[ecco-support] Budgets in ECv4, release 3

[Martha Buckley]

Hi Ou,
Thanks for the reply.  Can you tell me more about the type of budget that
you closed, e.g., the diagnostic (mass, heat, salt, etc) and the
layer/geographical area.  The reason that I am asking is that a budget over
a single grid cell may not need 64-bit precision, but when integrating over
depth, etc, the errors may accumulate.  I am currently looking at budgets
over the top 1000m for mass, heat, salt.

A very long time ago, I did a budget calculation using MITgcm output that
did not close due to 32-bit output, and I changed to 64-bit output and it
closed (since then I have always used 64-bit output so I haven't verified
that this is necessary).

thanks,
Martha



On Thu, Nov 2, 2017 at 12:03 AM, Wang, Ou (329B) <Ou.Wang at jpl.nasa.gov>
wrote:

> Hi Martha,
>
> I have closed the heat budget for the global ocean using the nctiles
> output available on ftp://ecco.jpl.nasa.gov/Version4/Release3/. The ratio
> of STD(tendency-convergences-forcing)/STD(tendency) is ~1e-7.
>
> Although the netCDF files are all in 64-bit precision, some of the
> original binary diagnostic outputs that were converted to the tiled netCDF
> files are in 32-bit precision. So apparently it is not necessary for the
> files to be in 64-bit precision to close budgets.
>
> Release 3 has geothermal flux. See Section 4.2.1 (page 10) of Chris
> Piecuch’s note
> <ftp://ecco.jpl.nasa.gov/Version4/Release3/doc/evaluating_budgets_in_eccov4r3.pdf> on
> budget calculation.
>
> Regards,
> Ou Wang
>
> From: <ecco-support-bounces at mit.edu> on behalf of Martha Buckley <
> marthabuckley at gmail.com>
> Reply-To: "ECCO support list, wider membership" <ecco-support at mit.edu>
> Date: Tuesday, October 31, 2017 at 7:35 AM
> To: "ECCO support list, wider membership" <ecco-support at mit.edu>
> Subject: [ecco-support] Budgets in ECv4, release 3
>
> Hi all,
> I am trying to close mass, heat, and salt budgets in ECCo v4 release 3.  I
> closed the budgets to machine precision ((tendancy - ocean convergences -
> surfaceFluxes)/tendancy  ~ 1e-7)) for an older iteration.  However, I
> unable to close the budgets to machine precision for the files on the
> website for release3 (the budgets visually look fine though).
>
> I'm trying to figure out if I have a bug in my code or if something
> changed with the new output.  A few questions:
> (1) Are the files 64-bit precision, which is needed to close budgets? This
> seems to be the case since ncdump says the files are "double".
> (2) Did anything change between "r2/iter3" and release3 that might explain
> this?
> (3) Has anyone closed budgets for release3 using the nctiles output
> available on the ECCO website?
> thanks,
> Martha
>
>
>
>
> --
> Martha W. Buckley
> marthabuckley at gmail.com
> http://sites.google.com/site/marthabuckley/home
>
> _______________________________________________
> ecco-support mailing list
> ecco-support at mit.edu
> http://mailman.mit.edu/mailman/listinfo/ecco-support
>
>


-- 
Martha W. Buckley
marthabuckley at gmail.com
http://sites.google.com/site/marthabuckley/home
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/pipermail/ecco-support/attachments/20171102/6df90fc3/attachment.html