tmnov tutorial : Calibration 4

return to Calibration 3      next to Calibration 5

Improve your understanding and strengthen your control
No deglint
Ghost pixels in calibration diagrams
 
 
 
Calibration 4: No Deglint
No Deglint   
-deglint/vRbaD/GlintM00.0       
-Extract/v/NIRmax200.0/mBPL2/mask_3           
-Calibrate/v/BdSNpZG/BDh_12_13_34       
  • You can disable the deglinting alltogether : GlintM=0
  • while the determination of -Lsw... is still obtained from using the glint regressions

Calibration 3 : GlintM=25  NIRmax=005.0
 
Calibration 4 : GlintM=0  NIRmax=200.0

Calibration 3 : GlintM=25  NIRmax=005.0
The thin grey line represents the glint regression line
Deglinting damages very shallow pixels
Calibration 4 : GlintM=0  NIRmax=200.0

Calibration 3 : GlintM=25  NIRmax=005.0
The thin grey line represents the glint regression line
 
Calibration 4 : GlintM=0  NIRmax=200.0
Cloud/haze contribue to the BPL calibration pixels
 


 
 
Calibration 4: Ghost on/off
Ghost was ON in the above
Now, turn it OFF
-Calibrate/v/BdSNpZg/BDh_12_13_34       
  • -Calibrate/v/BdSNpZG/BDh_12_13_34 turns ON the Ghost option
  • -Calibrate/v/BdSNpZg/BDh_12_13_34 turns OFF the Ghost option
  • The white square BPL pixels are the ghost pixels
  • BPL pixels for the Blue vs Red and the Green vs Red pairs are also plotted as white squares in the two-bands case Blue vs Green plot
    • and vice versa for the three-bands case
  • This is not an artifact caused by deglinting
  • Rather it is test of the BPL assumption : what we learn here is that, for this image
    • the shallow BPL pixels are greenish (LBgreen>LBblue and LBgreen>LBred), certanaily not neutral, over the 0-6 m depth range
    • when modeling by the pair Blue vs Green, we shall have to worry whether or not this applies over the whole 0-20 m shallow  depth range
 

Ghost is ON
Ghost is OFF
next to Calibration 5
 
 
 
 
 
 
 



Créer un site
Créer un site