Ekman pumping

[PaulSpence]

Evaluate Ekman pumping in Denman region

[PaulSpence]

Longterm mean Ekman pumping (1958-2019) in ACCESS-OM2-01 forced with JRA-55do (+ve upward)

[PaulSpence]

Time-series of annual mean Ekman pumping area integrated on shelf in Denman region

A slight decrease in downwelling on the shelf since 1980s.

[ribeiron]

Hope you don't mind I barge in, as I try to organize my thoughts for my own work below:

I see the same thing on the Vincennes Bay area (slight decrease in downwelling), but unfortunately, the integrated ekman transport within the shelf (south of 1000m isobath) doesn't correlate with the mCDW intrusions we see.

The average integrated ekman transport on the shelf (south of the 1000m isobath) for years of intrusions and years of no intrusions is -1354.01m and -1413.19, respectively.

[ribeiron]

The good news is, there is a paper in the Totten (Nakayama et al 2021) that says they have identified positive Ekman speed when ASC is weak and negative when it is not. I might have missed it, but they don't provide any correlation coefficients/relationship etc. In any case, I set up in a new wild goose chase this Easter and I found a correlation between Ekman pumping on the shelf break with both the shoaling of the isopycnals in Vincennes Bay (which correlates with mCDW intrusions) and weakening of ASC (which correlates with the isopycnal shoaling) + the actual intrusion. I am hoping this is the smoking gun I needed it, but I need to run this through Adele before I get to excited.

Hope this helps you on your own chase.

[PaulSpence]

@ribeiron nice plot! I will look into the temp relationship with sea ice production now.

[PaulSpence]

@ribeiron Thanks for the chat yesterday. I played around a bit with Ekman pumping (and its components) correlation with shelf temp in Denman and Vincennes Bay, but didn't find any strong correlations. Maybe my area box is bigger than yours (e.g. xt_ocean<-255 & xt_ocean>-247 for Vincennes Bay).

[PaulSpence]

[ribeiron]

When I split the ekman components in Vincennes Bay, the Y component (taux) is the one that correlates the most in every bit of the shelf that I choose. I am not sure what is the implication of this for the easterlies experiments, where we see that the meridional wind is the one driving the intrusions , but I will discuss that with Adele tomorrow (live) cause I just got to Camberra :) These are the numbers for the near shelf correlations (the orange bit of the ekman figure): Cor EKMAN X NEAR SHELF vs TEMP <xarray.DataArray ()> array(-0.3978305) Cor EKMAN X NEAR SHELF vs ISO 27.82<xarray.DataArray ()> array(-0.00919596) Cor EKMAN X NEAR SHELF vs ASC <xarray.DataArray ()> array(0.07822677)

Cor EKMAN Y NEAR SHELF vs TEMP <xarray.DataArray ()> array(0.46513873) Cor EKMAN Y NEAR SHELF vs ISO 27.82 <xarray.DataArray ()> array(-0.46909226) Cor EKMAN Y NEAR SHELF vs ASC <xarray.DataArray ()> array(0.31915267)

If a lag is introduced Cor EKMAN Y NEAR SHELF vs TEMP becomes 0.7 in 4 months. You can find the script here: "/g/data/v45/ns3783/scripts/Corr_separate_ekman.ipynb". I can't upload files to where your codes are stored here, but let me know your path on gadi, and I'll just transfer them to you there if you can't access my folder.

[ribeiron]

This is the figure for the anomaly for peak and low shelf temperature years for the ekman transport.

[ribeiron]

Here is the average of all years for each component and all together.

[ribeiron]

And here is the anomalies for each ekman component in years of peak and low shelf temperatures in vincennes bay