Bathymetry and water column correction
LANDSAT 8 OLIP 
at Shark Bay and Hamelin Pool, Western Australia
4395*5915, 30 m pixel size, UTM zone 49
scene LC80090452013133LGN01, April 18th 2013 downloaded from USGS
Work done september 2013

 
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

Please refer to Bora Bora and Sanaa
for use of the PANchromatic band for water column correction
 
Please refer to
  • HICO user’s Annual Report, August 2012 "Determining inherent optical properties, bathymetry and benthic types using HICO imagery within northeast and Western Australia " 
  • HICO User’s Annual Report, April 2014 "Examining the potential of detecting change in HICO-derived bathymetry: a case study of Shark Bay, Western Australia"
  • Bierwirth & Burne Landsat Shark Bay Onshor.pdf
  • Burling_thesis.pdf
  • "Intertidal and Subtidal Microbialites in the Shark Bay World Heritage area" Lindsay Collins, Jahnert & Collins 2011, 2012 
  • WAMSI Shark Bay Project Talk S Blake - 3 Apr 2012 web.pdf


 
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
 
home
The data
Deglinting
Optical calibration
Normalization
Retrieved depth                                           Hamelin Pool 
Water column correction
Using the PAN band
Difficulties
Command Line
Download the result





The data

Seven bands are used: Purple, Blue, Green, PAN, Red, NIR and SWIR1

FCC

TCC

GREEN band
Offshore:
the Leeuwin current carries turbid  waters

Panchromatic band, sum-aggregated to 30 m,
Note the smal group of clouds East of  Cape Peron








Deglinting

Glint regression based on SWIR1 band

 

Glint regression based on NIR band
 
  • Only a few scatters of small clouds need deglinting ast of Cape Peron.
  • Apart from these, no need for deglinting.
 
  • The PAN band exhibits very noisy correlation
    • I don't know why
    • Maybe poorly co-registered image after 4*4 sum aggregation
    • Or, as is the case with WV2 images: is there a time gap between PAN and other visible bands?







Optical calibration

Water type OIB of Jerlov
Spectral BOA reflectance of the brightest bottoms:
RM: 0.184 0.233 0.333 0.346 0.437 0.635 0.729
on the scale 0-1, as calibrated using the _MTL.txt textfile

Calibration diagram
for bands Blue, Green, Red and NIR

Calibration diagram
for bands Purple, GreenRed and NIR
Spectral operational 2K
  • This calibration yields the following operational
2K0.085/0.076/0.169/0.182/0.606/5.084/18.7
2Kblue  =0.076 m-1
2Kgreen=0.169 m-1
2Kred    =0.606 m-1
2Knir     =5.1    m-1


Final_Z = CoefZ*Z  - Tide_to_Datum
Operational wavelengths
  • These diagrams demonstrate that a satisfactory fit of the BPL mode has been achieved among all visible bands
  • -WL/440/480/560/590/620/865/1610
  • This was obtained by setting
    • WLpurple at 440 nm
    • WLblue     at 480 nm
    • WLgreen   at 560 nm
    • WLred       at 620 nm
Only WLred is set away from mid-waveband, at a very low value  like at Caicos

This clearly suits the data comfortably,
and yield depth results that approximately match those reported in
Burling_thesis.pdf

 
LsM
Conversion of LsM to BOA reflectance yields
  • RM: 0.184 0.233 0.333 0.346 0.437 0.635 0.729
    • this is very close to coral sand reflectance published by S. Maritorena 1994, and by C. Davies 2012
  • This high reflectance is only represented at very shallow depth: this might be gypsum deposits, as reported in "swwa_microbialites_lcollinsx.pdf"
LsM
Conversion of LsM to BOA reflectance
  • This very high reflectance
    • RM/0.184/0.233/0.333/0.346/0.437/0.635 0.729
    • is slightly less bright than the sand dunes on land
  • It is ~1.4 times brightest than that of the brightest bottoms at depths
    • RM/0.132/0.167/0.228/0.249/0.312/0.454/0.520
  • Settling for the high LsM carries the following benefits:
    • increases dramatically the S/N ratio of dark pixels
    • which in turn increases the operational depth range

Benefits of normalization
of the water column correction process
  • It is therefore desirable to adopt coral sand BOA reflectance as a way
    • to normalize the water column corrected spectral bottom reflectances, and therefore to allow for time series studies
    • to increase the S/N ratio of shallow bottoms, and therefore improve the precision of results at low bottom-reflected signal

Normalization

of the water column correction process