LANDSAT 8 OLIP 
A time series of Landsat 8 OLIP images
at Nosy Be, Madagascar

work done in january 2021 at 15 m GSD
 
Using the Panchromatic band for water column correction
to derive water depth and spectral bottom signature:

Landsat 8 OLIP bandset used for this work

Purple=1Blue=2Green=3PAN=4Red=5NIR=6 and SWIR1=7

   peer-reviewed and published in 2017    

home
 
 
 
image acquired on August 13th 2019

The data

TCC TOA

TCC TOA deglinted, no smoothing

Glint regressions, based on SWIR band
  • quite low level of glint/haze
  • R2 values are acceptable
  • Deglinting is far from perfect
  
PAN BOA deglinted
  • the atmosphere is thick and homogeneous
  • water volume reflectance is near null everywhere
 
UltraBlue BOA deglinted
  • the atmosphere is thick and heterogeneous
  • water volume reflectance over deep waters
    • is mostly at ~25, variable 
    • but decreases to 4 where Coastal 5 waters are observed at the end of  Ampasindava Bay 
 
BLUE TOA deglinted
  • the atmosphere is thick and heterogeneous
  • water volume reflectance over deep waters
    • is mostly at ~20, variable 
    • but decreases to 9 where Coastal 5 waters are observed at the end of  Ampasindava Bay 



Optical calibration over the whole image
 

Calibration 6,5,3,2 whole image
  • 0-15 m: K2/K3=0.60     OIB+0.8
  • >15  m: K2/K3=0.50     OIB+0.4

Calibration 6,5,3,1 whole image 
the UltraBlue band_1 behaves as 
expected

 

Calibration 6,5,4,2 whole image   
the panchromatic band_4 is under control
The calibration over the full image
captures the optical properties
of the clearest waters

But, zooming on local ROIs
reveals distinct spatial heterogeneity:
as documented below:
this cannot be overlooked
while inverting the simplified RTE.





 





Optical calibration under 8 local ROIs
 
Calibration under mask_E1_white
  • K2/K3~0.55
  • The Grand Serpent shoal is ~10 m deep
 
Calibration under mask_E2_red
  • 0-6 m: K2/K3 decreases from 1.0 to 0.8
  • 6-~15 m: K2/K3 stable   at ~0.8
  • >15    m: K2/K3  clearer at ~0.7
 
Calibration under mask_E3_turquoise
  • K2/K3~0.58
  • The Grand Banc de l'Entree shoal is ~10 m deep
 
Calibration under mask_E4_green
  • K2/K3~=1.0 down to ~5 m
  • becomes clearer at depth
  • this case is humbling!
 
Calibration under mask_E5_black
Baie des Russes
  • K2/K4=K3/K4~=0.38
  • therefore: K2/K3~=1.0 down to >6 m
  • bottoms are fairly dark
 
Calibration under mask_E6_purple
  • down to ~5 m :K2/K4=K3/K4~=0.38
    • therefore: K2/K3~=1.00
  • deeper than 5 m: K2/K3~=0.70 
    • waters become clearer at depth  
  • bottoms are fairly dark
 
Calibration under mask_E7_yellow
  • K2/K4~=0.18
  • K3/K4~=0.29
  • therefore: K2/K3~=0.6
  • this water type OIB+0.8 is present over the whole depth range in the vicinity of Nosy Iranja island 
  • bottoms are fairly bright
 
Calibration under mask_E8_pink
  • K2/K4~=0.9
  • K3/K4~=0.7
  • Kblue>Kgreen!
  • therefore: K2/K3~=1.3
  • this is water type Coastal C5 of Jerlov in  Ampasindava Bay
   
   



 
 


 



 



Créer un site
Créer un site