Mismatch between input geographic limits and output of l2gen for MODISA using V2022.0

Use this Forum to find information on, or ask a question about, NASA Earth Science data.
zhigang
Posts: 72
Joined: Tue Nov 10, 2020 8:03 pm America/New_York

Mismatch between input geographic limits and output of l2gen for MODISA using V2022.0

by zhigang » Wed Apr 13, 2022 11:10 am America/New_York

Hi,

Recently, I am confused about MODISA processing using V2022.0 OC processors. Here is an example:

I set the south=30.9, north=31.09, west=82.1, east=82.3 in l2gen to process A2002274073500.L1B_LAC. But output level-2 product is in lat=<20.28, 20.29>, lon=<91.96, 91.97>. The input boundary was mismatched with the output product. Such an issue is frequently for my processing task.

A sampled output file can be found using below link:
https://drive.google.com/file/d/1BE_D0MqYvtva9l9Yx5YEfCeWhFoMlLqU/view?usp=sharing

Could someone give me some suggestions?

Best Regards,
Zhigang
Attachments
Screenshot 2022-04-13 230849.png
Screenshot 2022-04-13 230849.png (272.67 KiB) Not viewed yet
Screenshot 2022-04-13 230820.png
Screenshot 2022-04-13 230820.png (312.96 KiB) Not viewed yet

Tags:

zhigang
Posts: 72
Joined: Tue Nov 10, 2020 8:03 pm America/New_York

Re: Mismatch between input geographic limits and output of l2gen for MODISA using V2022.0

by zhigang » Wed Apr 13, 2022 10:50 pm America/New_York

I reverted the V2021.1, it works now. I am not sure whether I wrongly set V2022.0 or this version has potential bugs.

bijaylaxmi
Posts: 8
Joined: Fri Apr 01, 2022 4:54 pm America/New_York

Re: Mismatch between input geographic limits and output of l2gen for MODISA using V2022.0

by bijaylaxmi » Mon Apr 25, 2022 12:22 pm America/New_York

Hi,
I am not able to get the correct output using V2022.0 and V2021.1 as well.
I set south=26, north=31, east=-87.5, and west=-92.5. The output has south=20, north=25.6, east= -85.4 and west=-92.2.

OB SeaDAS - knowles
Subject Matter Expert
Subject Matter Expert
Posts: 245
Joined: Mon Apr 07, 2008 4:40 pm America/New_York

Re: Mismatch between input geographic limits and output of l2gen for MODISA using V2022.0

by OB SeaDAS - knowles » Tue Apr 26, 2022 1:01 pm America/New_York

Please post the exact file name you are using. Also if possible post your par file ... or cut and paste the subsetting options which you used.

For example I just ran l2gen on file: AQUA_MODIS.20101010T180500.L1B.hdf

# SUBSETTING OPTIONS
north=40
south=35
east=-70
west=-75

The resultant file had the following Global Attributes:
subset.png
subset.png (28.79 KiB) Not viewed yet
Note: due to the slighted tilted nature of the MODIS scene swath the (min/max) geographic boundaries of the resultant file will be a slightly larger geographic box than the requested l2gen subset boundaries. This is to ensure that all pixels of the requested subset are contained in the resultant scene.

bijaylaxmi
Posts: 8
Joined: Fri Apr 01, 2022 4:54 pm America/New_York

Re: Mismatch between input geographic limits and output of l2gen for MODISA using V2022.0

by bijaylaxmi » Tue Apr 26, 2022 6:45 pm America/New_York

Here is one of the files I failed to obtain correct output. A2005335192500.L1B_LAC.

The parameters are as followed:

ifile=/Users/gulfcarbon/Documents/A2005335192500.L1B_LAC
geofile=/Users/gulfcarbon/Documents/AQUA_MODIS.20051201T192500.GEO.hdf
ofile=/Users/gulfcarbon/Documents/AQUA_MODIS.20051201T192500.L2.OC.nc

# SUITE
suite=OC

# PRODUCTS
l2prod=Kd_490 Rrs_1240 Rrs_1640 Rrs_2130 Rrs_748 Rrs_859 Rrs_869 Rrs_vvv aer_model_max aer_model_min aer_model_ratio angstrom aot_869 chlor_a humidity ipar nLw nLw_1240 nLw_1640 nLw_2130 nLw_748 nLw_859 nLw_869 nLw_vvv nflh ozone par pic poc rhos rhos_1240 rhos_1640 rhos_2130 rhos_748 rhos_859 rhos_869 rhos_vvv chl_ocx

# PROCESSING OPTIONS
aer_opt=-3
maskhilt=0

# SUBSETTING OPTIONS
north=31
south=26
east=-87.5
west=-92.5

# THRESHOLDS
cloud_thresh=0.018
cloud_wave=2130

# ANCILLARY INPUTS Default = climatology (select 'Get Ancillary' to download ancillary files)
icefile=/Users/gulfcarbon/ocssw/var/anc/2005/335/N2005335_SST_OIV2AVAM_24h.nc
met1=/Users/gulfcarbon/ocssw/var/anc/2005/335/N200533518_MET_NCEPR2_6h.hdf
met2=/Users/gulfcarbon/ocssw/var/anc/2005/336/N200533600_MET_NCEPR2_6h.hdf
met3=/Users/gulfcarbon/ocssw/var/anc/2005/336/N200533600_MET_NCEPR2_6h.hdf
ozone1=/Users/gulfcarbon/ocssw/var/anc/2005/335/N200533500_O3_AURAOMI_24h.hdf
ozone2=/Users/gulfcarbon/ocssw/var/anc/2005/336/N200533600_O3_AURAOMI_24h.hdf
ozone3=/Users/gulfcarbon/ocssw/var/anc/2005/336/N200533600_O3_AURAOMI_24h.hdf
sstfile=/Users/gulfcarbon/ocssw/var/anc/2005/335/N2005335_SST_OIV2AVAM_24h.nc

