Earthdata Forum

Hello All,

I recently put in a request for about 8 years worth of daily MODIS chlorophyll-a data (~2900 days/files) extracted from the area around the Galapagos Islands (~4 degrees of latitude by ~4 degrees of longitude). On the data ordering screen, it had all 2900 files listed as being staged for download. When I eventually got the link and downloaded, I found that there were only about 2400 days/files included in the download.

Is this an issue with the data download process or is it a lack of available data? Since sensing chlorophyll concentration is dependent on a lack of cloud cover, does the bulk download automatically filter out any days that are entirely composed of "No Data" (i.e. completely covered by clouds, and thus have no chlorophyll concentration data throughout that spatial extent)?

I took a look at the temporal coverage of missing data days, and it appears that the distribution of missing days is non-random, with September/October being consistently higher in missing data from year to year. This makes me think that it may be a result of auto-filtering, since the cloud cover could be seasonally high here, but I wanted to check and see if anyone else has run into this problem.

If it's not an auto-filtering issue, does anyone know how to download ~500 files that are not temporally similar? I retrieved the file names for all the missing files, but they're so spread out throughout the 8 years that I can't batch order them by simply selecting a range of dates. If there is a way to iteratively order files based on a list of file names, that would be ideal.

Thank you so much for your help!

Hi.

If you can provide one or more of the order IDs in question, i can look more into it.

For some basic information, if this was an order for L2 data, that is done using the operational L2 granules and a data-extraction utility that simply extracts a region from the full L2 file.  For granules on the edge of a region, the browser might indicate that the granule overlays the region, but the extraction utility might not be able to get any or enough of the region.  These files get marked with the status NO_EXTRACT.  I don't believe the L2 extraction utility cares about clouds.  If it can extract the region of interest from the full granule, that is what gets staged.

If this was a L3 order for map/image products, the system starts with the global binned product and then runs a mapping utility on the region of interest.  In this case, if there are not enough filled bins in the region, the mapper will bail out and again the file will get marked with the NO_EXTRACT status.

Regards,
john

Hi John,

Thanks for your response, here's my order ID: f020e0a0d4c7c957

It was indeed an L3 order for the mapped product. So, correct me if I'm wrong, but what you're saying is that the missing data is likely due to the satellite sweep not fully including the region of interest? Thanks for your assistance!

Best,
Calvin

Calvin,

For L3 orders, the global binned product would be given to the mapper program with the regional bounding-box coordinates you provided.  If no unflagged pixels are in that region, the mapper bails with NO_EXTRACT.  That could be caused by the gaps between orbits, especially in a daily file, or any of the flagged conditions that we mask out at L3: ATMFAIL,LAND,HILT,HISATZEN,STRAYLIGHT,CLDICE,COCCOLITH,LOWLW,CHLWARN,CHLFAIL,NAVWARN,MAXAERITER,ATMWARN,HISOLZEN,NAVFAIL,FILTER,HIGLINT

john