GOSI10: Local ME s-coordinates in the Nordic overflows area

[oceandie]

The aim of this issue is to assess the impact of the improved Noridc overflows on the the global circulation and biases.

Implementing localised Multi-Envelope vertical coordinates in the Greenland-Scotland ridge region can help to significantly improve the representation of the Nordic Seas overflows at eddy permitting resolutions - see Bruciaferri et al. 2023.

For this issue, we plan to run the following experiments:

1) long intergrations (45y) with GOSI9 eORCA025 forced with JRA with both zps and MEs 2) long intergration (45y) with GOSI9 eORCA12 forced with JRA with both zps and MEs 3) long integration (100y) with GC5 with both zps and MEs

If everything goes as planned and we decide to implement local MEs in the Nordic OVF region of GOSI10, then the following tests should be conducted to check that we can replicate the results of GOSI9 with JRA:

1) long intergrations (45y) with GOSI10 eORCA025 forced with ERA5 (?) with both zps and MEs 2) long intergration (45y) with GOSI10 eORCA12 forced with ERA5 (?) with both zps and MEs

Note that the experiments with GOSI10 should use the djc scheme instead of the prj to accurately compute horizontal pressure forces and the envelopes of the local MEs-coordinate system should be optimised using 3 months long HPG error tests (instead of a 1 month), as suggested in the review of the JAMES paper.

[oceandie]

All the experiments will differ from the standard GOSI9p8.0 of Guiavarc'h et al. in the following numerical settings/choices:

1) JRA forcing 2) TRIADS scheme for the isoneutral mixing 3) logarithmic bottom drag coefficient 4) Pressure Jacobian scheme for computing horizonatal pressure forces

In addition, the exp using local ME s-coord will switch off the BBL parameterization (i.e., ln_trabbl=.false.).

[oceandie]

The suites to run the experiments at 1/4 of degree of resolution are the following:

1) u-db981@274132 implements GOSI9-025 with $z^*ps$ coordinates and is run by @ukmo-cguiavarch 2) u-db983@274131 implements GOSI9-025 with local ME$s$ coordinates in the Nordic overflows region and is run by @oceandie

[oceandie]

The experiments are finished and the results are available on mass at moose:/crum/u-db981 and moose:/crum/u-db983 .

[oceandie]

Validation notes after running MARINE_ASSESS can be found here:

1. GOSI9-025 with $z^*ps$: u-db981_nemo_vs_u-cn082_nemo