Optical calibration, bathymetry, water column correction and bottom typing of shallow marine areas, using passive remote sensing imageries
 

CASI image of Menton, French Riviera
Work done in 2005 and 2008

Please refer to Jaubert et al. Marine Ecology Progress Series, Vol 263: 75-82, 2003

A late afternoon image, courtesy of HDI , collected october 2000.
This is a 12 bands image, 1843*1704 pixels, pixel resolution 1 meter.

The 4SM licensekey is rated at 579 US$ for this image.

Removing limb brightening

return to 4SM Study Cases
-WL467/508/523/535/544/554/565/574/585/595/644/845 in nanometers
Unfortunately, this dataset lacks bands in the red and green regions of the solar spectrum.
This does not help for an optimal deglinting, and also for calibrating/modeling!



 
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
 

 


Normalized bottom reflectance

the bright green area
is identified by Jaubert et al. as
a bed of Caulerpa Taxifolia seaweed,

while the pale orange areas
are identified as
beds of photophilic algae.

Clearly, radiometric quality is a must, 
or it shows badly!

 

 

Processing april 2008



Processing april 2008

 



Late afternoon flights entail sunglint from the SW
 


Deglinting could not remove limb brightening
 


NIR band
notice a slight adjacency effect

 


Red band
notice the effect of stirring of waters along the beach

Green band at 554 nm

Blue band at 467 nm
notice that there is no negative bottom contrast
Depths are computed down to ~ 20 m
see menton_Z250_modal_filtereed
see menton_wZ

Bottoms are quite dark

 



True color composite of water column corrected image
does not show much spectral diversity



"Normalized" image LSB
enhances various bottom types

 


Bottom typing
of water column corrected spectral image

Colors show the amount of spectral contrast
for each bottom type mapped. 




Regions of interest
for 13 bottom type signatures are shown

see menton_LBS_classified_color_modal_filtered

 
Bottom Typing
and modal filtering
in 4SM
13 classes ar distinguished, from 6 to 13


  • all pixels are classified
  • Limb brightening towards the East causes some trouble
  • Stirring of the waters along the waterline had to be countered
  • See all 13 spectral bottom type signatures: their color in the image to the left are taken from the "normalized" water column corrected  bottom image.


  • These classes should be binned into famillies of similar bottom types, and renumbered accordingly.
  • Then they can only be labelled through field data.
 
signature 6                                    signature 7
    
 
 signature 8                                    signature 9
 
signature 10                                   signature11
 
 
signature 12                                  signature13
 
 
signature 14                                  signature15
 
signature 16                                 signature17   signature 18
 
STATISTICS OF ALL SHALLOW BOTTOM TYPING: BOA radiance 0-200
CoefSIG=8.00 nbMISS=0 -E1/1843/1/1704  on Sat May 10 23:19:35 2008
SIGnbMax=18
Type_0 is Land pixel
Type_1 is Unclassified Shallow pixel
Type_2 is Optically Deep water pixel
Type_3 is No Data pixel
Type_4 is Wave Breaker pixel
Type_5 is Cloud/Shadow pixel
Type_6 and above is for Classified Shallow pixel

0.01% as type_1   Shallow_pixel_NOT_Classified

22.50% as type_10  SIG   0.0  35.7  49.3  54.4  52.2  51.4  50.8  45.6  47.9  47.9  47.9  47.9
B_47.9  SD_13.0    STD   0.0   6.5   6.2   6.6   6.0   5.8   6.2   6.3   6.2   6.2   6.2   6.2

20.21% as type_15  SIG   0.0  57.4  59.1  58.9  58.9  58.9  58.4  57.8  57.2  60.4  58.6  58.6
B_58.6  SD_2.2     STD   0.0   1.1   1.1   1.2   1.3   1.4   1.6   1.6   1.4   1.4   1.2   1.2

13.06% as type_13  SIG   0.0 271.1 316.8 304.8 274.6 291.9 214.5 214.5 214.5 214.5 214.5 214.5
B_291.8 SD_16.7    STD   0.0  71.6  85.5  80.5  69.5  76.5  25.5  25.5  25.5  25.5  25.5  25.5

8.16% as type_8   SIG   0.0  62.8  64.6  65.6  66.0  66.5  66.8  66.8  65.7  62.8  64.8  64.8
B_64.7  SD_1.5     STD   0.0   1.2   0.9   0.7   0.7   0.8   1.0   1.1   0.9   1.6   0.8   0.8

7.31% as type_12  SIG   0.0  52.8  54.1  53.4  52.8  52.5  51.9  51.3  51.3  56.7  53.3  53.3
B_53.3  SD_2.8     STD   0.0   1.5   1.5   1.5   1.5   1.5   1.5   1.5   1.5   1.6   1.5   1.5

6.55% as type_16  SIG   0.0  73.8  73.4  72.5  70.2  69.1  70.1  70.4  75.0  72.5  72.5  72.5
B_72.5  SD_2.2     STD   0.0   1.8   1.6   1.6   1.6   1.6   1.5   1.7   1.7   1.6   1.6   1.6

6.45% as type_14  SIG   0.0  75.8  76.5  78.5  78.4  78.4  78.4  78.3  77.7  76.8  77.3  77.3
B_77.3  SD_5.0     STD   0.0   3.8   3.7   3.8   3.9   3.8   4.0   3.8   3.9   3.8   3.8   3.8

5.20% as type_9   SIG   0.0  32.6  35.2  36.5  38.0  39.2  39.0  39.4  38.7  33.8  35.6  35.6
B_35.6  SD_7.3     STD   0.0   2.4   2.6   2.7   2.7   2.6   2.7   2.6   2.6   2.5   2.6   2.6

4.39% as type_7   SIG   0.0 110.2 109.2 106.9 105.3 104.5 103.2 103.6 105.5 112.9 107.9 107.9
B_107.9 SD_7.7     STD   0.0   9.3   8.6   8.2   7.9   7.6   7.3   7.5   8.0   9.7   8.5   8.5

2.28% as type_18  SIG   0.0  18.9  29.0  32.5  32.0  31.6  30.5  26.1  28.1  28.1  28.1  28.1
B_28.1  SD_8.1     STD   0.0   1.7   2.4   2.6   2.5   2.4   2.4   2.2   2.3   2.3   2.3   2.3

2.02% as type_11  SIG   0.0  16.6  19.3  18.5  17.2  17.9  17.9  17.9  17.9  17.9  17.9  17.9
B_17.9  SD_17.5    STD   0.0   2.9   3.4   3.3   3.0   3.1   3.1   3.1   3.1   3.1   3.1   3.1

1.01% as type_6   SIG   0.0 169.9 172.7 174.4 174.6 174.5 174.3 174.3 173.9 173.4 172.9 172.9
B_172.9 SD_6.3     STD   0.0  10.7  10.9  11.3  11.2  10.8  11.0  10.8  10.8  10.8  11.6  11.1

0.86% as type_17  SIG   0.0  10.9  12.0  12.7  13.4  14.1  14.1  14.4  14.0  10.9  12.2  12.2
B_12.2  SD_34.3    STD   0.0   3.4   4.0   4.4   4.5   4.6   4.8   4.7   4.3   3.3   4.1   4.1
STATISTICS OF ALL SHALLOW BOTTOM TYPING: BOA radiance 0-200
CoefSIG=8.00 nbMISS=0 -E1/1843/1/1704  on Sat May 10 23:19:35 2008
SIGnbMax=18
   Type_0 is Land pixel
   Type_1 is Shallow_pixel_NOT_Classified
   Type_2 is Optically Deep water pixel
   Type_3 is No Data pixel
   Type_4 is Wave Breaker pixel
   Type_5 is Cloud/Shadow pixel
   Type_6 and above is for Classified Shallow pixel

0.01% as type_1, i.e. Shallow_pixel_NOT_Classified

1.01% as type_6   SIG   0.0 169.9 172.7 174.4 174.6 174.5 174.3 174.3 173.9 173.4 172.9 172.9
B_173 SD_6.3    STD   0.0  10.7  10.9  11.3  11.2  10.8  11.0  10.8  10.8  10.8  11.6  11.1

4.39% as type_7   SIG   0.0 110.2 109.2 106.9 105.3 104.5 103.2 103.6 105.5 112.9 107.9 107.9
B_108 SD_7.7    STD   0.0   9.3   8.6   8.2   7.9   7.6   7.3   7.5   8.0   9.7   8.5   8.5

8.16% as type_8   SIG   0.0  62.8  64.6  65.6  66.0  66.5  66.8  66.8  65.7  62.8  64.8  64.8
B_65  SD_1.5    STD   0.0   1.2   0.9   0.7   0.7   0.8   1.0   1.1   0.9   1.6   0.8   0.8

5.20% as type_9   SIG   0.0  32.6  35.2  36.5  38.0  39.2  39.0  39.4  38.7  33.8  35.6  35.6
B_36  SD_7.3    STD   0.0   2.4   2.6   2.7   2.7   2.6   2.7   2.6   2.6   2.5   2.6   2.6

22.50% as type_10  SIG   0.0  35.7  49.3  54.4  52.2  51.4  50.8  45.6  47.9  47.9  47.9  47.9
B_48  SD_13.0   STD   0.0   6.5   6.2   6.6   6.0   5.8   6.2   6.3   6.2   6.2   6.2   6.2

2.02% as type_11  SIG   0.0  16.6  19.3  18.5  17.2  17.9  17.9  17.9  17.9  17.9  17.9  17.9
B_18  SD_17.5   STD   0.0   2.9   3.4   3.3   3.0   3.1   3.1   3.1   3.1   3.1   3.1   3.1

7.31% as type_12  SIG   0.0  52.8  54.1  53.4  52.8  52.5  51.9  51.3  51.3  56.7  53.3  53.3
B_53  SD_2.8    STD   0.0   1.5   1.5   1.5   1.5   1.5   1.5   1.5   1.5   1.6   1.5   1.5

13.06% as type_13  SIG   0.0 271.1 316.8 304.8 274.6 291.9 214.5 214.5 214.5 214.5 214.5 214.5
B_36  SD_16.7   STD   0.0  71.6  85.5  80.5  69.5  76.5  25.5  25.5  25.5  25.5  25.5  25.5

6.45% as type_14  SIG   0.0  75.8  76.5  78.5  78.4  78.4  78.4  78.3  77.7  76.8  77.3  77.3
B_77  SD_5.0    STD   0.0   3.8   3.7   3.8   3.9   3.8   4.0   3.8   3.9   3.8   3.8   3.8

20.21% as type_15  SIG   0.0  57.4  59.1  58.9  58.9  58.9  58.4  57.8  57.2  60.4  58.6  58.6
B_59  SD_2.2    STD   0.0   1.1   1.1   1.2   1.3   1.4   1.6   1.6   1.4   1.4   1.2   1.2

6.55% as type_16  SIG   0.0  73.8  73.4  72.5  70.2  69.1  70.1  70.4  75.0  72.5  72.5  72.5
B_73  SD_2.2    STD   0.0   1.8   1.6   1.6   1.6   1.6   1.5   1.7   1.7   1.6   1.6   1.6

0.86% as type_17  SIG   0.0  10.9  12.0  12.7  13.4  14.1  14.1  14.4  14.0  10.9  12.2  12.2
B_12  SD_34.3   STD   0.0   3.4   4.0   4.4   4.5   4.6   4.8   4.7   4.3   3.3   4.1   4.1

2.28% as type_18  SIG   0.0  18.9  29.0  32.5  32.0  31.6  30.5  26.1  28.1  28.1  28.1  28.1
B_28  SD_8.1    STD   0.0   1.7   2.4   2.6   2.5   2.4   2.4   2.2   2.3   2.3   2.3   2.3 

 

Calibration diagram

Jerlov water type ~OII+0.86
A difficult case, not for beginers:
 
  • locally very bright bottoms at the beach,
    • and fairly dark bottoms everywhere else
  • along the beach down to ~4-5m, water type seems to be ~OIII+0.55,
    • clearing to  water type ~OII+0.86 at depth











 


Removing the limb brightening 
on a raw flight line 512 pixels wide

Menton, French Riviera 1998
Raw 512 wide flight line OOE271 by HDI/CSM
This 512*3587 14 bands CASI image is a courtesy of HDI, 1998



Radiometric improvement

of raw 512 wide flightline data

  • Left : raw and de-Limbed/de-Rowed

  • Right:

    • upper: raw

    • lower: corrected

  • DeLimb: takes out the limb brightening caused by side-viewing away from vertical

  • DeRow : corrects for faint row miscalibration

  • This correction is based on radiometric properties sampled over opticaly deep lines at the top of the flightline

 

Radiometric improvement

of raw 512 wide flightline data

  • Left : raw and de-Limbed/de-Rowed

  • Right:

    • upper: raw

    • lower: corrected

  • DeLimb: takes out the limb brightening caused by side-viewing away from vertical

  • DeRow : corrects for faint row miscalibration

  • This correction is based on radiometric properties sampled over opticaly deep lines at the top of the flightline

  • There is still a short transient decrease of light though

 
Raw data
 
Delimbed data
two stripes are also Deglinted - no smoothing

The Delimbing-deglinting process

  • limb brightening is very important

  • skyglint is quite low: what is observed after delimbing is mostly random noise which may only be smoothed at a later stage




Illustration of the Delimbing process
on a 512 wide raw CASI flightline

from bottom to top:

  • Raw data: the limb brightening is very obvious: it must be removed prior to geocoding of the image

  • Delimbed/Derowed:note that in spite of the derowing of individual rows, there remains groups of adjacent rows which are slightly brighter or darker: this shows how important for shallow water work that the imaging instrument be optimaly tuned

  • Delimbed/Derowed/Deglinted/Smart-smoothed

  • Delimbed/Derowed/Deglinted/NOT-smoothed: in spite of all efforts, there remains all sorts of noise features and also remnants of imperfect deglinting: therefore it is very important that the data be smoothed prior to modeling

 

  • This image preparation is done "on the fly" while modeling, as 4SM reads the raw U8 or U16 data and applies all the required radiometric improvements on a pixelwise basis prior to modeling