Optical calibration, bathymetry, water column correction and bottom typing of shallow marine areas, using passive remote sensing imageries
Bathymetry and water column correction
at Caicos Bank, Bahamas
4018*4149, 30 m pixel size, UTM zone 18, downloaded from USGS

Panchromatic band
Using the Panchromatic band for water column correction to derive water depth and spectral bottom signature:

Landsat 8 OLIP bandset used for this work

Coastal_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

Some recents additions to 4SM capabilities:
Calibration optimization now operational
Vegetation Line for Lw and La


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
I keep digging
until suitable data
become available
This is a time series of Landsat 8 OLI+PAN images
work done july-september 2016

Seatruth DTM was purchased 
from Harris-Ellis_Knowledge_Transfer

The Harris-Ellis dataset was used 
for research on carbonate sands bodies
on the Bahamas:

Satellite Imagery and Visualization of the Caicos Platform

Location of depth soundings
used to produce a crude DTM
Purple depth points are satellite derived
Others are derived from soundings and nautical chart

  • DTM indicates that optical calibration using Jerlov data is correct and suitable
  • And also that using the PAN solution is correct
  • Co-registration is excellent
  • Three main sources of error:
    • DTM very coarse
    • Biased retrieval: Htide is un-realistic; this still needs investigation
    • Genuine depth change over time
  • DTM is too coarse for a fair estimation of RMSE
    • there is nothing like a LIDAR or MBES DTM for estimating RMSE
  • This DTM only useful in the 1-17 m depth range
    • lacks depths 17-30 m
  • DTM allows for invaluable insight into 4SM modeling practise
    • 4SM now tuned for Landsat 8 OLIP data, next WV2P

Remarkable findings and improvements
This series of images offers
a good opportunity
for strengthening
4SM tools
Regressions: the slope of the regressions is found to be ~=1 :
  • this owes nothing to the existing DTM,
  • but rather offers a nice confirmation that YES:
  • thanks to Jerlov's data, no need for field data for optical calibration?
SAM: water column
corection results
mapped to
spectral angle
offer a good
opportunity for
time series studies
MAY 2013   RMSE=0.52 m
SWIR deglinting very good
RED bottom detection down to ~8-9 m==>low KRED
Excellent calibration
PAN solution better than GREEN solution
DTM seatruth: no need for field data
BOA reflectance==>SAM bottom classification
JANUARY 2014   RMSE=0.57 m
SWIR deglinting very good
RED bottom detection down to ~8-10 m==>low KRED
Excellent calibration
BOA reflectance==>SAM bottom classification
FEBRUARY 2014  RMSE=0.61 m
MARCH 2014   RMSE=0.63 m
PAN solution safer, avoids artifacts
"Whiting" or resuspension?

AUGUST 2014   RMSE=0.59 m
SEPTEMBER 2014   RMSE=0.73 m
A bad image: complex atmosphere, poor deglinting
Stratified waters are correctly  accounted for in 4SM: very clear in the SE corner
DTM is faulty in the SE corner
KBLUE/KGREEN=0.42 ==> WaterType OIB+0   ==>   No need to decrease KRED

Something is very wrong: deeper than ~10 m: calibration not under control CoefKPAN=1.25
JANUARY 2015   RMSE=0.93 m
Bad atmosphere
4SM model is not  under good control

Vegetation Line==>Lw=Lsw-La==>BOA==>BOA Reflectance
JULY 2015   RMSE=0.69 m
Bad atmosphere
Htide=1.5 m: something's wrong!
Very green in august
Htide=0.7 m

BEST IMAGE                           OCTOBER 2015  RMSE=0.56 m
Vegetation Line==>Lw=Lsw-La==>BOA==>BOA Reflectance
K1/K2=(K1/K3)/(K2/K3) must be satisfied

Bottom typing
KCOASTAL  must be slightly decreased


APRIL 2016   RMSE=0.66 m
very dense and heterogeneous atmosphere
several very bright spots
Lsw for PAN needs some increase
MAY 2016   RMSE=0.66 m
The histogram is a better way to estimate a tide height,  rather than the regression