Inconsistent wind stress between CESM2 piControl from CMIP6 archive and that from Campaign

[apauling]

Hello,

I was recently trying to compare the surface zonal wind stress from a run of CESM2 I have done to that from the CMIP6 piControl run. I first got the variable tauu from the CMIP6 archive, but it looked strange compared to TAUX from my run. (I have flipped the sign of TAUX in my run so that it is consistent with the convention from CMIP6). I then got TAUX from /glade/campaign/collections/cmip/CMIP6/timeseries-cmip6/b.e21.B1850.f09_g17.CMIP6-piControl.001. This looks much closer to TAUX from my own run, and noticeably different to that from the CMIP6 archive over land. I have attached a plot comparing tauu from the CMIP6 archive and TAUX from the timeseries output (with the sign flipped).

Was something different used for the surface zonal wind stress over land for the CMIP tauu variable?

Thanks!

 

[islas]

Hi Andrew,

The CMIP6 variable includes all stress between the surface and the atmosphere, so it includes the stresses due to the sub-grid orographic schemes as well. I think if you some up TAUX, TAUBLJX and TAUGWX from the CESM2 output you should find better agreement with the CMIP6 tauu variable.

Isla

 

[strandwg]

To expand on Isla's comment, "tauu" is defined as "-1*(TAUX+TAUBLJX-TAUGWX)" so if you make the same calculation you'll get better results.

 

[apauling]

Hi Isla and Gary,

Thank you both for your quick responses, I will try using that combination of variables.

Cheers,

Andrew