At this stage, you can go straight 
to final modeling
 
 
Something else you can do
if you have some spare time:

Display two BPLs
Substitution : play a bit!
So far, we have identified two distinct cases:
  • BPL pixels under Mask_4 are written in textfile tarawa-subset_m4.cal
  • BPL pixels under Mask_3 are written in textfile tarawa-subset_m0.cal

Let'sdisplay the two sets of BPL pixels in the same calibration diagram
  • disable the two -CP... arguments
    • @CP/196.96/157.85/058.64_0.25m
    • @CP/185.04/130.29/029.71_0.50m
  • disable Z in -Calibrate/v/BdSNpzg/BDh_12_13
dir *.cal
  • remove tarawa-subset_m0.cal
  • remove tarawa-subset_m4.cal
  • you need tarawa-subset_m0.cal
    • for this run -Extract/v/mBPL3/mask_3                       right NOW
    • Mask=0; mBPL=3: this sweeps all marine pixels masked at 3 or 4 (and up to 20)
    • writes tarawa-subset_m0_255.gli 
    • writes tarawa-subset_m0.cal
    • the current calibration diagram before substition is shown below left
  • you need tarawa-subset_m4.cal
    • for this run  -Extract/v/mBPL3/Mask_4                      right NOW
    • Mask=4; this sweeps only marine  pixels under mask 4
    • writes tarawa-subset_m4_255.gli 
    • writes tarawa-subset_m4.cal
dir *.cal
you have two calibration data text files:
  • tarawa-subset_m0.cal mBPL=3 for the whole lagoon
  • tarawa-subset_m4.cal mBPL=4 under Mask=4
  • gedit  tarawa-subset_m4.cal  
    • select/copy all BPL pixels M...
      • M_004  1  2   169    17 137.6  88.0   8.5
      • ....
      • M_004  2  3    87   500  32.0  14.0   6.0
      • that's 375 lines
    • paste into a new text document
    • find and replace all M by m: this shall make them display as DPL pixels
      • M_004  1  2   169    17 137.6  88.0   8.5    (a BrightestPixel)
      • ...
      • M_004  2  3    87   500  32.0  14.0   6.0
        • become
      • m_004  1  2   169    17 137.6  88.0   8.5    (a DarkestPixel)
      • ...
      • m_004  2  3    87   500  32.0  14.0   6.0
    • ==>select/copy all records in this new text document
  • gedit  tarawa-subset_m0.cal 
    • select all m_...   records in tarawa-subset_m0.cal
    • paste
      • this substitution overwrites all existing m... records
    • save  tarawa-subset_m0.cal
    • close the new text document
  • Enable both BPL and DPL pixels in -Calibrate/v/BDSNpzg/BDh_12_23
    • crosses  shall denote these "DPL" pixels
    • squares shall denote the    "BPL" pixels
  • Disable -Extract... in the command line
    • -extract/v/mBPL3/mask_4 
    • -Calibrate/v/BdSNpZg/BDh_12_13 
 
  • Run the Calibration
  • The new calibration diagram after substition shown below right

The result: two BPLs

mBPL>=3      before substitution    ====>
 

====>mBPL>=3      after substitution
 


Jerlov's water type OI :
the clearest in town!
<================
  • The waters which get through the reef rim from the outer reef system 
  • appear to be of the clearest type of marine waters, according to Jerlov's classification :
  • Oceanic type I, with Kgreen/Kred~=0.175
 
  • These waters appear to mix very quickly with lagoon waters 
 
  • Lagoon watesr would be
  • Oceanic type III, with Kgreen/Kred~=0.26
<================
 

 
 
 
Something else you can do
Now, using the mask mSE,
you should be able to produce this kind of plot
see below



This bi-plot of Green band vs NIR band shows:
  • the glint signal over deep waters, in line with the cloud/shadow signal:
    • after all, this light  was produced by the sun, and reflected by water surface (glint)  or by water vapor (clouds)
    • when the glint signal is poorly correlated, one may include cloud/haze in order to produce a stronger glint regression
    • this also applies quite often to situations where an adverse atmospheric adjacency effect must/can be "deglinted".
  • the exponential decay over very shallow bright bottoms
  • a patch of cocconut groove: bright in NIR, dark in Green
  • scattered bareland pixels

Red vs NIR




 

 



Créer un site
Créer un site