# MISCELLANEOUS
resolution=250

The resultant file has the following global attributes:
Attachments
Screen Shot 2022-04-26 at 5.40.01 PM.png
Screen Shot 2022-04-26 at 5.40.01 PM.png (32.97 KiB) Not viewed yet

zhigang
Posts: 72
Joined: Tue Nov 10, 2020 8:03 pm America/New_York

Re: Mismatch between input geographic limits and output of l2gen for MODISA using V2022.0

by zhigang » Tue Apr 26, 2022 9:37 pm America/New_York

Hi Daniel,

Did you examine the subset options in V2022.1? I have a l2 file generated in V2022.1:

https://drive.google.com/file/d/1BE_D0MqYvtva9l9Yx5YEfCeWhFoMlLqU/view

which includes all inputs used in the processing.

Same configuration worked in V2021.1 but failed in V2022.1.

Regards,
Zhigang

OB SeaDAS - knowles
Subject Matter Expert
Subject Matter Expert
Posts: 245
Joined: Mon Apr 07, 2008 4:40 pm America/New_York

Re: Mismatch between input geographic limits and output of l2gen for MODISA using V2022.0

by OB SeaDAS - knowles » Thu Apr 28, 2022 8:43 am America/New_York

Hi Zhigang,

There is not yet an OCSSW tag V2022.1. Here are the OCSSW operational tags (V2022.0 being the latest tag) as of today (Apr 28, 2022):
V2020.0
V2020.1
V2020.2
V2021.0
V2021.1
V2021.2
V2022.0

OB SeaDAS - knowles
Subject Matter Expert
Subject Matter Expert
Posts: 245
Joined: Mon Apr 07, 2008 4:40 pm America/New_York

Re: Mismatch between input geographic limits and output of l2gen for MODISA using V2022.0

by OB SeaDAS - knowles » Thu Apr 28, 2022 9:17 am America/New_York

Hi bijaylaxmi,

l2gen can subset the standard resolutions, however for the case of MODIS with the extra resolution options, the 250m and 500m resolution options does not currently subset correctly to the requested geographic boundaries. For MODIS at 1km resolution, the l2gen subset option works fine.

SeaDAS-OCSSW does have tools specifically designed for subsetting which will do what you are trying to do. You can first extract the L1A file using the OCSSW program "l1aextract_modis" or in the GUI (Menu > SeaDAS-OCSSW > extractor). Then process the L1A extracted file to a L1B extracted file. Then when you run l2gen on the extracted L1B file, the geographic area will be correct and you can run l2gen at any of the available resolutions.

bijaylaxmi
Posts: 8
Joined: Fri Apr 01, 2022 4:54 pm America/New_York

Re: Mismatch between input geographic limits and output of l2gen for MODISA using V2022.0

by bijaylaxmi » Thu Apr 28, 2022 3:07 pm America/New_York

Hi Knowles,
I was using the extractor in ocssw earlier. And I have got the correct outputs.
I am having the issue since I have updated to the V2022.0. Now OCSSW also doesn't work with the older versions.
Here I have shared the output coordinates for the same MODIS file exacted using ocssw extractor.
Attachments
Screen Shot 2022-04-28 at 1.59.15 PM.png
Screen Shot 2022-04-28 at 1.59.15 PM.png (33.46 KiB) Not viewed yet

OB SeaDAS - xuanyang02
Subject Matter Expert
Subject Matter Expert
Posts: 323
Joined: Tue Feb 09, 2021 5:42 pm America/New_York
Answers: 1

Re: Mismatch between input geographic limits and output of l2gen for MODISA using V2022.0

by OB SeaDAS - xuanyang02 » Mon May 02, 2022 3:11 pm America/New_York

Hello bijaylaxmi,

I tried your file A2005335192500.L1A_LAC with V2022.0. Here is what I did

1) modis_GEO on A2005335192500.L1A_LAC -> AQUA_MODIS.20051201T192500.GEO.hdf

2) extractor on A2005335192500.L1A_LAC with (SWlon=-92.5, SWlat=26, NElon=-87.5, NElat=31) -> AQUA_MODIS.20051201T192500.L1A.sub.nc

3) modis_GEO on AQUA_MODIS.20051201T192500.L1A.sub.nc -> AQUA_MODIS.20051201T192500.GEO.sub.hdf

4) modis_L1B on AQUA_MODIS.20051201T192500.L1A.sub.nc -> AQUA_MODIS.L1B_LAC, AQUA_MODIS.L1B_QKM, AQUA_MODIS.L1B_HKM

5) V2022.0 has some issues with modis_L1B output filenames and guessing HKM/QKM files, so you need to rename the three L1B files
AQUA_MODIS.L1B_LAC -> A2005335192500.L1B_LAC.sub
AQUA_MODIS.L1B_QKM -> A2005335192500.L1B_QKM.sub
AQUA_MODIS.L1B_HKM -> A2005335192500.L1B_HKM.sub

6) l2gen on A2005335192500.L1B_LAC.sub + AQUA_MODIS.20051201T192500.GEO.sub.hdf with resolution=250

And here is the result:
Screen Shot 2022-05-02 at 3.10.31 PM.png
Screen Shot 2022-05-02 at 3.10.31 PM.png (32.5 KiB) Not viewed yet

Post Reply