A 4903*5463 Landsat-ETM 4 bands 8-bits image of Salar de Uyuni, Bolivia, dated April 3rd, 2001
Image L71233074_07420010403, courtesy of USGS_Glovis; this image subset is 147 km across , pixel size 30 m

Salar de Uyuni is a slat flat located in the Atacama desert in Bolivia, at an altitude of ~3700 m.
Because depths hardly exceed ~ one meter, the deep water radiances may not be estimated for the visible bands.
This is not too much of a problem though, as the Lw term is mostly important at depth over dark bottoms,
and cloud shadows allow for a rough estimate of atmospheric path radiance.
No-deglinting was required; some radiometric glitches though in the ETM image

A 4903*5463 Landsat-ETM 4 bands 8-bits image, dated April 3rd, 2001

Raw image, true color composite

  • Salar de Coipasa is located at the upper left corner of that scene

Raw image, false color composite

  • This image was taken shortly after a major input of water, as the whole salar is covered with water

Bottom reflectance, true color composite

  • The scene is seen "as if there was no water"
  • The bottom of the salar is said to be lined with a layer of pink algae (Wikipedia) on which flamingoes feed.

  • The darkening of the bottoms is likely to be related to
    • river discharge, bringing nutrients and sediments
    • and subsequent development of a pink algal mat, high in red, low in blue
  • rather thanthe effect of depth or water stratification


Bottom reflectance, normalized color composite

  • This clearly shows that the bottoms are lined with a reddish algal mat
  • Artifact: where radiances are saturated, the computed bottom signature becomes greenish instead: this is an artifact

Computed depth in decimeters

see also computed depth in centimeters

  • All depths are computed using the NIR band, with 2K[nir]= ~ 4.3 m-1
  • Therefore, the estimation of operational wavelengths and optical water type is of secondary importance
  • Depth over saturated pixels is still acceptable, but model inversion yields a faulty greenish signature!

Average bottom brightness
  • Salar de Coipasa is located in the NW corner of this image
  • in April2001: a wet situation
    • uniform and very dark bottom is most unlikely
    • absence of a moat in the depth image is most unlikely
    • waters probably are turbid
  • in June2002: a dry situation
    • image Z in centimeters: the remnant moat is distinctly present
    • image B: the algal mat has been scorched to white

Calibration diagram for the full image

  • Optical water type ~OII+0.6 in a high water situation
  • Wavelengths used are 477.6, 588.5, 661.3, 834.8 nm
  • Calibration appears to be quite reliable, in spite of the fact that the water is probably a saturated brine
  • Depths do not seem to exceed 1.2 m
  • Saturation: the radiances are badly saturated: this does not help!
  • Wikipedia for Salar de Uyuni: the interpretation of this weird plot relies on the knowledge that
    • the bottoms are lined with a pink algal mat
    • the water depth in the moat hardly exceeds 1 m

Calibration diagram of linearized data

  • This plot shows how the path radiance is tentatively specified using
    • Lsw[nir]=La[nir]
    • the Soil Line for all pairs of bands
  • Once the spectral path radiance is estimated, then common sense is applied to specify the deep water radiance for all bands



#\rm *cal; nice -20  4SM.3.21 -Process/26_JAN_2010    -OriginUSGS_Glovis                \
-dB/L71233074_07420010403/31_1_0_0/4_1/4903_5463/572.925_-2129.505/1_1            \
-Mis/Andes/Salar_de_Uyuni/Landsat/ETM/USGS/UTM_19_008/0.030_0.030/03_APR_2001        \
@URL/file:///home/rich/RSPwebsite/RSP.4SM                        \
-MakePIX                                        \
@import/data*L71233074_07420010403/dbnc_1_0_0_0/R8221_L7121/import_543.000_-2129.400     \
-LS0255.00/255.00/255.00/255.00                                \
-Lm00001.0/0001.0/0001.0/0001.0                                \
-LsM/052.7/294.4/350.7/178.2_CoefLM=0.80    \
-Lsw0052.0/032.0/020.0/010.0_1.036        \
-La/0042.0/029.0/020.0/010.0_1.036        \
-AutoCAL/print_2_3/MapBPL/MakeNewMask/getGlint/glintM42.0/Clouds /Ext_600_1200_1_2540


#Following is the full CommandLine=====mBPL=2==================================
#\rm *cal
nice -20  4SM.3.21 -Process/26_JAN_2010    -OriginUSGS_Glovis            \
-DB/L71233074_07420010403/31_1_0_0/4_1/4903_5463/572.925_-2129.505/1_1        \
-Mis/Andes/Salar_de_Uyuni/Landsat/ETM/USGS/UTM_19_008/0.030_0.030/3_APR_2001    \
@-URL/file:///home/ymorel/RSPwebsite/website/                    \
-makePIX/FILE/0/Z20                                \
-LS0255.0/255.0/255.0/255.0        \
-WLM/513.9/600.7/692.0/898.0        \
-WLm/441.3/519.4/630.6/771.6        \
-cWL/0.500/0.820/0.500/0.500        \
-cWL/0.500/0.850/0.500/0.500        \
-M/@0001/@0002/@0003/00004        \
@CP/238.1/209.8/145.3/11.9_1.00m    \
-CP/250.1/204.8/145.3/11.9_1.00m    \
-LsM/294.7/294.4/350.7/178.2_CoefLM=0.80    \
-Lsw0052.0/032.0/020.0/010.0_1.036        \
-La/0042.0/029.0/020.0/010.0_1.036        \
-Lsw0052.0/032.0/020.0/010.0_1.036        \
-La/0034.0/023.0/020.0/010.0_1.036        \
@dLsw00000/00000/00000/-0.40            \
-Lm00001.0/001.0/002.0/000.5            \
-KK2_3_0.60/Knir4.500/Bm400_BG500_GR600_RN700                    \
-KK1_3_0.40/Knir4.500/Bm400_BG500_GR600_RN700                    \
-Z/MSL0.00/zplus/cSL1.18/ZMax0.00/cdZ1.00/PAN_1_4/sIG_0w0/N_1/cZ1.000        \
-B/tclNe5.00/LBref200_100/Bmin0/cLM1.0                        \
@deglint/vRba/DegTol1.00/F1/L4/GlintM25.0/FN/mDEGLINT1                \
-extract/v/NIRband4/NIRmax255.0/mBPL2/mSOIL21/Ext_1_4903_1_5463            \
-Calibrate/v/BdSNpz/BDh_12_13_34                        \
-Model/mask_3    @-Tc1/2                                \
-Smooth/1/D/smart+                                \
-dCAL/A/1_3/16    @-E1/4903/1/5463 @-E1/4903/1/5463                \
-Recode/3_InChannel_5_into_2_OutChannel_5                    \


