4SM TM tmnov-tutorial-model1-2

As the AutoCAL process succeeded,
you may go directly to Calibration 1
or
do a few things here to learn from your image

Things you can do

Order of precedence among 4SM module arguments

-MakePIX
-Import
-Recode
-DN
-DD
-Deglint
-Extract
-Calibrate
-ProfileAB
-ProfileZZ
-RegressZZ
-Classify
-Model

#Following is the simplified AutoCal CommandLine=====mBPL=2 ======on Sat Nov 25 14:06:38 2017

nice -20 ./4SM.8.02 -Process/25_Nov_2017 -Origin/FUGRO-NPA @@

-DB/tmnov/31_3s_0s_0s/4_1/1265_1560/208.350_2398.770 @@

-Mis/Bahamas/Caicos/Landsat/TM/FUGRO-NPA/bOA/UTM_+19_008/0.0300_0.0300/22_NOV_1990 @@

@WLM/518.0/610.0/693.0/905.0 @@

@WLm/452.0/529.0/624.0/776.0 @@

@WL/0485.0/569.5/658.5/840.5 @@

-Lsw/060.5/017.4/011.1/006.9_Lsw_AutoCal @@

@dLsw/0-1.3/00-0.5/00-0.3/00-0.1_dLsw @@

-dLsw/000.0/000.0/000.0/000.0 @@

-Lw/0010.0/001.5/000.0/000.0_Lw_AutoCal @@

-M/@000001/00002/00003/00004_M @@

-Lm/0001.0/001.0/001.0/255.0_Lm @@

-LsM/203.6/112.8/136.1/111.3_LsM_AutoCal @@

@CP/192.58/99.02/87.90/10.43_0.75m @@

-KK1_2_0.5141/Knir4.5/CoefK_1.0/mask_3 @@

-Z/NA0.00d/N_1/nDR_34_1000_zDTM/mask_3 @@

-B/tclNe/cNorm2.50/Bmin0/cLM1.00 @@

-deglint/vRbaD/GlintM25 255 @@

@AutoCAL/R2_0.95/Land_-5.0_1.18_4s/GlintM25 @@

-extract/v/rawBDH/FullBDH/NIRband4/NIRmax255.0/mBPL2 /mask_3 @@

-calibrate/voAI/BDSNpZG/BDh_12_13_34 @@

-Model/v/rt1.00/dTM/ZM30/mask_3 @@

-smooth/5/D/Smart+ -LL10 @@

@-DN300/100/3/comment @-DDshp @-DNshp @-Eshp @-E/1/1265/1/1560 @@

-ProfileAB/LzrvWBc/profile_1/chAB_1_2_3_4 @@

@ProfileAB/lZrvWBc/profile_1/chAB_1_2_3 @@

@ProfileAB/lzRvWbC/profile_3/dTM/WZ_2_3/B_0_201/ZC_0.0_8_1_1/* @@

@RegressZZ/vwpC/seatruth*depth_points/CXY_1_0_0/CR_1.0_0.0/ZM/WZ_2_3/B_0_201/ZC_0_15_1_1/ALL @@

exit

Note: @disabled lines in italic may be removed

Things you can do:
run profiles to get a feeling of how things work
Profiles
White profile_1.shp Blue profile_2.shp Green profile_3.shp Red profile_4.shp
Syntax ProfileAB for PlotLS -ProfileAB/LzrvWBc/profile_1/chAB_1_2_3_4 for PlotZB -ProfileAB/lZrvWBc/profile_1/chAB_1_2_3 for PlotZ23 -ProfileAB/lzRvWbc/profile_3/dTM/WZ_2_3/B_0_201/ZC_1.5_7_1_1/*

