbl_kind, &
warmsnow = 0.77_dbl_kind, &
coldsnow = 0.81_dbl_kind

So, I would not use this option for your experiment.

 

[deepashree_dutta@student_unsw_edu_au]

Actually, if you are using shortwave = 'dEdd', then the albicev and albsnowv will not do anything. For this type of experiment I would definitely change shortwave to 'default' and lower all of the albedo values to the same as the ocean (0.06). Now, keep in mind that the concentration is read in for an F compset and there will still be ice. So, that means that snow can still accumulate on the sea ice. However, if all of the albedos are 0.06, then it means the snow and ice are "transparent" and look like ocean. I haven't used albedo_type = 'constant' before. However, the issue is that it will set the albedos to internal values:

real (kind=dbl_kind), parameter :: &
warmice = 0.68_dbl_kind, &
coldice = 0.70_dbl_kind, &
warmsnow = 0.77_dbl_kind, &
coldsnow = 0.81_dbl_kind

So, I would not use this option for your experiment.

Thanks a lot for your suggestion!