Currently, COBALT-related input parameters are minimally checked via the generic namelist-parsing functions, but there aren't many (any?) checks to make sure values make sense. Should we add a routine that performs more nuanced validation?
This may include simple checks like enforcing non-negative restrictions on rates, making sure that parameters that are used in tandem vary in a consistent manner (as the case2 example discussed this morning... if we opt to make the default values 0 rather than to add a flag, then we would want to check that users remembered to change all 3 at once), possibly providing realistic ranges for certain parameters (or ratios between parameters... many of the small/medium/large functional group differences are more to do with their relative rates than true functional differences).
Currently, COBALT-related input parameters are minimally checked via the generic namelist-parsing functions, but there aren't many (any?) checks to make sure values make sense. Should we add a routine that performs more nuanced validation?
This may include simple checks like enforcing non-negative restrictions on rates, making sure that parameters that are used in tandem vary in a consistent manner (as the case2 example discussed this morning... if we opt to make the default values 0 rather than to add a flag, then we would want to check that users remembered to change all 3 at once), possibly providing realistic ranges for certain parameters (or ratios between parameters... many of the small/medium/large functional group differences are more to do with their relative rates than true functional differences).