How to properly use ICEPACK 1.5.1 to simulate melt pond fraction during the MOSAiC drift?

[XudongLi]

Hi All,
I am currently attempting to use ICEPACK 1.5.1 to simulate melt pond fraction during the MOSAiC drift, but the simulation results appear very strange, as shown in the figure. I proceeded as follows:
(1) I downloaded the forcing data according to the user manual, namely MOSAiC_atm_drift1_precip_MDF_20191015_20200731.nc and MOSAiC_ocn_drift1_MDF_20191006_20200731.nc.
(2) The input parameters in icepack_in were modified based on the settings in set_nml.ocnmosaic and set_nml.atmmosaic.
(3) Because my focus is on melt pond fraction, I changed the simulation start date to 21 June 2020, while keeping all other parameters at their default values.

I also tested ICEPACK 1.5.2, but the results are similar to those from version 1.5.1. I have uploaded my input file icepack_in as well as the runtime log files.
I would greatly appreciate any comments or suggestions on this issue. Thank you very much!

best,
Xudong

 

[dbailey]

Part of the issue is that Icepack only outputs apnd and hpnd where the total pond volume = apnd*hpnd. These are indices 21 and 22 in your trcr array in the history files. This is the total amount of pond water and not the radiatively active pond fraction (apeff). However, you can edit the history output to add apeff to the history files or the log files. It would be worth checking the pond lid refreezing (ipnd) as well. Looks like you are getting some rapid refreezing events. It could also be on days when it was snowing from the MOSAiC data. Have you tried other forcing to see if it looks more reasonable?

 

[XudongLi]

Part of the issue is that Icepack only outputs apnd and hpnd where the total pond volume = apnd*hpnd. These are indices 21 and 22 in your trcr array in the history files. This is the total amount of pond water and not the radiatively active pond fraction (apeff). However, you can edit the history output to add apeff to the history files or the log files. It would be worth checking the pond lid refreezing (ipnd) as well. Looks like you are getting some rapid refreezing events. It could also be on days when it was snowing from the MOSAiC data. Have you tried other forcing to see if it looks more reasonable?

Hi dbailey,
Thank you for your comments on this issue. I tested the model using the default SHEBA data, but the MPF results are still very strange (see fig). The ipnd variable also does not show frequent pond water refreezing. I will try adding apeff to the output files. Thank you again.

best,
Xudong