[MITgcm-devel] exf_interp.F

[Patrick Heimbach]

Hi Matt,

sorry for late reply.
It would be great to have this available, if you have some time to look into this.

You may have already noticed that routine ext_interpolate_read may need repackaging
in case you would want to keep the I/O to the reduced/coarser res. grid active
(you probably want to).
You may also find that without "help" TAF will choke on some of the array indices
used in exf_interp.F

A different approach could be through coupling via the tapelev files, in three stages:
* high-res forward model produces tapelev_high_res output
* an interpolation routine computes tapelev_coarse_res from tapelev_high_res
* the coarse res. adjoint model, instead of producing its own forward tape levels,
  skips the forward simulation and reads the already existing interpolated tapelev files
* MPI would control the synchronization of the three parts
  (similar to the use of MPI in atmos/ocean coupling).

There are pros and cons for either approach.

-Patrick

On Jun 10, 2012, at 4:23 PM, Matthew Mazloff wrote:

> Hello,
> 
> It appears this won't be trivial as there is no vector IO for active variables.  An active_read_vector would have to be added.  Though if this were added it would likely be very useful.  Any thoughts on whether this is worth implementing?
> 
> Thanks
> Matt
> 
> 
> 
> 
> On Jun 10, 2012, at 10:06 AM, Matthew Mazloff wrote:
> 
>> Hello,
>> 
>> I would like to try using controls at a coarser resolution than the model grid.  Before I go ahead and try this I thought it wise to ask if there is any (taf) reason I won't be able to use exf_interp.F on control variables?
>> 
>> Thanks
>> Matt
>> 
>> _______________________________________________
>> MITgcm-devel mailing list
>> MITgcm-devel at mitgcm.org
>> http://mitgcm.org/mailman/listinfo/mitgcm-devel
> 
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel

---
Patrick Heimbach | heimbach at mit.edu | http://www.mit.edu/~heimbach
MIT | EAPS 54-1420 | 77 Massachusetts Ave | Cambridge MA 02139 USA
FON +1-617-253-5259 | FAX +1-617-253-4464 | SKYPE patrick.heimbach


-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 1588 bytes
Desc: not available
URL: <http://mitgcm.org/pipermail/mitgcm-devel/attachments/20120615/2478b660/attachment.p7s>