Negril85 tutorial
Prepare your image of Negril Shores

return to Negril tutorial        next to Calibration 1
 
 

 

 

Download and unzip 4sm tutorial TM negril85.zip
with the following content

  • yourimage.tif                    contains Landsat wavebands TM1 to TM4
  • negril85.mcdd                   contains some EASI commands
  • negril85_mSE.005.bkp    contains a backup copy of the mask mSE
  • negril85.sh                         contains the commandline script
  • negril85XYZoriginal.txt    contains the seatruth data, comprised of three sounding lines
  • negril85XYZmodified.txt  contains the seatruth data: some points are relocated
  • 4 Shape Files                      prepared with OpenEV, used to control -Recode operations
Database structure is dbnc=27 ,1,0,0

Channel and pseudocolor pallet description may be found in textfile negril85.mcd



We want to
  • create the database,
  •  import the TM bands,
  •  create the mask mSE in channel 5
 

 


OPENEV users

  • Now, make sure your script has the following lines
nice -20 4SM.2.06 -PROCESS/06_JUL_2008/MASK_0                    
-dB/PCI/negril85/31_1_0_0/4_1/300_500/104.3618125_192.214125/1_1    
-Origin/U_of_Upsala                            
-Mis/Jamaica/Negril/TM/TM/NA/UTM_19_008/.030_.030/01_JAN_85          

@-URL/file:///Full/Path/Of/Your/Own/URL           
-makePIX/FILE/8/Z20                            
-import/yourimage/dbnc_4_0_0_0/300_500/1,4_1,4                 
-recode/0_InChannel_1_into_239_OutChannel_5NoData            
-recode/0_InChannel_2_into_239_OutChannel_5NoData            
-recode/0_InChannel_3_into_239_OutChannel_5NoData            
-recode/0_InChannel_4_into_239_OutChannel_5NoData            
-recode/0_InChannel_5_into_2_OutChannel_5_Data_Pixels            
-makeLand/4.0/1.5/showLand                        
-recode/2_InChannel_5_into_248_OutChannel_5/land.shp           
-Recode/2_InChannel_5_into_240_OutChannel_5/deep.shp           
-recode/248_InChannel_5_into_21_OutChannel_5/mSOIL.shp           
-recode/2_InChannel_5_into_1_OutChannel_5/shallow.shp           
@-DN69/174_DeepWater @-DN116/381/2_BrightLand

  • Please run the command line script   negril85.sh    -in the following order-   to
    1. -MakePIX: create the database negril85.pix
    2. -Import: import the four TM channels from   yourimage.pix into the database negril85.pix
    3. -Recode: map a few weird NoData pixels to 239             in channel 5
    4. -Recode: map all other pixels to 2                                    in channel 5
    5. -MakeLand: code land areas to 248                                 in channel 5
    6. -Recode: clean the land areas                                          in channel 5
    7. -Recode: code some mSOIL at 21                                    in channel 5
    8. -Recode: code deep waters at 240                                    in channel 5
    9. -Recode: code some shallow areas to 1                            in channel 5

PCI users

  • Learn about MaskSE
  • Deep : start with isolating optically deep waters under mSE=240.
  • Shallow : go on with separating between shallow water under mSE=2: use the "seed" tool.
    • Code_2 for general shallow waters
    • Code_1 if desired for specific shallow areas.
  • Dryland : go on with filling dry land at mSE=248
  • mSOIL : go on with identifying bare soil pixels at mSE=21, from the brightest to the darkest, in view of selection of the Soil Line.
  • Alternately, you may copy file negril85.005.bkp onto negril85.005

provisional Lsw and La
  • Run the script with the following enabled  Syntax DN
    • -DN69/174_DeepWater: get a provisional estimate of average optically deep water radiance:  
      • ==> -Lsw066.5/017.4/011.3/005.5_DN69_174_DeepWater
        • ==> -La066.5/017.4/011.3/005.5 only valid for the Red and NIR bands

provisional LsM
  • Run the script with the following enabled
    • -DN116/381/2_BrightLand: get a provisional estimate of  average bright bare land radiance: 
      • ==> -LsM104.0/048.5/058.2/074.5_DN116_381_2_BrightLand 
      • both La and LsM values shall be fine-tuned upon calibration

We now have estimated a provisional Soil Line for the pair Red/Nir
  • Segmentation of the image: in a bi-dimensional histogram of Red versus Nir, 
    • the Soil Line is a straight line with a slope=(LsM[3]-La[3])/(LsM[4]-La[4])
      • which starts at point {11.3;5.5}, the provisional atmospheric path radiance La
      • and runs through point {58.2;074.5}, the provisional LsM
    • all bare land pixels are displayed along the Soil Line, inside a cone bounded by slope/CSL and slope*CSL: see Syntax Z
      • all pixels inside this cone shall be used upon extraction to estimate the Soil Line by NIR
    • all shallow pixels are displayed on the left hand side
    • all vegetated pixels are displayed on the wright hand side

 





Créer un site
Créer un site