Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

Fractional sea ice with fixed surface temperatures

This post was from a previous version of the WRF&MPAS-A Support Forum. New replies have been disabled and if you have follow up questions related to this post, then please start a new thread from the forum home page.


New member

I am trying to setup idealised simulations of flow over broken-up sea-ice, using the fractional_seaice option of WRF. I am using the "real" version of WRF, but manually create the met_em-files through a separate tool. I am trying to set up the simulations such that all ice and all ocean have the same surface temperature, respectively, and then would like to vary how ice and ocean are distributed in the domain. For partially ice-covered grid cells I am linearly blending the surface temperatures of ocean and ice in the input data. Surface energy balance is switched off (sf_surface_physics=0), as is radiation, so I am expecting surface temperatures to remain unchanged throughout the simulation.

The procedure seems to work in general, and in particular real.exe produces input and boundary files that look correct.

WRF itself however produces unexpected results. I tested both with WRF-3.9.1 and WRF-4.3 (using the same inputs with force_use_old_data=T), but the following symptoms remain the same.
  • With the YSU parameterisations (sf_sfclay_physics = bl_pbl_physics = 1), surface temperatures remain unchanged as expected. But for fractional sea ice cover between approximately 0.5-0.9, the simulated heat fluxes become more negative with more open water. This seems rather counter-intuitive, but may be it is just a weird corner case for this parameterisation and everything is actually working as designed. Initial stratification is stable (though not excessively so), so some downward mixing of heat over ice is expected.
  • With the MYJ parameterisations (sf_sfclay_physics = bl_pbl_physics = 2), it looks to me like I have uncovered a bug. Surface temperatures are not constant, but (increasingly) rapidly decrease over partially ice-covered grid cells in the initial couple of time steps until they fall below TICE_MIN and thus are reset to 221.4 K. It seems like MYJ is overwriting TSK by TSK_ICE at every time step, but then fails to reset TSK back to its blended value.

Can you comment on whether the observed behaviour of the YSU parameterisations is as designed? Or am I still doing something wrong in the way I prepare the input files? I am happy to provide a sample met_em-file, but it seems I am not allowed to attach them here. Further, can you comment on whether the observed behaviour of the MYJ parameterisation is a bug as I suspected?

Many thanks in advance for your help!