Difference between revisions of "STATSGO2-soil data"

From Snow
Jump to navigationJump to search
m
 
(37 intermediate revisions by the same user not shown)
Line 10: Line 10:
 
State data is spatially coarser than county or survey level data
 
State data is spatially coarser than county or survey level data
  
−
== Step 2: Soil Data Viewer ==
+
SURRGO format information [http://soildatamart.nrcs.usda.gov/SSURGOMetadata.aspx]
 +
 
 +
== Step 2: Install Soil Data Viewer ==
 
You need
 
You need
 
1) Install ArcGIS 9.2 or later for soil data viewer 5.2 on a Windows XP system
 
1) Install ArcGIS 9.2 or later for soil data viewer 5.2 on a Windows XP system
 
2) Install soil data viewer version 5.2 [http://soildataviewer.nrcs.usda.gov/download.aspx]
 
2) Install soil data viewer version 5.2 [http://soildataviewer.nrcs.usda.gov/download.aspx]
 +
 
== Step 3: Prepare Access database ==
 
== Step 3: Prepare Access database ==
 
1) Open Access template
 
1) Open Access template
Line 28: Line 31:
 
[http://soildataviewer.nrcs.usda.gov/userguide.aspx]
 
[http://soildataviewer.nrcs.usda.gov/userguide.aspx]
  
−
== Step 5: Karl's method for estimating soil properties using SDV ==
+
Note: Good practice to export layer because ArcMap won't remember them when your re-open your map
 +
 
 +
== Step 5: An Example of estimating soil properties using SDV in the Silver Fork of the American ==
 
Definitions:
 
Definitions:
  
−
a) Dominant Condition (DCD)
+
'''a) Dominant Condition (DCD)'''
 +
 
 +
'''b) Dominant Component (DCP)'''
 +
 
 +
'''c) Weighted Average (WA)'''
 +
 
 +
'''d) All Layers (ACP)'''
 +
===1) Soil Physical Properties: Surface Texture===
  
−
b) Dominant Component (DCP)
+
Useful for a) naming the soil .def file b) understanding depth to soil restrictive layer c) understanding ksat at surface
  
−
c) Weighted Average (WA)
+
===2) Soil Qualities and Features: Depth to Any Restrictive Layer (RHESSys requires soil depth)===
  
−
d) All Layers (ACP)
+
a) Generate Depth DCD, DCP, WA, ACP using both < and > for tie break rule with no cutoff (8 scenarios)
  
 +
b) Compare these to summer Landsat imagery and use expert judgment to determine the most accurate layer
  
−
'''1) Soil Physical Properties: Surface Texture'''
+
c)  Depth To Any Restrictive Layer results for Silver Fork
−
Useful for a) naming the soil .def file b) understanding depth to soil restrictive layer c) ksat at surface
+
*ACP, > 127,143,153 - not much spatial variability, bedrock not represented
 +
*ACP, < 0 - basically junk
 +
*'''WA - tie break doesn't matter, 17 for texture un-weathered bedrock'''
 +
*DCP - tie break doesn't matter, very shallow in upper basin
 +
*'''DCD, > zeros where surface texture is bedrock, 77,127,201'''
 +
*'''DCD, < zeros where surface texture is bedrock, 77,127,201 but shallower than DCD, > (obviously)'''
  
−
'''2) Soil Qualities and Features: Depth To Selected Soil Restrictive Layer'''
+
===3) Create a Raster that matches the shapefile ===
 +
a) To Raster > Polygon to Raster (used Ksat's FID). Use MAXIMUM_AREA, set env. variables to size of dem
 +
b) Reclass - to get rid of No Date (map to 0)
  
−
a) Generate Depth DCD, DCP, WA, ACP using both < and > for tie break rule with no cutoff (8 scenarios)
+
===4) Soil Physical Properties: Saturated Hydraulic Conductivity (RHESSys requires Ksat at surface and decay with depth)===
 +
a) Generate Depth WA  for '''ALL LAYERS-ACP,slow''' and for '''Depth Range-WA,slow''' using both slowest and fastest for tie break rule with no cutoff (8 scenarios)
  
