Worldview 3 Processing

[gtimmeney]

Hi Quentin,

With the recent updates, I have found some issues in processing WV03 imagery for DSF, adjacency correction, and using a limit polygon. It seems that the code assumes that there is also a SWIR band with this imagery, but this is often not the case in certain coastal areas, particularly near the Red Sea.

Some of the issues I have ran into are:

1. polygon as geojson or limit coords, even when in the same UTM projection as the WV03 satellite image are reprojected into WGS and cause acolite to fail.

2. DSF is not configured for fixed, tiled, segmented, or resolved (line 745 in acolite_l2r)

3. inputfile_swir=None seems to not be recognized, and the code still searches for a swir band, causing the atmospheric correction to fail.

Any input or support would be greatly appreciated.


Here is an example of my settings file, if that might assist.

inputfile="path2"_WorldView3_2019_09_17_08_42_10_L1R.nc
output="path2"_Rrs_v2
adjacency_correction=True
atmospheric_correction=True
inputfile_swir=None
limit=26.60770706391749, 36.232893828929406, 26.622379739179074, 36.24809567532168
aerosol_correction=dark_spectrum
dsf_aot_estimate=Fixed
dsf_residual_glint_correction=True
l2w_parameters=Rrs_*
l2w_export_geotiff=True
l2w_mask_negative_rhow=True


##EarthDataInformation
##here
##here

##data management
l2w_data_in_memory=True
l2w_export_geotiff=True
netcdf_projection=True
export_geotiff_match_file=None
l1r_delete_netcdf=True
l2t_delete_netcdf=True
l2r_delete_netcdf=True
l2w_delete_netcdf=True
rgb_rhot=False
rgb_rhos=False
map_title=False
map_l2w=False
output_rhorc=False
output_lt=False

[quinten]

Hi Graeme

For the commercial sensors I don't always have many scenes available for testing. If you could share via email your specific input bundle for testing/development purposes I can take a look at it specifically.

In this case there seem to be two potential issues in your settings file:
1) you specify a L1R ACOLITE file as input, which means the limit (or polygon) will be ignored
2) I think the "dsf_aot_estimate" setting should not be capitalized, so dsf_aot_estimate=Fixed must be dsf_aot_estimate=fixed

Are you using a recent clone of the GitHub code or one of the binary releases?

[Edit] I should note that the option adjacency_correction=True is currently not publicly available, and not configured for WorldView!

Quinten

[gtimmeney]

Hi Quentin

Good news is I ended up getting it to work!

I am using the most recent binary release.

No matter what options I set, even with the MUL file as the input, I cannot get limt= or polygon= to work on any worldview imagery.

The issue was the dsf_aot_estimate=Fixed.

I will follow up via email about sharing the image, I have had a number of issues with worldview recently and would like to share the image so you can use it for testing and troubleshooting.

Best,

Graeme

[quinten]

Hi Graeme

If you could share the data that would be the most efficient way for me to troubleshoot!

I think for WorldView subsetting is currently only supported if the input data are in a geographic projection. If they are not there is an option to attempt a reprojection with "worldview_reproject=True" but this setting is not well tested.

Quinten