Jane Meiforth
New Member
Posts:6  
| 03 Feb 2021 07:05 PM |
|
|
Dear Team at Harris Geospatial,
when I add the zenith and azimuth angles for an image from the off-nadir WorldView-2 sensor taken in New Zealand, the reflectance values in the coastal and blue bands decline to values under zero even in vegetated areas outside shadow and water bodies. I am already using parameters with low water vapor and aerosol according to the clear atmosphere in New Zealand (US standard atmosphere, maritime aerosol, 70km visibility). I wonder if I did something wrong with calculating the zenith and azimuth angles. Is this correct?
Azimuth: Wv02 satellite metadata: meanSatAz 353.8 degree >> FLAASH Azimuth 6.2 degree
off-Nadir: WV02 satellite metadata: meanoffNadirView angle 17.9 degree >> FLAASH Zenith Angle: 162.1 degree
The default nadir parameters (azimuth 0 degree, zenith 180 degree) lead to more realistic spectra. Or can it be that the off-Nadir correction in FLAASH does not match the conditions of the clear atmosphere in New Zealand?
With kind regards,
Jane
|
|
|
|
MariM
Veteran Member
Posts:2396 
| 04 Feb 2021 08:58 AM |
|
|
I think your azimuth should be 353.8-180 = 173.8 (as viewed from the ground) here.
Also, if you know your scene is very clear, you can push the visibility up to 100km which can sometimes lift those coastal and blue bands out of the negative.
What are the average radiance values in these areas? If it is still very low even in vegetated areas, you can still have poorly modeled pixels in those areas.
|
|
|
|
Jane Meiforth
New Member
Posts:6
| 05 Feb 2021 01:22 AM |
|
|
Hi Mari, the correction in the azimuth angle solved the problem with the negative values in the first bands! Many thanks , Jane
|
|
|
|
Jane Meiforth
New Member
Posts:6
| 18 Feb 2021 07:00 PM |
|
|
Hello again,
I just received this information from Maxar regarding the azimuth angle in the metadata of WorldView-2 satellites: " The azimuth of the satellite as seen by an observer sitting on the target (AOI) measured in a clockwise direction from north. Since our map-projected products are always Map-North up, "up" direction of the center line is always North."
So that means that the azimuth is already the angle seen from the observer and can be directly used in FLAASH when it is under 180 degree. And I assume that azimuth angles over 180 degree need to be calculated like : (180 - (360 - azimuth angle)) to achieve the values necessary for the FLAASH parameters.
Can you confirm this?
Many thanks,
Jane
|
|
|
|
MariM
Veteran Member
Posts:2396
| 19 Feb 2021 09:14 AM |
|
|
If that is their definition, it is also the definition in FLAASH so it could be used directly. Yes, azimuth is defined as +/- 180, so you will adjust for a 360 degree value to fall in this range.
|
|
|
|
Jane Meiforth
New Member
Posts:6
| 20 Feb 2021 03:01 PM |
|
|
Thank you. I just realized that azimuth values over 180 degree need to be calculated like this to integrate them in FLAASH: (0-(360-azimuth angle) Many thanks, Jane
|
|
|
|