Reduce tracer biases in the Black Sea

[gustavo-marques]

The purpuse of simulations 024 and 025 is to check if narrowing and shallowing the Bosphorus and Dardanelles straights reduces the tracer biases in the Black Sea. These runs should be compared against 022 and 023, respectively.

Below are temperature and salinity drifts from the WOA18 annual mean for the Black Sea. The biases are signfincatly reduced in 024 and 025. Interstingly, the near-surface salinity bias flipped sign in the new runs. I think we should incorporate the changes made in these runs into our next baselines.

Temperature bias in the Black Sea

Salinity bias in the Black Sea

cc'ing @klindsay28, @fobryan3

[klindsay28]

Here are some line plots of 1000m averages over the Black Sea for tracers:



The topography and width edits are successfully removing the spurious deep trends in thetao and so.
Additionally, agessc over this initial segment appears to be equal to time, while in the control runs it had a smaller trend, indicative of younger Med water being transported in to the Black Sea.
These are all in the desired direction.

Here are some plots that I don't quite understand when taken together:



So sst is slightly up and sos is down with the mods, but mlotst and `oml (not shown) are up.
I'm not sure why the mods lead to deeper mixing. I'm not particularly concerned by this, but do find it curious.