Landsat 8 at La Parguera, Puerto Rico
1757*1219, 30 m pixel size, UTM zone 19, december 27th 2013
scene LC80050482013361LGN00, December 27th 2013. Image courtesy of the U.S. Geological Survey


image december 27th 2013
work done in may 2016

home


 


 
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
 
   
Depth Retrievals along Green Profile
PAN solution is a confirmed substitute to the GREEN solution

PAN solution vs GREEN solution

Profile Green
LIDAR backdrop
  • Segment D
    • deeper than ~5-10m, depths are badly underestimated
  • Segment C
    • depths are well estimated along the whole segment
  • Segment B
    • depths are badly underestimated from B1 to B2
  • Segment A
    • depths are well estimated in the 0-15 m depth range
PAN solution

Modeling by the "PAN solution"
The RED profile shows LIDAR depth seatruth
QUITE REMARKABLY,
THE PAN BAND IS NOT NOISY HERE
GREEN solution

Modeling by the "GREEN solution"
The RED profile shows LIDAR depth seatruth

PAN and GREEN solutions yield
approximately the same depth retrieval

 
GREEN vs PAN solution

PAN and GREEN solutions yield
approximately the same depth retrieval



Minute differences allow for
spectral features to appear
that the GREEN solution cannot reveal
GREEN vs PAN solution

Minute differences allow for
spectral features to appear
that the GREEN solution cannot reveal



ZGREEN-ZPAN
over ROI 2 at 15-25 m


 



ZGREEN-ZPAN
over ROI 5 at depths 0-10 m

This shows that
  • differences are mostly of +/-a few tens of centimeters, but can reach +/-a few tens of decimeters
  • therefore, using the PAN solution in place of the GREEN solution yields a spectral bottom signature which is more contrasted, as illustrated below
    • this can only increase the potential for spectral bottom typing
  • I look forward to pan-sharpenned version of such processing
This confirms
that the PAN solution
is trustworthy
to considerable advantage
 




 


Water column correction

TCC
deglinted, smoothed,
water column corrected and normalized
  • FOUL: areas badly affected by foul waters show in purple hues
    • this is an indication that water clarity in the blue range have been badly overestimated
  • Vegetation: areas which exhibit a greenish bottom signature show in green hues, likely bottom vegetation
  • Coral sand: bright areas are possibly to be coral sands

TCC
deglinted, not smoothed,
water column corrected

no smoothing here

TCC
deglinted, not smoothed,
and water column corrected
  • LEFT: the dynamic range 0-250 of the scaled image has been applied to the shallow water column corrected bottom signatures, TOA
    • details are obscured over bright areas
  • RIGHT: the dynamic range 0-250 of the scaled image has been halved to the shallow water column corrected bottom signatures, TOA
    • this reveals fine details over bright areas
 

 



Seatruth

ZLIDAR - Z4SM
  • RED:    underestimated depth retrievals
  • BLUE: overestimated    depth retrievals
  • LIGHT tones: whether blue or red, signal more acceptable depth retrievals

A tolal failure?






Some nice seatruth regressions from 0 to 25 m

and some not so nice
NO SMOOTHING
This proves that the simplified RTE is trustworthy
the way it is operated in 4SM


ROI 2
at 15-25 m

this is over the western deep outer ridge

 


ROI 4
at 2-7 m
this is over the western part
of La Parguera Marine Reserve,
where W. Hernandez calibrated his work

 

ROI 1
at 4-13 m



ROI 3
at 3-16 m


ROI 5
over La Parguera Marine Reserve


ROI 6






ZLIDAR

TCC water column corrected
LIDAR depths have been applied/forced here




ZLIDAR - Z4SM
RED is where
4SM underestimated the bottom depth

 
LIDAR depths have been applied/forced here
  • this allows us to visualize where water column correction forces unrealistic bottom signatures
  • because/where the water leaving reflectance is significantly higher than assumed to be , then water column corrected signatures arespiked to the extent of the exponential of this artifact
  • what we see here is that the GREEN water leaving reflectance dominates 
    • therefore, we may conclude that chlorophyll content is the main source of our problems
    • as reported abundantly in the litterature on SouthWest Puerto Rico.
  • where waters are true to the assumed OII water type, hues are dark and dull in this TCC rendition
  • where chlorophyll artifact is strongest, hues are brightest in this TCC rendition
    • note that this mimics the ZLIDAR-Z4SM image