meris yellow substance scaling

By hesselmans Date 2010-09-13 09:18

I downloaded meris RR level 2 products from the ESA website (merci), format N1.
The product contains chlorophyll data (called algal_1 and algal_2), suspended sediment (total_susp)
and yellow substance (yellow_subs).
I displayed the above 4 constituents using SeaDAS.
Values for algal_1, algal_2 and total_susp were in the expected range.
However, a rather large number of pixels had negative yellow substance values.

ESA documentation indicates that the logarithm of the 4 constituents is stored.
In the SeaDAS code (src/c4idl/libmerisread/epr_dddb.c) I could find reference to this
logarithm transform for algal_1, algal_2 and total_susp, but not for yellow_subs.
Can someone confirm that one should apply 10^yellow_subs to get the proper value for yellow_subs in SeaDAS
or am I mistaking?

Regards, Gerard

By @sean Date 2010-09-13 13:26

Gerard,

The example N1 file I have shows that yellow_substance is NOT store as the log10 of the value.
However, this file predates the latest MERIS product spec document I found, which as you point
out, indicates that yellow_substance IS stored as the log10 of the value. Seems a change was
made. So, I suppose the answer to your question is sometimes...

Sean

By hesselmans Date 2010-09-15 09:03

Sean

It seems I have to perform the 10^ operation myself whenever needed.
The MERIS L2 data include product confidence descriptors.
For yellow substance it is PCD_16 (see epr_dddb.c).
Is PCD_16 somehow included in the l2_flags provided by SeaDAS?

Regards, Gerard

By @sean Date 2010-09-15 12:45

Gerard,

SeaDAS will display the MERIS L2 flags via:
Utilities->Data Visualization->L2 Flags Display

And, yes, the PCD_16 flag is available.

Sean

By hesselmans Date 2010-09-15 14:01 Edited 2010-09-15 14:03

Sean

Glad that PCD flags are available.
On the Level 2 Flags Display I can switch between
"l2gen L2 Flags", "MODIS SST Flags" and "MERIS L2 Flags".
The first item for "l2gen L2 Flags" is ATMFAIL,
whereas the first item for "MERIS L2 Flags" is CLOUD.
Does this mean that whenever I load the variable l2_flags of MERIS L2 data into memory, using the IDL routine get_meris_data,
the first bit is the CLOUD flag.
whereas in case of SeaDAS or MODIS data, using the appropriate reading routine, the first bit of l2_flags corresponds to the ATMFAIL flag?

Regards, Gerard

By @sean Date 2010-09-15 15:52

Yes. The load routine does nothing to the l2 flag array when it is read in,
the L2 flags display function simply interprets a MERIS l2 flags array
differently from a SeaWiFS l2 flags array - because they're different :-)

Sean

By unhson Date 2011-12-08 04:14

Hello,

I applied MERIS L2 flags (in Level 2 Flags Display) on a MERIS L2 images in GUI (e.g. cloud, land, etc). However, is there the function in command line? I searched but couldn't find it. Thanks.

Ali

By @sean Date 2011-12-08 16:06

Yes. The load_graph command is what you'll need. See:

http://seadasdev.domain.pub/seadas/doc/seadisp/load_graph_cmd.html

For a simple example of using l2flags via command line, see:

http://oceancolor.gsfc.nasa.gov/forum/oceancolor/topic_show.pl?pid=4889#pid4889

Sean

By unhson Date 2011-12-08 20:18

Thanks for your answer.
Actually, I already tried 'load_graph' command.
But it doesn't work well unlike GUI.

For example, after loading ['chlor_a','l2_flags] from MERIS L-2 files..
I successfully displayed land (or cloud) over the chl-a image. But when I tried it in command..

>display, band_no=1
>load_graph, 2, fbuf=1, ftype='BAND', bit=[1], color=[1]

it doesn't work. (I'm using seadas 6.1 on mac os-x).
am I wrong?

Ali

By @sean Date 2011-12-09 01:03

bit 1 is ATMFAIL. bit 10 is CLDICE.

Sean

By unhson Date 2011-12-09 01:14

yes, bit 1 is ATMFAIL for 'l2gen L2 flags'.
but for 'MERIS L2 Flags', it looks different. is that right?
thanks.

Ali

By @sean Date 2011-12-09 14:38 Edited 2011-12-09 14:40

Correct. I was lost in a fog, saw l2_flags and assumed you were using OBPG processed L2 files.
The MERIS l2_flags are indeed different. After fiddling with the GUI and load_graph a bit, I'm
not convinced the GUI is entirely correct.

I was able to confirm that load_graph can produce flag overlays for the MERIS products, however.
I don't have time to dig too deeply into it (the MERIS flag scheme is a little more complicated than
the simple 32 bitwise values used by OBPG), but here are a few key flags and their corresponding
bit value (replace the bit value with BIT in the following: load_graph, 2, fbuf=1, ftype='BAND', bit=[BIT], color=[3])

Land bit 24
Ocean bit 22
Coastline bit 14
High Glint bit 3
Suspect bit 21 (I think)
Cloud bit 1

Sean