Landsat 5073x5123 pixels TM_20090110
at Geraldton, Western Australia

back to the Geraldton page
 
 
 


TM_20090110
full image
processed Dec 17 2011


This exceptional image extends over 150 km from the Abrolhos Islands to the south of Dongara
It gives some precious insight into the "glint" problem
It also hints at a possible explanation for the effective wavelength problem with Landsat images

NIR band raw
Notice the orderly increase
of the NIR path radiance from 8_NW to 16_SE.
This does not appear to caused by adjacency effect

BLUE band deglinted
The blue band exhibits
heterogeneous deep water radiances.
This probably results from variable content in Chlorophyll
Pink is 68
Green is 63
 
<<== Profile_yellow : deglinting is OK
The increase of path radiance from NW to SE
is clearly visible
in the NIR, Red and Green bands,



while the Blue band
does not exhibit such an orderly pattern

 

GREEN band deglinted
The green band is remarkably homogeneous,
except along the coast
where long shore currents cause turbidity
Light blue is 20
Light green is 19

Some inconsistencies remain though

RED band deglinted
The red band is remarkably homogeneous,
except along the coast
where long shore currents cause turbidity
Green is 13
Light green is 14

Light blue is 15
Some inconsistencies remain though

TCC raw
  •  Lsw
  • Glint
  • profile_yellow
  • profile_4
  • profile_purple

TCC deglinted
 

Path radiance Glint regressions
  • ABOVE this plot records the general increase of path radiance from West to East by 9 DNs along the blue shapefile
 
  • BELOW this plot records the swell modulated surface glint is comprised of ~4 DNs at most in the subset to the right
Swell modulated Glint regressions


==> Swell and path glints have
very similar spectral composition

Deglinting must correct for
  • the general pattern of the path radiance
  • the local     pattern of the swell modulated sea-surface clutter


Seatruth
along the coast,
using a depth profile provided by R. Chiles


Coast
Plot_ZZ   and Regress_ZZ
cWL[green]=0.45
Kikj=0.47
RLBgb1.300_0.500
This applies along the coast


 



Z2 vs Z3
over the islands






Islands
Plot_Z23   and Regress_Z23
cWL[green]=0.58

Kikj=0.47
RLBgb1.300_0.900
Islands must be processed separately


 

 
Conclusion 1
  • the optical properties are stable
    over this image, as KiKj=0.47 suits
    both the islands and the coast
Conclusion 2
the differences are two-fold
  • cWL[green] : the effective wavelength for the Green band depends on path Radiance properties
  • RLBgb : bottoms are distinctly greener along the coast that at the islands
Coast
nice -20  4SM.4.08  -Process/30_OCT_2011  -Origin/FUGRO-NPA               
-DB/TM_20090110/31_1_0_0/4_1/5073_5123/158.415_-3127.245/1_1                          
-Mis/Australia/Geraldton/Landsat/TM/FUGRO-NPA/UTM_50_008/0.030_-0.030/10_JAN_2009   
-LS/0255.0/255.0/255.0/255.0           
-cWL/0.500/0.450/0.500/0.500       
-M/@000001/00002/00003/00004           
-LsM/232.6/137.5/181.1/178.4           
-Lsw/62.68/17.71/12.09/06.00           
-dLsw000.0/000.0/000.0/000.0           
-Lw/0013.6/001.5/000.0/000.0_AutoCAL       
-Lm/0000.5/000.4/003.0/255.0           
-KK1_2_0.4700/Knir4.500/mask_3           
-Z/LAT0.40/N_1/RLBgb1.300_0.500/cZ1  2.9   
-B/tclNe/Bmin1/cLM1              
-deglint/vRbaD/F2/Q1/L4/GlintM25.0       
-extract/v/rawBDH/fullBDH/mBPL2/mask_3       
-calibrate/v/BDSNpZg/BDh_12_13_14       
-Model/mask_3                   
-Smooth/1/D/Smart+               

 
Islands
nice -20  4SM.4.08  -Process/30_OCT_2011  -Origin/FUGRO-NPA               
-DB/TM_20090110/31_1_0_0/4_1/5073_5123/158.415_-3127.245/1_1                          
-Mis/Australia/Geraldton/Landsat/TM/FUGRO-NPA/UTM_50_008/0.030_-0.030/10_JAN_2009   
-LS/0255.0/255.0/255.0/255.0           
-cWL/0.500/0.580/0.500/0.500       
-M/@000001/00002/00003/00004           
-LsM/232.6/137.5/181.1/178.4           
-Lsw/62.68/17.71/12.09/06.00           
-dLsw000.0/000.0/000.0/000.0           
-Lw/0013.6/001.5/000.0/000.0_AutoCAL       
-Lm/0000.5/000.4/003.0/255.0           
-KK1_2_0.4700/Knir4.500/mask_3           
-Z/LAT0.40/N_1/RLBgb1.300_0.900/cZ1_islands   
-B/tclNe/Bmin1/cLM1                
-deglint/vRbaD/F2/Q1/L4/GlintM25.0       
-extract/v/rawBDH/fullBDH/mBPL2/mask_3       
-calibrate/v/BDSDpZG/BDh_12_13_14       
-Model/mask_3                   
-Smooth/1/D/Smart+               

 
geraldton_TM_20090110_full_autocalibration
AutoCalibration on the full image
all raw pixels in the BDH

 
geraldton_TM_20090110_calibration_full
Calibration on the full image
deglinted marine pixels only in the BDH

 










 



Créer un site
Créer un site