Optical calibration, bathymetry, water column correction and bottom typing of shallow marine areas, using passive remote sensing imageries
busy? see 4SM slides
4SM seatruth

Your comments are invited             4SM conclusions               See you on 4SM blog


 
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
 
 

Seatruth evidence
reasonable performances can be achieved,
provided the image is worthy of shallow water work.

 
WV2 Buck Island Reef, USVI                                                     may 2013
  Bahrain march 2011
  Oahu, Hawaii, USA july 2011
  Princess Cays, Bahamas  
     
CASI Heron Island, GBR, Australia december 2013
  Prince Edward Island, Nova Scottia, Canada february 2002
     
TM/ETM Caicos Bank, Bahamas  
  Geraldton, Western Australia december 2011
  Negril Shores, Jamaica  
     
IKONOS Lee Stocking Island, Bahamas september 2002
  Dubai july 2009
     
SPOT Moorea Island, French Polynesia 2008





The two seatruth plots below
are typical of what can be achieved by 4SM
  • Use of Jerlov's data for optical calibration  really does a good job:
    • see how the majority of pixels are assigned a retrieved depth which is reasonably accurate
    • accuracy is reasonable
    • precision decreases as the bottom depth increases
  • But foul waters at particular locations cause severe artifacts:
      
 
 




 
 

Seatruth dataset formats suitable for 4SM

Seatruth dataset can be of several sorts
depth points
depth sounding lines
text file or shapefile
digital terrain model

OVERLAY
Seatruth datasets must be in the same projection and datum as the spectral image

4SM shall discard the data which map outside the spectral image
 
 



1 - Depth sounding lines : the general case
  • XYZ textfile for depth sounding lines
  • Each separate depth sounding line starts with a header
  • Any number of profiles
  • Any number of points in each profile
> LineName_1 -
X    Y    Z
X    Y    Z
 .....

> LineName_2 -
X    Y    Z
X    Y    Z
X    Y    Z
 .....
  • Best is one shapefile
    • in the same georeferencing as the spectral image
    • with an attribute for Depth in meters, positive
    • comprised of any number of separate profiles/polylines
    • any number of depth points per profile/polyline
    • each polyline may be comprised of several segments
  • 4SM shall call shp2text to
    • read this shapefile for XUTM   YUTM   Z_in_meters
    • name the separate profiles with suffix A to Z
    • convert into a suitable text file as above


 




2 - Depth points : formatted as a one-line dataset

  • XYZ textfiles for scattered depth points
  • This is the basic format
  • XUTM   YUTM   Z separated by space
    • XUTM   and YUTM in meters
    • Z   Depth in meters, positive

 
  • Best is one shapefile
    • in the same georeferencing as the spectral image
    • with an attribute for Depth in meters, positive
    • comprised of any number of separate profiles
    • any number of depth points per profile
  • 4SM shall call shp2text to
    • read this shapefile for XUTM   YUTM   Z_in_meters
    • name the separate profiles with suffix A to Z
    • convert into a suitable text file as above





3 - DTM for gridded digital terrain model

  • DTM stands for gridded digital terrain model
    • DTM to be generated from scattered depth points or from depth sounding lines
    • grid spacing to match that of the spectral image
    • datum/georeferencing to match that of the spectral image
    • NoData grid points to be mapped to 0
    • Depth in centimeters, positive
  • 4SM shall crop/import the DTM to overlay the spectral image
    • This becomes one more S16 channel in the spectral image,
    • and shall be loaded as such in a S16 channel, with depths in centimeters