Profile 1 PlotLS
-ProfileAB/LzrvWBc/profile_1
check on Lsw for the NIR band
Please run your script with the following: -deglint/.........../GlintM25 -Model/v/..... -smooth/5..... -ProfileAB/LzrvWBc/profile_1/chAB_1_2_3_4	........................Conclusions.......................... Glintm=Lswnir=6.9 is convenient GlintM=25 should be reduced as required
Profile_1 deglinting the raw data	Deglinting range Glintm=Lswnir=6.3; GlintM=25 .......................PlotLS along Profile 1-B...................... 0-11 km Glintm=Lswnir=6.9 is convenient hardly any glint clutter is observed over optically deep waters ......................PlotLS along Profile 1-A...................... 0-1.5 km the NIR radiance exceeds GlintM=25 no deglinting is applied over this very shallow intertidal mud flat 1.5-9.0 km the amount of "glint" decreases afterwards this is not acceptable, unless we suspect a trick of the "Adjacency effect" quite unlikely in this environment GlintM=25 should be reduced to Lswnir+~5 9.0-35.0 km the amount of glint over Glintm=6.9 is hardly worth noting Glintm=Lswnir=6.9 is convenient Lsred decreases to ~2 DNs over deep water radiance Lswred~=11.1 35.0-41.0 km the profile goes through three very shallow ridges, which are unduely "deglinted" GlintM=25 should be reduced

Profile 1 PlotZB
-ProfileAB/lZrvWBc/profile_1
Please run your script with the following -Model/mask_2 -Smooth/5/D/Smart+ -ProfileAB/lZrvWBc/profile_1/chAB_1_2_3	..................Conclusion.................. Increase CoefLM from -B/.../cLM1.00 to -B/.../cLM1.10 then run the Calibration There is some potential for bottom typing
Profile_1 depth and spectral water-column corrected reflectance	In the 3-band case (Red solution), Z is increased until the ratio LBblue/LBred equals 1 this is because BOA radiances are normalized to the range 0-200 please refer to Morel & Favoretto, 2017 Greenish: it turns out that most bottoms in this profile exhibit a distinctly greenish spectral bottom signature this is quite a secure statement wherever Z was computed using the Red solution this site is a feeding ground for turtles and dugongs 14-20 km some pixels are modeled by the Green solution as water depth reaches 5.5 m please disable -smooth....... this reveals even more Green solutions, as the Red solution reaches its limits at Lmred=1.0 increase/decrease Lmred, see what happens -Lm/0001.0/001.0/000.5/255.0 -Lm/0001.0/001.0/002.0/255.0 25-33 km water column corrected reflectances are saturated CoefLM1.10 shall correct for that try it now: run -Calibrate... with the following -B... argument: -B/..................../cLM1.10 see Syntax_B
Now, please try -Smooth/5/D/Smart+ try .../Profile_2/... try .../Profile_3/... try .../Profile_4....	Too many water column corrected signatures are saturated, especially in Profile_3: use -B/...../CoefLM1.10 in to increase to -LsM/218.9/122.5/148.6/121.7_LsM_cLM=1.100 this does not affect the retrieved depths

Profile 3 PlotZ23
-ProfileAB/lzRvWbc/profile_3/dTM/WZ_2_3/B_0_201/ZC_1.5_7_1_1/*
Z2 vs Z3
This tool compares two solutions, like Green solution versus Red solution Z2 vs Z3 After all, we wish these two solutions to yield nearly same retrieved depths The answer is: most Green solution depths are underestimated by ~0.6 m	One work around is to increase LsMgreen manually from -LsM/218.9/122.5/148.6/121.7 to -LsM/218.9/130.5/148.6/121.7 as illustrated on the left plot The plot below also reveals that depth under-estimation in the 2-4 depth range affects darker bottoms more while bright bottoms behave nicely
Z2 vs Z3 along Profile_3 -LsM/218.9/122.5/148.6/121.7	Z2 vs Z3 along Profile_3 -LsM/218.9/130.5/148.6/121.7

**Profile 3** PlotZ23
-ProfileAB/lzRvWbc/profile_3/dTM/WZ_2_3/B_0_201/ZC_1.5_7_1_1/*

-**LsM/218.9/130.5/148.6/121.7**
PlotZB: Z and spectral LB still some saturation from waypoints A1 to A2: -LsM... should be increased further	PlotZ23: Z2 vs Z3
Please try that now, on all four profiles Please check how the calibration diagram is affected Please check how -DD926/340 is affected: the Red solution is still retrieving a depth=0.14 m but it produces a highly biased spectral watercolumn corrected signature: this us not acceptable, as this is a BPL pixel	..................Conclusion.................. I quite like this last finding about raising LsMgreen to 130.5 After all, first improve/consolidate the output then eventually come up with some acceptable and generic explanation This solution is not acceptable, though, as it violates the BPL assumption which assumes that the water column corrected BPL bottom reflectances should be spectrally neutral