−
b) Compare these to summer Landsat imagery and use expert judgment to determine the most believable layer
+
b) '''All Layers Ksat''' results for American Silver Fork
 +
*DCD - tie break doesn't matter, portions with surface texture bedrock have no data. Range is 91 um/s, 29 um/s 18 um/s
 +
*DCP - tie break doesn't matter, large portions of upper basin have no data. Range is 91 um/s to 28 um/s for lower
 +
*WA - tie break doesn't matter, Ksat is 60 um/s in upper basin and 26 um/s in lower basin. More spatial variability than ACP >
 +
*ACP < - Ksat is too slow
 +
*ACP > - Ksat is 91 um/s in upper basin and 28 um/s in lower basin. Less spatial variability than WA
  
−
c) results for Silver Fork: Tie break doesn't matter for all but ACP.
+
c) '''50 cm layer increment - WA, slow''' results for American Silver Fork (note 50cm increment chosen given depth to bedrock in step 1 - coarser might be appropriate for lower basin). Weighted Averaged used - this means lower/upper is the same so only 5 cases (0-50cm,51-100cm,101-150cm,151-200cm, 201-250cm).
−
ACP, > was too shallow;
+
*0-50cm 64 um/s to 25 um/s
−
ACP, < was too shallow;
+
*51-100cm: 64 um/s to 25 um/s
−
WA, < was to shallow;
+
*101-150cm: 69.4 um/s, 44.9 um/s, 27.5 um/s, 25.7 um/s, 16.0 um/s, 9.2 um/s -
−
DCP, > bimodal (0 and 201) very shallow in upper basin;
+
*151-200cm: 54.8 um/s, 32.8 um/s, 20.7 um/s, 14.5 um/s, 9.2 um/s -
−
DCD, < bimodal (0 and 201) deeper than DCP, >. Areas with zero depth to bedrock are those classified as un-weathered bedrock in Surface Texture
+
*201-250cm: 9.2 um/s - this is for the lowest elevation unit in the basin which is likely the deepest
  
−
'''3) Soil Physical Properties: Saturated Hydraulic Conductivity (RHESSys requires Ksat at surface)'''
+
d) '''25 cm layer increment - WA, slow''' results for American Silver Fork (note 50cm increment chosen given depth to bedrock in step 1 - coarser might be appropriate for lower basin).
  
−
-Plot for each 100cm
+
*export attributes table, open with excel and save 2 columns (# and ksat) as csv
  
−
'''4) Soil Physical Properties: Percent Clay (used in RHESSys only for carbon cycling)'''
+
*csvread in matlab and plot Ksat for each soil unit (21 for Silver Fork, 3 are lakes) plot Ksat w/ depth
  
−
'''5) Soil Physical Properties: Percent Sand (used in RHESSys only for carbon cycling)'''
+
*determine m (how?) from Ksat in MATLAB
  
−
'''6) Soil Physical Properties: Percent Silt (used in RHESSys only for carbon cycling)'''
+
===5) Soil Physical Properties: %Clay, %Sand, %Silt (used in RHESSys only for carbon cycling)===
 +
a) Focus is not on carbon cycling - simplify and use WA, lower and upper were the same for all 3 variable
 +
Clay - range is 2.7% to 12.7%
 +
Sand - range is 66.3% to 80.4%
 +
Silt - range is 15.9% to 24.1%

Latest revision as of 23:17, 10 February 2010

Step 1: Get soil data

Download state data from here: [1]

Download county or survey level data from here: [2]

Both sources should include a) tabular data b) spatial data c) Microsoft Access template

State data is spatially coarser than county or survey level data

SURRGO format information [3]

Step 2: Install Soil Data Viewer

You need 1) Install ArcGIS 9.2 or later for soil data viewer 5.2 on a Windows XP system 2) Install soil data viewer version 5.2 [4]

Step 3: Prepare Access database

1) Open Access template 2) Enable macros 3) Type in full path name to tabular data 4) Import should run automatically

Step 4: Working with Soil Data Viewer in ArcGIS

1) Open ArcGIS - View > Toolbars > Soil Data Viewer Tools 2) Click on the Soil Data Viewer button 3) Browse to Access database 4) Choose the soil data layer to connect to the database

