It would be convenient for the assimilation code to automatically set the assimilation cycling time (start and stop dates) based upon the the observation times of the observation sequence files. This is applicable for non-regular timing of observations (e.g. leaf area) in which observations are bi-monthly or every 8 days. There could already be some code/approach out there to address this, so looking for more feedback.
Is your feature request related to a problem?
No, not really.
Describe your preferred solution
Modify the assimilate.csh script to query the date for each of the observation sequence files, and use that to set the STOP_DATE variable using the ./xmlchange within the caseroot folder. Give the user the option to use this 'automatic' assimilation cycling approach which queries the observation sequence file date, or default back to the 'manual' approach.
Describe any alternatives you have considered
If the observations are regular (i.e. daily, monthly) then it is easy to set the STOP_N and STOP_OPTION variables to produce the appropriate assimilation cycling time.
For non-regular timing of observations, it is my experience that users general make ad-hoc modifications to the assimilate.csh script specific to their application. This just produces a more general, formal approach.
Use case
It would be convenient for the assimilation code to automatically set the assimilation cycling time (start and stop dates) based upon the the observation times of the observation sequence files. This is applicable for non-regular timing of observations (e.g. leaf area) in which observations are bi-monthly or every 8 days. There could already be some code/approach out there to address this, so looking for more feedback.
Is your feature request related to a problem?
No, not really.
Describe your preferred solution
Modify the assimilate.csh script to query the date for each of the observation sequence files, and use that to set the STOP_DATE variable using the ./xmlchange within the caseroot folder. Give the user the option to use this 'automatic' assimilation cycling approach which queries the observation sequence file date, or default back to the 'manual' approach.
Describe any alternatives you have considered
If the observations are regular (i.e. daily, monthly) then it is easy to set the STOP_N and STOP_OPTION variables to produce the appropriate assimilation cycling time.
For non-regular timing of observations, it is my experience that users general make ad-hoc modifications to the assimilate.csh script specific to their application. This just produces a more general, formal approach.