[ERROR] cmorized ocean vertical velocity

[YanchunHe]

Quote A. Gjermundsen

I wonder if there is something wrong with the cmorized wo variables.
when looking at wo in gn grid the values range from -0.0012 to 0.001 m/s. For the same variable, experiment and time period in the cmorized gr files, wo values range from -0.0016 to 1.88e8. It seems to me that all bottom values (or where there is bathymetry) are given in the range 1e7 to 1.9e8 instead of nan-

E.g.
/nird/projects/NS9034K/CMIP6/CMIP/NCC/NorESM2-LM/piControl/r1i1p1f1/Omon/wo/gn/latest/wo_Omon_NorESM2-LM_piControl_r1i1p1f1_gn_160001-160912.nc

/nird/projects/NS9034K/CMIP6/CMIP/NCC/NorESM2-LM/piControl/r1i1p1f1/Omon/wo/gr/latest/wo_Omon_NorESM2-LM_piControl_r1i1p1f1_gr_160001-160912.nc

[YanchunHe]

You are right, @adagj . The noresm2cmor tool uses the surface grid to mask the ocean grid of missing values, so that it fails to identify the missing value points in vertically interpolated data of marginal seas.

It is hard to redo the cmorization of wo for all the experiments, but should be possible to report an issue at ESGF, that resetting all values larger than 1 should be a quick fix.

Alok, could you report this issue at ESGF for all the published wo on gr grid (data on gn grid are OK). @monsieuralok

I can reprocess wo on gr for all CMIP and SenarioMIP experiments, if you find useful, Ada!

[adagj]

@YanchunHe
For my sake, you don't need to reprocess anything. I set all wo larger than 1 to nan (as you also suggested) in my scripts and the results seem reasonable. I just wanted to let you know :)

(ps. these large wo values are not just marginal seas, but all "bottom" values)

Ada

[YanchunHe]

Alok, could you report this issue at ESGF for all the published wo on gr grid (data on gn grid are OK). @monsieuralok

A quick fix is to set wo to NaN when it is larger than 1.

[monsieuralok]

@YanchunHe would this issue will be fixed in future version? or if already some data are published with fixed?

[monsieuralok]

@YanchunHe https://errata.es-doc.org/static/view.html?uid=62a25f76-78f6-e46f-9a7c-e890ee22e518 I have mentioned it on errata.

[YanchunHe]

Thanks Alok! Cmorized datasets (after version v20210521) in the future will not have this issue. Already published data will be not changed. This errata should be enough.