*Is your feature request related to a problem? Please describe.
It is possible to run CESM in a multi-instance configuration, which generate ensemble members of a particular simulation. Currently CUPiD only recognizes history file output its timeseries tool from a single instance of the model. There is functionality to process timeseries for multiple individual cases in one go, but not to handle the multi-instance history files (tagged "model_000X.hX', for example). The ability to process timeseries files for multi-instance cases would be very useful for data assimilation applications, which tend to produce multi-instance history files at high temporal frequencies and high overhead, occupying lots of storage space.
Describe the solution you'd like
I'd like to be able to process all of the history files associated with each instance of the model case into their own timeseries files.
Describe alternatives you've considered
I've hacked a legacy version of the CESM postprocessing timeseries tool pre-CUPiD to identify the individual instance files via the associated xml file. This option is no longer officially supported.
I've processed the timeseries files from history files manually using nco; this option is time consuming and not optimized.
Additional context
This project is proposed as part of the SEWG winter meeting (2026).
*Is your feature request related to a problem? Please describe.
It is possible to run CESM in a multi-instance configuration, which generate ensemble members of a particular simulation. Currently CUPiD only recognizes history file output its timeseries tool from a single instance of the model. There is functionality to process timeseries for multiple individual cases in one go, but not to handle the multi-instance history files (tagged "model_000X.hX', for example). The ability to process timeseries files for multi-instance cases would be very useful for data assimilation applications, which tend to produce multi-instance history files at high temporal frequencies and high overhead, occupying lots of storage space.
Describe the solution you'd like
I'd like to be able to process all of the history files associated with each instance of the model case into their own timeseries files.
Describe alternatives you've considered
I've hacked a legacy version of the CESM postprocessing timeseries tool pre-CUPiD to identify the individual instance files via the associated xml file. This option is no longer officially supported.
I've processed the timeseries files from history files manually using nco; this option is time consuming and not optimized.
Additional context
This project is proposed as part of the SEWG winter meeting (2026).