Question about the typhoon-induced sea surface cooling in ROMS-WRF coupled simulation

[leo_lee]

Hello everyone,

I'm trying to simulate a super-typhoon in the WRF-ROMS coupling model through COAWST, now I have something unclear, and need your help.

By examining the model output, I compared the simulated SST with the high-resolution microwave SST data obtained as raw sensor data and as daily optimally interpolated data (MWOI), it is denoted that the SST from the ROMS is about 1-2℃ colder than the MWOI SST. I heard that the over-mixing problem often occurred when using ROMS to simulate typhoon processes so that over-cooling happened. I have extracted the surface forcing from the WRF outputs, and ran the ROMS-only simulation, the over-cooling problem still existed. I have used both the GLS mixing scheme and the KPP scheme, the over-cooling problem didn't get much improved.

Therefore, if I want to adjust the parameters in ROMS mixing scheme to improve the simulated sea surface cooling, which parameters should i adjust to alleviate the over-mixing problem?

Thank you in advance.

Leo

[jcwarner]

it looks like the ocean SST is higher in roms all around than in the GHRSST. so maybe your initial conditions are not matching the SST data even before the storm. How did you set the initial stratification? perhaps that could be too thin of a surface mixed layer allowing the cooler temp to mix upwards more easily.

suggest you look at some vertical profiles of the ocean along the TC track, in time, to see how the vertical structure of temperature is changing.
how well do the winds from WRF compare to observed? stronger winds can cause more mixing.
-j

[jivica]

There are many variables in the play;
a) what about your surface model layer thickness? SST is usually referenced to 0.5m and if your model cell is ~3m then you are comparing different things.
b) if you are using WRF winds then possibly BULK_FLUX is not good (Cd is more complex than it is coded currently).
For the case of strong winds, I changed ROMS bulk_flux to be more realistic (using Hwang, 2020 formulation) producing wind saturated Cd values - important for strong winds (>30m/s).
Similar is done inside WRF if you used 2-way coupled system, note that WRF is sending *wind stress* to ROMS.

Cheers,
Ivica

[ben108472]

REMSS OISST is interpolated product based on microwave, IR channel.
If the satellite data is not continuous, the boundary will be very noise.
The situation is extreme in the region when the SST gradient is very high.
This case is Trami, so SST cooling is very obvious due to its slow translation speed.