sea ice albedo (r_ice r_snw and r_pnd) value for present day

[littledddna@gmail_com]

Dear all, I have a question about namelist variable r_ice in the ICE model. I have a LGM run and I want to do another sensitivity run with present day sea ice cover. Looks like I can achieve that by modifying r_ice, r_snw and r_pnd in the ice model. But I am not sure what the new value should be for present day. Any suggestions or links that I can learn about it?

Thanks very much in advance

 

[dbailey]

Moving this to Paleo forum. This is a very good question. We set r_snw, r_ice, and r_pnd based on standard deviations from present day observations. We mostly adjust r_snw generally somewhere between 1.25 and 1.6. This is mostly to compensate for biases in the incoming shortwave. This says that the nonmelting snow grain radius will be as follows:

rsnw_nm = rsnw_nmbase - r_snw * rsnw_std

where rsnw_nmbase = 500um and rsnw_std = 250um. For a LGM climate, do you expect weaker incoming shortwave than present day? I would say start with present day values and run for 10 years or so. Then check the top of the atmosphere imbalance and the sea ice thickness. If this is what you expect, then stay with the r_snw from present day. If you need thicker ice, then adjust r_snw up a bit. @jiangzhu has done LGM simulations and can speak to more of the tuning.

 

[wangsz@fio_org_cn]

Moving this to Paleo forum. This is a very good question. We set r_snw, r_ice, and r_pnd based on standard deviations from present day observations. We mostly adjust r_snw generally somewhere between 1.25 and 1.6. This is mostly to compensate for biases in the incoming shortwave. This says that the nonmelting snow grain radius will be as follows:

rsnw_nm = rsnw_nmbase - r_snw * rsnw_std

where rsnw_nmbase = 500um and rsnw_std = 250um. For a LGM climate, do you expect weaker incoming shortwave than present day? I would say start with present day values and run for 10 years or so. Then check the top of the atmosphere imbalance and the sea ice thickness. If this is what you expect, then stay with the r_snw from present day. If you need thicker ice, then adjust r_snw up a bit. @jiangzhu has done LGM simulations and can speak to more of the tuning.

I found in ice_shortwave.F90, r_snw/r_ice/r_pnd are used to change the albedo. But we can also change the sea ice and snow albedo by setting albicev/albicei/albsnowv/albsnowi in ice_nml. So what's the difference between these two groups of parameters?

Thanks in advance :)

 

[dbailey]

The second set of albedos only works with shortwave = 'default' or 'ccsm3'. That is these were used for the older CCSM3 shortwave and not the delta-Eddington.

 

[littledddna@gmail_com]

Dear David,

I've come to check this and ask you again though it was a long time ago:) R_SNW is in the CICE namelist, but r_ice and r_pnd are not. Should I change the default value (which is zero) in ice_ic.F90 if I want to change the albedo using delta-Eddington？

Thanks very much again

 

[littledddna@gmail_com]

I just add these variables in namelist_definition.xml and change r_ice and r_pnd to -2. It worked. Hopefully, the sea ice will decrease in the model.