As noted in #125 and elsewhere it is possible for tempest extremes to accept input files with different time ranges and run to produce output.
It will error if something requests unavailable data, but will write to a file that is subsequently useable.
This could lead to unexpected erors or behaviour in a number of places so we should ideally build in a check to the Python code to ensure that time ranges are consistent across all files rather than relying on the user to provide sensible data.
This particularly pertinent as some datasets are no provided in consistent files for all variables (e.g. CEDA).
As noted in #125 and elsewhere it is possible for tempest extremes to accept input files with different time ranges and run to produce output.
It will error if something requests unavailable data, but will write to a file that is subsequently useable.
This could lead to unexpected erors or behaviour in a number of places so we should ideally build in a check to the Python code to ensure that time ranges are consistent across all files rather than relying on the user to provide sensible data.
This particularly pertinent as some datasets are no provided in consistent files for all variables (e.g. CEDA).