-import/data*L71233074_07420010403/dbnc_1_0_0_0/R8221_L7121/chIn1,4/chOut1,4/dbiw_997_3_4903_5463    \
-Prepare/scale/delimb/fill/smooth/linearize                    \
-Classify_3.0_0F/mode_5_11_16                            \
-Classify_3.0_0F/Mode_5_11_16                            \
-RegressZZv/L71233074_07420010403XYZ_GPS/0.001_0_0/1.000_0.000/ZM/2_4/0_201/0_30/ALL        \
-RegressZZV/L71233074_07420010403XYZ/0.001_0_0/1.000_0.000/ZM/2_4/0_201/0_30/*            \
-ProfileZZv/L71233074_07420010403/0.001_0_0/1_0                            \
-ProfileABZ/profile_1/442_4427/1610_4634/1_3_9_11                \
-Recode/2_InChannel_5_into_1_OutChannel_5/mask_1                \
-Recode/2_InChannel_5_into_3_OutChannel_5/mask_3                \
-Recode/2_InChannel_5_into_240_OutChannel_5/deep                \
-Recode/2_InChannel_5_into_248_OutChannel_5/land                \
-Recode/248_InChannel_5_into_21_OutChannel_5/mSOIL                \
-Recode/GT25_InChannel_4_into_234_OutChannel_5/NameShapeFile            \
-Recode/LT4_InChannel_4_into_249_OutChannel_5                    \
-Prepare/scale/delimb/fill/Smooth/linearize                    \
-Import/GetImportKey_65814@42187860852760994610_gezirat/apr4casi/dbnc_0_0_12_0/5363_6566/1,12_65,76    \
-Archive                                \
-DD480/2756BrightShallow                                \
-DN250/1387_2BrightLand                                \
-DDD144/148/113/126_7                                \

kghostview -scale 0.6667 -landscape L71233074_07420010403_4_3_2_1.eps &

Modify Channel Descriptor


cm01=" 1  1_Raw_478 nm";                    dboc=1  ; r mcd
cm01=" 2  2_Raw_575 nm";                    dboc=2  ; r mcd
cm01=" 3  3_Raw_661 nm";                    dboc=3  ; r mcd
cm01=" 4  4_Raw_835 nm";                    dboc=4  ; r mcd
cm01=" 5 Special_Effect_mask; use pctmSE";            dboc=5  ; r mcd
cm01=" 6 wZ: use pctwZ";                    dboc=6  ; r mcd
cm01=" 7 wR: use pctwR";                    dboc=7  ; r mcd
cm01=" 8 Z0->250 in dm: use pctZ250";                dboc=8  ; r mcd
cm01=" 9 Z: use pctZ20";                    dboc=9  ; r mcd
cm01="10 B: use pctB";                        dboc=10 ; r mcd
cm01="11 Bottom Typing: use pctClassified";                    dboc=11 ; r mcd
cm01="12 LSBCC_blue ";                        dboc=12 ; r mcd
cm01="13 LSBCC_green";                        dboc=13 ; r mcd
cm01="14 LSBCC_red  ";                        dboc=14 ; r mcd
cm01="15 dZED: use pctZ250";                    dboc=15 ; r mcd
cm01="16 void";                            dboc=16 ; r mcd
cm01="17 void";                            dboc=17 ; r mcd
cm01="18 void";                            dboc=18 ; r mcd
cm01="19 void";                            dboc=19 ; r mcd

cm01="20  1_Linearized_478 nm";                    dboc=20 ; r mcd
cm01="21  2_Linearized_575 nm";                    dboc=21 ; r mcd
cm01="22  3_Linearized_661 nm";                    dboc=22 ; r mcd
cm01="23  4_Linearized_835 nm";                    dboc=23 ; r mcd

cm01="24  1_Ldeglinted_478 nm";                    dboc=24 ; r mcd
cm01="25  2_Ldeglinted_575 nm";                    dboc=25 ; r mcd
cm01="26  3_Ldeglinted_661 nm";                    dboc=26 ; r mcd
cm01="27  4_Ldeglinted_835 nm";                    dboc=27 ; r mcd

cm01="28  1_LSB_478 nm";                    dboc=28 ; r mcd
cm01="29  2_LSB_575 nm";                    dboc=29 ; r mcd
cm01="30  3_LSB_661 nm";                    dboc=30 ; r mcd
cm01="31  4_LSB_835 nm";                    dboc=31 ; r mcd

cm01="32_ZG: Zshallow in centimeters, others at -1";        dboc=32 ; r mcd

Créer un site
Créer un site