Optical calibration, bathymetry, water column correction and bottom typing of shallow marine areas, using passive remote sensing imageries
 
a 1024*1024 SPOT 1 XS 3-bands image (1986, processed in 1994 and 2008)
a 1383*842 EO1-ALI 6-bands image (2003, processed in 2010)
of Moorea Island, French Polynesia
 


 
1 - NO NEED for field data, nor for atmospheric correction
2 - this is demonstrated in this website, using a variety of hyper/multi spectral data
 
Requirements are
1 - homogeneous water body and atmosphere
2 - some coverage of optically deep water
3 - some coverage of dry land
 
Problems are
1 - the precision on estimated depth is found wanting, because the noise-equivalent change in radiance  of accessible data is too high for shallow water column correction work 
2 - radiance data should be preprocessed by the provider at level 1 in order to improve S/N ratio
3 - exponential decay: the deeper/darker the bottom, the poorer the performances
 
So
I keep digging
until suitable data
become available
 
 
Please refer to Stephane Maritorena,
"Remote sensing of water attenuation in coral reefs: a case study in French Polynesia":
IJRS, vol 17, no 1, pp 155-166, 1996

 
This is the image that MARITORENA 
could not calibrate using Lyzenga's method 
in his PhD work.
Please see below that
the 4SM method allows for  
an easy and reliable optical calibration  
of shallow areas for this image.
DGPS navigated precision depth profiling
courtesy of SHOM (Tahiti, 1994)
An average seatruth depth has been computed manually from SHOM's precision depth sounding profiles at each 20*20 m seatruth pixel for Temae, Nuupere and Orau lagoons (IFREMER-SPT, Tahiti, 1994).

 


Image_B


Image_Z


N=1073, RMS=0.45 m

  • Smart-Smoothing does a good job here on all seatruth pixels
  • Too many seatruth depths in the 2-3.0 m depth range result in biased regression line
  • Optical calibration is optimal needs no adjustments


N=1046, RMS=0.67 m

  • No Smoothing yield poor results here on all seatruth pixels
  • Too many seatruth depths in the 2-3.0 m depth range result in biased regression line


Smart-Smoothing does a good job
particularly here only on darkest pixels 0<B<100

  • Too many seatruth depths in the 2-3.0 m depth range result in biased regression line
  • Most darker pixels are affected by coral growths amid sandy bottoms: this makes their "seatruth depth" a delicate concept.


No Smoothing yield poor results
particularly here only on darkest pixels 0<B<100

  • Too many seatruth depths in the 2-3.0 m depth range result in biased regression line
  • Most darker pixels are affected by coral growths amid sandy bottoms: this makes their "seatruth depth" a delicate concept.
 
Optical calibration diagram

Optical calibration diagram
SPOT

SPOT

WL[green]=550 nm,
WL[red]=650 nm

K[green]/K[red]~=0.26

an optical water type OII+0.75 of Jerlov

2K[green]=0.225 ==> ZMax=23.3 m
2K[red]=0.867 ==> ZMax= 6.1 m

ALI

ALI

WL[blue]=482.5 nm,
WL[green]=576.0 nm
WL[red]=666.0 nm

K[blue]/K[green]~=0.56

an optical water type OII+0.11 of Jerlov

2K[blue ]=0.140 ==> ZMax=39 m
2K[green]=0.248 ==> ZMax=23 m

ALI seatruth regression for Temae lagoon

The 1986 SPOT image data is affected by
  • electronic system noise: striping , crosstalk in the NIR band (right)
  • adjacency effect in the NIR band (right)

<== Green, Red and NIR bands, histeq enhancement
<== RGB, Image_Z and Image_B
This limits the potential of the image 
for a shallow water application
The 2003 ALI image data
is of excellent radiometric quality

Processing this image hardly require any smoothing

Only the glint require proper attention

  
Seatruth


TEMAE lagoon
 

Stability: these profiles exhibit remarkable stability over the years
Clear waters: the southern channel has clear waters in the 2003 ALI image


Seatruth

Seatruth regression for Temae lagoon


NUUPERE lagoon
RIGHT: seatruth profiles from West to East

Stability: these profiles exhibit remarkable stability over the years
Nothing will do: the northern channel has turbid waters in the 2003 ALI image; results there are artifacts


ORAU
RIGHT: seatruth profiles from West to East

Stability: these profiles exhibit remarkable stability over the years
Nothing will do: the southern channel has turbid waters in the 2003 ALI image, much more than in the 1986 SPOT image; results there are artifacts


ALI
True Color Composite
of the normalized bottom reflectance

 

ALI
Classified into seven shallow bottom types

unclassified shallow pixels are in red

[up][home]