Welcome to the Earthdata Forum! Here, the scientific user community and subject matter experts from NASA Distributed Active Archive Centers (DAACs), and other contributors, discuss research needs, data, and data applications.
by dem1 » Tue Aug 16, 2022 5:31 am America/New_York
We found that with R2022 the granule start time in the new filenaming convention is not consistent with R2018: with R2018 seconds were always set to 00, but with R2022 there is a mix of 00 and 01, is it normal?
We also found files with seconds different than 00 or 01 and minutes not multiple of 5 with R2022 where they were multiple of 5 and seconds to 00 with R2018, example (files found using the File Search API):
AQUA_MODIS.20050416T025349.L2.OC.nc <== this one seems to be A2005106025000.L2_LAC_OC.nc with R2018
AQUA_MODIS.20210511T120509.L2.OC.nc <== this one (at least) seems not available in the L2 browser
Yes, this is normal (and has been the case for SST L2s since 2019). We decided to stop lying to you
The start time in the L2 files is derived from the first valid scan line in the scene. Quite often that is indeed on the 5 minute boundary of the arbitrary chunking that is done to the MODIS data, however, the MODIS scan rate isn't an exact multiple of seconds, so sometimes the time bleeds into the next second. Other times, for one reason or another, the sensor doesn't have any valid data for a fair bit of time into the 5 minute chunk, so the L2 filename time will have a value that isn't so close to the 5 minute boundary as the L1 filename time would have you believe it was.
As for the one you noted appears missing from the L1/2 browser, it's there...but it's a significantly smaller file than the 5 minute variety and only the northern portion of the scene has valid data: