Problem generating L3 SSTs from NASA MYD00F data.

[smasse]

We are currently processing Aqua/Terra data collected with our antenna
up to level SST and Ocean Color using SeaDAS 7.5. We are trying to
use NASA MOD00F and MYD00F data if and when the antenna is
disabled for more than a day. We downloaded four (4) MYD00F granules
for one day that cover our area of interest and proceesed them to level 3.
The Ocean Color L3 products were generated but the SST products were not.
Note that l2bin exits with a code 110 when processing L2 SST products.
I have attached the output of the l2bin and l3bin commands.
Thanks,

Serge
attachment 1

[OB.DAACx - SeanBailey]

Serge,
You log file shows you set flaguse=LAND,~HISOLZEN
That would be what you would want for binning nighttime SST.  The "~" inverts the flag, so ~HISOLZEN will exclude all data that is NOT flagged as HISOLZEN.  Not something you want to do on the bright side of the earth.

Sean

[smasse]

Hi Sean,
I will check these flags. However, the same flags are successfully used  to generate binned data
for direct broadcast granules. See the attached log file 20181203.log.
Thanks,
Serge
attachment 1

[OB.DAACx - SeanBailey]

Without seeing the source files I can't comment on why it seems to work for your DB data.  The HISOLZEN flag is used to split day and night data for global binning purposes.  You can probably get away with not including it in the flaguse parameter. 

Sean

[smasse]

Is there a document on the use of these flags?
If that would, I can upload the L2_SST from DB data for 20181203 and the L2_SST from NASA data for 20181118
that I used with l2bin. However, it is a substantial amount of data.
Serge

[OB.DAACx - SeanBailey]

The bin product description includes general information about the application of the L2 flags.  The "~" used to invert the flag application in the binner...well that's documented in the code :wink:

An example ncdump -h output for both the DB sourced and NASA L0 sourced L2 files should be sufficient (and small enough to attach to the forum - though please NOT inline)

Sean

[smasse]

See attachments
Serge
attachment 1

attachment 2

attachment 3

attachment 4

attachment 5

attachment 6

[OB.DAACx - SeanBailey]

Serge,

These are a range of day, night and mixed (i.e. both day and night) granules.
Those with the  attribute "Day or Night"  set to "Night" or "Mixed" will work with your l2bin process as you've defined it (with flaguse=LAND,~HISOLZEN), but those with "Day" will not:

$ grep "Day or Night" *dmp
A20181118153000.gran.L2_SST.hdf.dmp:            :Day or Night = "Day" ;
A20181118153500.gran.L2_SST.hdf.dmp:            :Day or Night = "Day" ;
A20181118171000.gran.L2_SST.hdf.dmp:            :Day or Night = "Day" ;
A20181118171500.gran.L2_SST.hdf.dmp:            :Day or Night = "Mixed" ;
A20181203101624.gran.L2_SST.hdf.dmp:            :Day or Night = "Night" ;
A20181203211747.gran.L2_SST.hdf.dmp:            :Day or Night = "Day" ;

Sean

[smasse]

I tried l2bin with 'flaguse=LAND,~HISOLZEN`, 'flaguse=LAND', and without 'flaguse' on the command line.
In all 3 instances, l2bin exits with a code of 110.
Serge

[OB.DAACx - SeanBailey]

I assume we're still talking SST?  Have you confirmed that there are data to bin?  The SST binning uses the quality flag to determine what to bin.  If the quality flag is 2 or less, the data will be binned.  If it is 3 (or 4...but rarely will it be 4) then the data are excluded from binning.  The l2gen code will set the quality flag to 3 for a number of reasons, one of which is the use of climatology for the reference SST product.  Make sure your l2gen process includes NRT ancillary data (the getanc.py script provided with SeaDAS will identify and retrieve these for you)

Sean