Arctic temperature Bias in G-MOM

[whokim]

I have generated Arctic temperature bias plots, comparing the two vertical coordinates in G compset (ie., g.e23_a16g.GJRAv4.TL319_t232_zstar_N65.2024.004 and g.e23_a16g.GJRAv4.TL319_t232_hycom1_N75.2024.005) along with a compatible POP 1º simulation. In short, all simulations have a similar degree of cold biases at the Atlantic Water layer (~200-600 m), but Hycom1 appears to have a weaker bias than zstar.

Eurasian and Canadian Arctic temperature bias relative to WOA18 for MOM and WOA13 for POP as a function of time and depth (3rd and 5th cycles are missing for POP)_

4th cycle-averaged temperature bias at ~50 m, ~300 m (roughly the core of the Atlantic Water layer), and ~900 m relative to WOA18 for MOM and WOA13 for POP

[dabail10]

Here are some plots from CC Bitz. She is finding that there is a lack of frazil formation just north of Svalbard. Looks like the "heat available" or negative part of frzmlt is very big. On the order of -200 W/m^2. This is preventing ice growth right above the region where the warm salty Atlantic water enters the Arctic Ocean.