L2gen - cant remove HILT flags MODIS

[paulajcu]

Hi
I am using L2gen to process MODIS aqua for retrieving high-turbidity flood plume data. I need to remove HILT which is saturated at all the highest turbidity pixels in the L2 products, the exact region I want to quantify with a math band algorithm. I have tried 'fixes' from this forum but nothing has worked so far. I'm using seadas version 9 GUI.
Can anyone help?
Cheers
Paula

[OB SeaDAS - knowles]

If your question regards running l2gen, then setting "maskhilt=1" (which is the default and really doesn't need to be set if your suite is "OC") will result in a level-2 file with HILT flagged pixels masked out.

If your question regards using a math expression based SeaDAS tool with an input level2 file (for instance MathBand) then the math expression to only include chlor_a band values where HILT is not flagged would be:
"if (! l2_flags.HILT) then chlor_a else NaN"

Danny

[paulajcu]

Hi Danny
Thanks for replying.
I want to run L2gen without masking HILT. I want to see the values of the pixels, not mask them. Turning off maskhilt is not achieving this. They are still coming out masked.
Cheers
Paula

[OB SeaDAS - knowles]

Do the pixels which you see as HILT flagged also contain other flags which are set to true (such as STRAYLIGHT, CLDICE, LAND)? Look at Pixel Info > Flags to verify this.

If you do see pixels which are only flagged HILT then send the exact MODIS filename and ideally a picture of the region where the flags are being set.

Also send the exact processing parameters which you are using to run l2gen so we can try to replicate what you are seeing.

Danny

[paulajcu]

These (attached) are the processing parameters I am using; the rest are default.
The turbid pixels only flag as HILT (and PRODFAIL).
Filename is A2009047045000.L1A_LAC which I have processed through to AQUA_MODIS.20090216T045000.L1B.sub.hdf to run l2gen.
Attached is true colour image of same region, same modis overpass.

[OB.DAAC - SeanBailey]

When the HILT flag is set for MODIS, one or more of the bands are saturated. It is often that the 869nm band is saturated, this band is used in the atmospheric correction. Disabling the masking of HILT will not change the fact that the data are saturated and a saturated 869nm band will cause the aerosol selection process of the atmospehric correction to fail. It may be possible to use alternate bands for the atmospheric correction (e.g. 859nm, or one of the SWIR), but that comes with its own issues (e.g. the vicarious calibration may need to be rederived and greatly increased uncertainty due to low SNR for these bands).

Sean