[MITgcm-devel] Problems with diagnostics package and mnc

[Chris Hill]

Ed,

  I don't think that makes sense.
  A user writing a field needs to indicate the time associated with that 
field. MNC, at the lowest levels, shouldn't need to know what time the 
model is at, just what time the field has/wants.

Chris
Ed Hill wrote:

> If possible, I'd really like to have a variable within a common block
> (even if its an element-per-thread vector as Chris describes) that gives
> us a "real time" (eg. with actual time units--whatever they are!) that
> can be written as a coordinate variable by MNC.
> 
> Ed
>