Soil data viewer online user guide is here: [5]

Note: Good practice to export layer because ArcMap won't remember them when your re-open your map

Step 5: An Example of estimating soil properties using SDV in the Silver Fork of the American

Definitions:

a) Dominant Condition (DCD)

b) Dominant Component (DCP)

c) Weighted Average (WA)

d) All Layers (ACP)

1) Soil Physical Properties: Surface Texture

Useful for a) naming the soil .def file b) understanding depth to soil restrictive layer c) understanding ksat at surface

2) Soil Qualities and Features: Depth to Any Restrictive Layer (RHESSys requires soil depth)

a) Generate Depth DCD, DCP, WA, ACP using both < and > for tie break rule with no cutoff (8 scenarios)

b) Compare these to summer Landsat imagery and use expert judgment to determine the most accurate layer

c) Depth To Any Restrictive Layer results for Silver Fork

  • ACP, > 127,143,153 - not much spatial variability, bedrock not represented
  • ACP, < 0 - basically junk
  • WA - tie break doesn't matter, 17 for texture un-weathered bedrock
  • DCP - tie break doesn't matter, very shallow in upper basin
  • DCD, > zeros where surface texture is bedrock, 77,127,201
  • DCD, < zeros where surface texture is bedrock, 77,127,201 but shallower than DCD, > (obviously)

3) Create a Raster that matches the shapefile

a) To Raster > Polygon to Raster (used Ksat's FID). Use MAXIMUM_AREA, set env. variables to size of dem b) Reclass - to get rid of No Date (map to 0)

4) Soil Physical Properties: Saturated Hydraulic Conductivity (RHESSys requires Ksat at surface and decay with depth)

a) Generate Depth WA for ALL LAYERS-ACP,slow and for Depth Range-WA,slow using both slowest and fastest for tie break rule with no cutoff (8 scenarios)

b) All Layers Ksat results for American Silver Fork

  • DCD - tie break doesn't matter, portions with surface texture bedrock have no data. Range is 91 um/s, 29 um/s 18 um/s
  • DCP - tie break doesn't matter, large portions of upper basin have no data. Range is 91 um/s to 28 um/s for lower
  • WA - tie break doesn't matter, Ksat is 60 um/s in upper basin and 26 um/s in lower basin. More spatial variability than ACP >
  • ACP < - Ksat is too slow
  • ACP > - Ksat is 91 um/s in upper basin and 28 um/s in lower basin. Less spatial variability than WA

c) 50 cm layer increment - WA, slow results for American Silver Fork (note 50cm increment chosen given depth to bedrock in step 1 - coarser might be appropriate for lower basin). Weighted Averaged used - this means lower/upper is the same so only 5 cases (0-50cm,51-100cm,101-150cm,151-200cm, 201-250cm).

  • 0-50cm 64 um/s to 25 um/s
  • 51-100cm: 64 um/s to 25 um/s
  • 101-150cm: 69.4 um/s, 44.9 um/s, 27.5 um/s, 25.7 um/s, 16.0 um/s, 9.2 um/s -
  • 151-200cm: 54.8 um/s, 32.8 um/s, 20.7 um/s, 14.5 um/s, 9.2 um/s -
  • 201-250cm: 9.2 um/s - this is for the lowest elevation unit in the basin which is likely the deepest

d) 25 cm layer increment - WA, slow results for American Silver Fork (note 50cm increment chosen given depth to bedrock in step 1 - coarser might be appropriate for lower basin).

  • export attributes table, open with excel and save 2 columns (# and ksat) as csv
  • csvread in matlab and plot Ksat for each soil unit (21 for Silver Fork, 3 are lakes) plot Ksat w/ depth
  • determine m (how?) from Ksat in MATLAB

5) Soil Physical Properties: %Clay, %Sand, %Silt (used in RHESSys only for carbon cycling)

a) Focus is not on carbon cycling - simplify and use WA, lower and upper were the same for all 3 variable Clay - range is 2.7% to 12.7% Sand - range is 66.3% to 80.4% Silt - range is 15.9% to 24.1%

Retrieved from "http://wiki-snow.eri.ucsb.edu/index.php?title=STATSGO2-soil_data&oldid=213"