I am developing an Operational Forecast System for the Cook Inlet region of Alaska. In the winter time, air temperatures can fall to -40C or below, the humidity can fluctuate between 30%-100%, there are winds up to 35 m/s (~80 mph) and the air pressure fluctuates between 960mb and 1060mb. So conditions are pretty extreme and Hernan has said on multiple occasions that the Bulk Flux formulation in ROMS (based on the TOGA-COARE experiments in the equatorial Pacific) is only valid for mild meteorological conditions.
In conditions such as these, should we continue to use the Bulk Flux formulation in ROMS or is there a better practice to follow (e.g. impose wind stress and net heat fluxes from an met model, run ROMS in a ocean-atmosphere coupled mode, etc.)?