Task: Implement finding nearest time if a xarray Dataset has time dim and multiple time points. If there is time dim in xarray dataset but only 1 time, then matching to nearest time is not necessary (skip).
Here is a sample netcdf with 366 days for testing
https://github.com/fish-pace/point-collocation/tree/main/examples/fixtures/multi_time.nc
Here is a notebook that shows some sample points and finding nearest lat/lon/time
https://github.com/fish-pace/point-collocation/blob/main/examples/multi_time.ipynb
Background: In some cases, a granule will have multiple times. For example, a yearly granule with 12 months or 365 days. A lat/lon/time will match to that granule, but the time needs to be matched to nearest time.
Notes:
- time should always be 1D. but might not be uniform.
- although unusual, it is possible that 2D lat/lon would have multiple times.
- if there is only 1 time, then matching to nearest time is not necessary
- cf_xarray should be used to find the time dim, and a reasonable fallback used.
- report time dim with geolocation in open_dataset() and anywhere else that geolocation is reported. Do not report time dim if the granules do not have a time dim
Task: Implement finding nearest time if a xarray Dataset has time dim and multiple time points. If there is time dim in xarray dataset but only 1 time, then matching to nearest time is not necessary (skip).
Here is a sample netcdf with 366 days for testing
https://github.com/fish-pace/point-collocation/tree/main/examples/fixtures/multi_time.nc
Here is a notebook that shows some sample points and finding nearest lat/lon/time
https://github.com/fish-pace/point-collocation/blob/main/examples/multi_time.ipynb
Background: In some cases, a granule will have multiple times. For example, a yearly granule with 12 months or 365 days. A lat/lon/time will match to that granule, but the time needs to be matched to nearest time.
Notes: