Optical calibration, bathymetry, water column correction and bottom typing of shallow marine areas, using passive remote sensing imageries
CASI image of Bay of Porquerolles, French Riviera
Image courtesy of HDI , collected october 2000. This is a 12 bands image, 3148*2664, pixel resolution 1 meter.
The 4SM licensekey is rated at 1003 Euros for this image.


Please refer to Jaubert et al. Marine Ecology Progress Series, Vol 263: 75-82, 2003
Healthy Caulerpa Taxifolia seagrass  are invading very dark Posidonia meadows

work updated april 2008

return to 4SM Study Cases



 
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
 

-WL467/508/523/535/544/554/565/574/585/595/644/845 in nanometers
Unfortunately, this dataset lacks bands in the red-to-green region of the solar spectrum.
Only one blue band.
This does not help for an optimal deglinting, and also for calibrating/modeling.

Lucky that we have one NIR band!

The best calibration of CASI data I have seen to-date, april 2008 with 4SM.2.06.
Lots of clean bright sands here, at all depths.
I wish I was given the chance of a proper 4SM work on this dataset:
starting from raw flightlines, myself on my own machine.
 

The best calibration
of CASI data in town!


 See that wavelengths adopted for this bandset 
fit well an intermediate curve 
of Jerlov's classification of marine waters:

Jerlov's water type was ~OII+0.28 in october 2000
  Path radiance:

see in particular that it is very clear how 
the atmpspheric path radiance
is determined by use of 
the SoilsLine by NIR here.
 

The worst mosaic in town!

This image was not corrected 
for limb brightening,
nor for variable illumination,
nor for bad rows,
nor for fancy ILS pixels.
 
 

The worst mosaic in town!

Deglinting-Smoothing can't make this image
worthy of shallow water work,
although this marine park sanctuary site
was quite a bit of an opportunity.
Certainly worth a proper pre-processing/mosaicking:

I wish I was given the chance
of a proper 4SM work on this dataset,
starting from raw flightlines,
myself on my own machine.

There is a lot to learn from such a project,
both for data collection/pre-processing
and for 4SM processing.


 

Cleaned_Equalized_Deglinted image

Some cleaning/preprocessing of raw data 
was really worth the pain.
BUT REMEMBER
GARBAGE IN ==> GARBAGE OUT anyway.


Color Composite
of Normalized Spectral Bottom Reflectance
According to Jaubert et al.,
one large and well defined bright green bed 
of Caulerpa taxifolia lies at a former wreckage place.




Computed Depth image_Z250
image_Z250 modal filtered
image_wZ: depth is computed down to ~15 m

 
 


Average Bottom Reflectance image_B
mostly extermely dark bottoms


Color composite of

the "water column corrected" image_LBS
like if at low tide

Mostly very dark bottoms:
this site is said to have a lot of seaweeds



Histogram equalization of image_LBS

helps to see throuh

and its normalization helps further


 

Negative bottom contrast ...

Negative bottom contrast abound
over very dark bottoms
here at 467 nm


... a real challenge for modeling!

Negative bottom contrast abound
over very dark bottoms
here at 523 nm







Now we can run
a bottom type classification in 4SM

as of april 2008
This was done in 4SM without any access to seatruth,

including a final modal filtering

Under such adverse conditions,
the equalization has been a real challenge.

Too bad the deglinting was not able
to clean properly the heavy sun glint.

Flying into or away from the sun
would have been advisable indeed
in view of shallow water work!
 

23 bottom types are mapped:
mostly three famillies of bottom types:
  • Sands with vegetation:  types 6 to 11
  • "green" vegetation:        types 12 and 13
  • "brown" vegetation:       types 14 to 23
STATISTICS OF
SHALLOW BOTTOM TYPING

BOA radiance 0-200

on Tue Apr 29 06:21:54 2008

 


 

 
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

 
 1.05% as type_1, i.e. Shallow_pixel_NOT_Classified

 2.27% as type_6   SIG 163.7 168.5 179.6 167.7 166.6 164.2 165.0 168.6 163.4 167.1 169.7 167.6
BGSL_168 SD_5.3    STD  12.3   8.8   8.1   7.9   8.2   8.0   8.1   8.8   7.8   9.7   9.8   9.4

 
3.41% as type_7   SIG 145.6 155.3 170.4 162.0 163.2 162.5 164.6 170.4 160.3 159.8 148.3 155.9
BGSL_156 SD_2.5    STD   3.6   3.7   4.3   4.1   3.9   4.1   3.8   4.8   4.0   4.0   3.7   3.6

 
5.33% as type_8   SIG  89.7 109.4 118.0 105.5 103.5 100.3 101.3 105.5 102.4 105.5  97.0 102.0
BGSL_102 SD_3.3    STD   3.5   3.7   3.8   3.3   3.2   3.0   3.0   3.1   3.2   3.3   3.8   3.5

 
1.90% as type_9   SIG  74.6  95.0 103.6  93.2  92.5  89.8  91.5  94.8  91.2  92.0  81.8  88.2
BGSL_88  SD_3.5    STD   3.3   3.3   3.5   3.1   3.0   2.9   2.9   2.8   2.9   3.1   3.4   3.1

 
2.88% as type_10  SIG  67.1  83.9  92.3  83.4  82.4  80.3  81.4  86.2  82.5  83.6  72.9  78.6
BGSL_79  SD_4.1    STD   3.4   3.6   3.8   3.3   3.0   3.0   3.0   3.2   3.1   3.3   3.4   3.4


 7.54% as type_11  SIG  44.9  51.2  57.6  55.8  57.8  57.6  60.4  61.3  55.3  51.8  47.6  51.4
BGSL_51  SD_8.5    STD   3.3   4.2   4.8   4.6   4.6   4.8   4.8   4.9   4.5   4.3   3.7   4.0

 
8.75% as type_12  SIG  27.7  25.3  28.6  27.4  28.4  28.4  29.9  31.7  28.7  28.3  26.1  27.3
BGSL_27  SD_7.7    STD   1.6   2.1   2.3   2.2   2.3   2.3   2.3   2.4   2.3   2.1   1.8   2.0

 
3.81% as type_13  SIG  16.0  18.6  21.1  18.4  19.4  19.2  20.8  21.8  20.1  19.6  16.2  18.1
BGSL_18  SD_5.4    STD   1.0   1.0   1.1   1.0   1.0   0.9   1.0   1.0   1.0   0.9   1.0   1.0


13.31% as type_14  SIG  39.3  32.3  35.2  32.9  34.3  33.3  36.5  36.0  35.2  35.5  35.4  35.4
BGSL_35  SD_12.5   STD   4.5   4.4   4.5   4.1   4.4   4.5   4.9   4.5   4.7   4.4   4.4   4.4


18.68% as type_15  SIG  22.0  19.0  20.0  16.7  16.8  16.2  16.9  18.1  17.1  18.7  19.9  19.5
BGSL_20  SD_14.5   STD   3.3   2.9   3.2   2.9   2.7   2.5   2.4   2.7   2.8   3.2   2.8   3.0

 
8.09% as type_16  SIG  18.6  11.7  14.2  13.3  15.1  13.1  17.1  14.7  15.0  15.0  15.0  15.0
BGSL_15  SD_6.6    STD   1.4   0.9   0.9   0.9   1.0   0.9   1.2   0.9   0.9   0.9   0.9   0.9

 
8.19% as type_17  SIG  16.8   9.4  11.4  10.7  11.8  11.1  13.7  12.9  12.7  12.7  12.7  12.7
BGSL_13  SD_5.6    STD   0.8   0.7   0.8   0.7   0.7   0.7   0.6   0.7   0.7   0.7   0.7   0.7

 
3.05% as type_18  SIG   9.1   4.5   5.1   6.2   6.4   5.8   8.9   9.4   8.3   4.4   6.6   6.6
BGSL_7   SD_12.1   STD   0.8   0.8   0.9   0.7   0.8   0.7   0.8   0.8   1.0   0.8   0.8   0.8

 
5.76% as type_19  SIG  13.2   0.8   3.1   2.7   5.1   4.1   8.1   7.2   3.9   6.1   6.1   6.1
BGSL_6   SD_17.4   STD   1.8   0.5   0.9   0.9   1.2   1.0   1.5   1.1   1.0   1.1   1.1   1.1

 
4.18% as type_20  SIG   8.2   2.0   3.3   3.7   5.1   5.0   7.4   7.4   5.3   3.3   5.0   5.0
BGSL_5   SD_27.5   STD   1.3   1.2   1.6   1.4   1.5   1.5   1.7   1.6   1.5   1.4   1.3   1.3

 
1.09% as type_21  SIG   4.4   3.1   4.7   2.6   4.3   3.7   6.2   5.4   3.8   1.1   3.9   3.9
BGSL_4   SD_10.7   STD   0.2   0.4   0.5   0.4   0.4   0.4   0.5   0.5   0.6   0.5   0.3   0.3


 0.39% as type_22  SIG   4.3   3.5   3.9   1.5   2.9   2.1   4.3   3.4   3.4   3.4   3.4   3.4
BGSL_3   SD_8.3    STD   0.2   0.2   0.3   0.3   0.3   0.2   0.2   0.5   0.5   0.5   0.5   0.5

 
0.33% as type_23  SIG   9.3   0.5   0.4   2.2   2.3   1.3   4.3   4.9   3.4   4.0   4.0   4.0
BGSL_4   SD_5.6    STD   0.4   0.2   0.3   0.2   0.3   0.3   0.2   0.3   0.3   0.1   0.1   0.1








Type 6


Type 7


Type 8


Type 9


Type 10


Type 11








Type 12


Type 13


23 bottom type
signatures
are extracted
from select
areas of interest
of the
"water column
 corrected"
image








Type 14


Type 16


Type 18


Type 20


Type 22


Type 23