[MITgcm-support] Merging Floaters' Trajectory Output

[Bertrand Delorme]

Hi everyone,

I am using the FLT package to track the 3D advection of some floaters in my simulation. I am only interested in the floaters’ trajectory. I run my simulation in parallel with MPI using 96 cpus. To get all outputs into a single file, I was using one of these two methods in the past:

  *   MNC package + gluemncbig to get them in a single netcdf file
  *   useSingleCpuIO=.TRUE. in data to get them in a single “.data” file
However, it seems that none of these methods affects the outputs from the FLT package. I have 96 files that looks like that: float_trajectories.001.XXX.data where XXX goes from 001 to 096. How can I merge these files together? I have tried using the joinmds script as:
“joinmds float_trajectories 001“ in the run directory but got a “No dataset found” error message.

As a side question, are we expecting the model to takes longer to run when using the MNC package rather than the traditional MDSIO?

Thank you,

Bertrand Delorme
PhD Candidate Stanford University



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.mitgcm.org/pipermail/mitgcm-support/attachments/20201126/5dd19e50/attachment.html>