Mira Berdahl (Apr 21 2022 at 19:30): Mira Berdahl (Apr 21 2022 at 19:30):Hi,
I'm trying to read the N_HEAT variable from POP ocean output for the first time, and having an issue when specifying the transport region. The data looks fine when I choose transport_reg = 0 (global). However, when I choose transport_reg (atlantic), all the data comes back as nans. When I look at the file in ncview, I see some nan's (which I expected to see), but there are also actual data there as well that apparently aren't being read or detected. I have attached a screenshot showing a simple example of this.
Does anyone have experience with reading N_HEAT and know how to get around this?
Thanks!
Screen-Shot-2022-04-21-at-12.23.46-PM-2.png
Stephen Yeager (Apr 21 2022 at 21:24):This is because the diffusion component is missing for the Atlantic region and so the "Total" component is also missing. The diffusion component is generally small, so a good approximation is to compute the total advection. This will yield a reasonable plot:
N_HEAT.isel(transport_reg=1).isel(transport_comp=[1,3,4]).sum('transport_comp')
Oh, excellent. Thanks for the explanation!
Mira Berdahl (Apr 21 2022 at 22:24):Actually, one follow-up... any idea why this component is missing? How worrisome is it that we don't know the diffusion component's contribution?
Mira Berdahl (Apr 22 2022 at 01:55):Or was this omission intentional for some reason?
Stephen Yeager (Apr 22 2022 at 12:51):I've forgotten why this component is missing. Perhaps @Keith Lindsay knows? I can also ask Gokhan when I get a chance.
Keith Lindsay (Apr 22 2022 at 13:33):There is a comment in the code that states
! - because southern boundary diffusive transports are not available,
! the total and diffusive transport components are not computed
! for regions.
This comment was last modified in March 2006. So it was either written then, or predates that. I think the meridional transport was written by Gokhan. I later wrote code to write out lateral diffusive tracer fluxes. This was added to the code in April 2010. It seems that one could have gone back and updated the meridional transport code to use the lateral diffusive tracers fluxes on the regional boundary, but this was never done.
Mira Berdahl (Apr 22 2022 at 16:50):Thanks @Stephen Yeager and @Keith Lindsay , this is good to know.
Last updated: May 16 2025 at 17:14 UTC