Explaining spatial variation of correlations in density space about asc_and_heat_transport HOT 1 OPEN

Attaching my slides from the OMO meeting here. Morrison_ASC_heat_transport.pdf

For the Getz and Totten transects, it looks like the velocity structure in the Totten remains unchanged between strong and weak ASC, while the ASC penetrates deeper in the Getz while it speeds up. Velocities on the Totten are also much greater than on Getz.

Could there be a threshold where for a weaker ASC, the CSHT and ASC speed are correlated, while for stronger ASC there is little difference? i.e. for ASC stronger than a certain threshold, the along-isopycnal pathway for CDW intrusions is still blocked, at least on an interannual timescale? I guess this is kind of inline with what you were thinking @adele-morrison, where the changing sign of the ASC could make a difference.

Based on Adele's results, I redid the correlations only including locations where the ASC speed remains below a certain threshold in the given time series (mask has dimensions (lon_bins, density)). The hypothesis is that regions with lower ASC speed have greater correlations.

Since we mainly care about the CDW layer, the filled plot shows the correlations in the CDW layer for locations where the ASC speed is below a threshold velocity. As the threshold increases, more longitudinal bins are included. Although the highest correlation regions are included at lower thresholds, the link between low threshold speed and high correlation does not hold past threshold u= 0.05m/s. Even when looking at threshold u = 0.04 m/s, there are still regions of lower correlation (r^2 ~ 0.3-0.45) that are included. The threshold velocity I think could be an explanation for the lack of correlation in certain regions e.g. Totten, but definitely is not the full story.

The blank spaces are explained by the lower panel, with the correlations in each density layer. In the blank regions, the CDW layer does not exist, which might be filled by the correlation of the next densest layer. So in the WAP and Weddell, the water at the 1000m isobath is lighter than the 27.82 kg/m3 bin.

The same plot as above using annual averaged correlations is below:

Motivating slide from Adele's presentation below:

from asc_and_heat_transport.