**Profile 3** PlotZ23
-ProfileAB/lzRvWbc/profile_3/dTM/WZ_2_3/B_0_201/ZC_1.5_7_1_1/*

**-LsM/218.9/122.5/148.6/121.7**
**-Model/v/Rt1.0/dTM/ZM30**
PlotZ23: Z2 vs Z3	PlotZ23: Z2 vs Z3
-Model/v/Rt1.0/dTM/ZM30 Syntax_Model This consists in applying to the Green (or PAN) solution for this scene the Blue/Green ratio of bottom signatures which have been observed to prevail for the Red solution over the Bahamas Bank: please refer to Landsat 8 at Lee Stocking Island. Please check how -DD926/340 is affected: the Red solution is still retrieving a depth=0.14 m, and the BPL assumption is verified.	..................Conclusion.................. The .../Rt1.0/... specification works well on this Landsat TM scene of Caicos Bank. It was used successfully on Landsat 8 scenes at San Lorenzo Channel

Conclusions
Glintm=Lswnir=6.9 is convenient GlintM=25 should be reduced as required Increase CoefLM from -B/.../cLM1.00 to -B/.../cLM1.10 or even more Raising LsMgreen is not acceptable, but using the UseRatio look-up table shows promise There is some potential here for bottom typing
Here is your script at this stage #Following is the simplified AutoCal CommandLine=====mBPL=2 ======on Wed Nov 29 14:41:27 2017 nice -20 ./4SM.8.02 -Process/29_Nov_2017 -Origin/FUGRO-NPA @@ -DB/tmnov/31_3s_0s_0s/4_1/1265_1560/208.350_2398.770 @@ -Mis/Bahamas/Caicos/Landsat/TM/FUGRO-NPA/bOA/UTM_+19_008/0.0300_0.0300/22_NOV_1990 @@ -Lsw/060.5/017.4/011.1/006.9_Lsw_AutoCal @@ @dLsw/0-1.3/00-0.5/00-0.3/00-0.1_dLsw @@ -dLsw/000.0/000.0/000.0/000.0 @@ -Lw/0010.0/001.5/000.0/000.0_Lw_AutoCal @@ -M/@000001/00002/00003/00004_M @@ -Lm/0001.0/001.0/001.0/255.0_Lm @@ -LsM/203.6/112.8/136.1/111.3_LsM_AutoCal @@ -LsM/218.9/122.5/148.6/121.7_LsM_cLM=1.100 @@ -KK1_2_0.5141/Knir4.5/CoefK_1.0/mask_3 @@ -Z/NA0.00d/N_1/nDR_34_1000_zDTM/mask_3 @@ -B/tclNe/cNorm2.50/Bmin0/cLM 1.10 @@ -deglint/vRbaD/GlintM25 255 @@ -extract/v/rawBDH/FullBDH/NIRband4/NIRmax255.0/mBPL2 /mask_3 @@ -calibrate/voAI/BDSNpZG/BDh_12_13_34 @@ -Model/v/Rt1.0/dTM/ZM30/mask_3 @@ -smooth/5/D/Smart+ -LL10 @@ @-DN300/100/3/comment @-DDshp @-DNshp @-Eshp @-E/1/1265/1/1560 @@ -ProfileAB/LzrvWBc/profile_1/chAB_1_2_3_4 @@ -ProfileAB/lZrvWBc/profile_1/chAB_1_2_3 @@ -ProfileAB/lzRvWbC/profile_3/dTM/WZ_2_3/B_0_201/ZC_0.0_8_1_1/* @@ @RegressZZ/vwpC/seatruthdepth_points/CXY_1_0_0/CR_1.0_0.0/ZM/WZ_2_3/B_0_201/ZC_0_15_1_1/ALL @@ @-DD300/500 @ -DD926/340* exit