You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When loading a FCI AMV NetCDF product witht he fci_l2_nc reader the time attributes represent the following:
start_time: end_time of the first image/repeat cycle used for the wind tracking
end_time: end_time of the last image/repeat cycle used for the wind tracking
These are time stored in the FCI L2 AMV filenames.
However, it would make more sense to have the following:
start_time start_time of first image/repeat cycle
end_time end_time of last image/repeat cycle
wind_time end_time of second image/repeat cycle
Then the start_time and end_time would represent the full duration of the data used for the tracking which makes sense.
We also propose to add the attribute wind_time representing the end_time of the second image/repeat cycle.
This logic should then be harmonized among the EUMETAST AMV datasets read with the fci_l2_nc, fci_l2_bufr and seviri_l2_bufr readers.
We should also make sure to add test for the delta between the start_time and end_time to make sure that this is always consistent and breaks if there would be a corresponding change in the L2PF filename convention.
The text was updated successfully, but these errors were encountered:
When loading a FCI AMV NetCDF product witht he
fci_l2_nc
reader the time attributes represent the following:start_time
: end_time of the first image/repeat cycle used for the wind trackingend_time
: end_time of the last image/repeat cycle used for the wind trackingThese are time stored in the FCI L2 AMV filenames.
However, it would make more sense to have the following:
start_time
start_time of first image/repeat cycleend_time
end_time of last image/repeat cyclewind_time
end_time of second image/repeat cycleThen the
start_time
andend_time
would represent the full duration of the data used for the tracking which makes sense.We also propose to add the attribute
wind_time
representing theend_time
of the second image/repeat cycle.This logic should then be harmonized among the EUMETAST AMV datasets read with the
fci_l2_nc
,fci_l2_bufr
andseviri_l2_bufr
readers.We should also make sure to add test for the delta between the
start_time
andend_time
to make sure that this is always consistent and breaks if there would be a corresponding change in the L2PF filename convention.The text was updated successfully, but these errors were encountered: