Least Cost Path (Carnivores)

** Currently under construction **

Least-Cost Path Model for Grizzly & Wolverine in North-central Washington State
Grizzly Team: Tate Mason, Temsha Huttanus, Shane Skaar, Tempe Regan, Michelle Jeffries
Wolverine Team: Allie Anderson, Shawn Smith, Heidi Ware, Bryce Robinson

Methods described here were modified from Singleton et al. (2002), “Landscape permeability for large carnivores in Washington: A geographic information system weighted-distance and least-cost corridor assignment”, U.S. Forest Service Research Paper PNW-RP-549 PDF.

The work here involves three parts, a.) the acquisition and preparation of 5 input rasters, b.) creation of the “travel cost” raster, and c.) calculation of a specific least cost “travel path” from Point A to Point B, as determined by the user. Each of the 5 input rasters are classified and each class assigned a weighting factor (a coefficient between 0.0 and 1.0). The analysis is performed using 90m pixels, so all 5 input rasters must be standardized to this cell size prior to calculation. The travel cost equation is:

Elevation x
Slope x Population Density x Road Density x Habitat Type

MAP SETUP
1.) Set up ArcMap:
– Open new, blank .mxd document.
– Set Data Frame Coordinate System to a projected system (Ex: UTM_NAD83_Zone11).
– Set Data Frame Display Units appropriately (Ex: Meters).
– Turn on Spatial Analyst extension.
– Turn off Background Processing (Geoprocessing menu > Geoprocessing Options).
– Set working folder or set Environments for Workspace and Scratch Space.
– Determine your study area boundary. Create a polygon shapefile of the study area.

ACQUIRE & PREPARE DEM
2.) Acquire 30m ASTER DEM from EarthExplorer. Data arrives in GCS_WGS84. Unzip files. Load DEM(s) into ArcMap.

3.) Optional: Mosaic DEMs (Data Management > Raster > Raster Dataset > Mosaic to New Raster tool). See additional info HERE.

4.) Fill sinks and holes in DEM (Spatial Analyst > Hydrology > Fill).

ELEVATION RASTER
5.) Resample DEM to 90m pixels (Raster > Raster Processing > Resample tool). Singleton et al. (2002) used 90m pixel in their analysis.

6.) Convert DEM to integer format (Spatial Analyst > Math > Trigonometric > Int tool).

7.) Reclassify elevation (meters) using 4 classes and the following value ranges, Break Values, and new values:

0-1000 –> 1
1000-1500 –> 2
1500-2000 –> 3
2000-max –> 4

SLOPE RASTER
8.) Create slope raster in percent rise from 30m DEM – the filled, projected one NOT the one resampled to 90m (Spatial Analyst > Surface > Slope).

9.) Reclassify Slope raster based on criteria provided in Singleton et al. (2002)…….

HABITAT LAYER
10.) For this project, we’ll used Landfire data (landfire.cr.usgs). Download

11.) Group Landfire habitat/vegetation classes (field name?) to reasonably match categories shown in Singleton et al. (2002)…….Do this in Excel…

12.) Find/Replace PERM values with those from Singleton et al. (2002)…

13.) Join Excel table to Landfire raster…

HUMAN POPULATION DENSITY
14.) Acquire human population data (census blocks) from Census.gov. Look for the latest TIGER/Line Shapefiles. Download “Block” data for the entire state, then clip to study area boundary (Geoprocessing > Clip tool).
Link: http://www.census.gov/geo/maps-data/data/tiger.html.

15.) Clip census data to study area boundary (Geoprocessing menu > Clip tool).

16.) Calculate population density by census block polygon:

– Open Attribute table
– Add new field, AREA_KM2, Float format
– Calculate geometry (km2) for the field
– Add another new field, POPDENSE, Float format- Use Field Calculator to divide POP10/AREA_KM2 to get density in persons per km2

17.) Convert population density shapefile to a raster:

– Input = clipped population density shapefile
– Field = POPDENSE
– Output = blk10pop
– Output cell size = 30

18.) Focal Statistics tool to determine population in 0.9 km2 radius neighborhood. The output will contain a value representing the number of people within 900m of each pixel.

– Input = blk10pop
– Output = blk10pop_fs
– Neighborhood = Circle
– Radius = 900
– Units = Map (make sure Display Units are Meters)
– Statistics Type = SUM

19.) Reclassify population layer using 5 classes, Break Values of 10, 25, 50, 100, 10000, and New Values of 1,2,3,4,5 respectively. So the value ranges for the five classes are assigned new values 1 through 5:

0-10 –> 1
10-25 –> 2
25-50 –> 3
50-100 –> 4
100-10000 –> 5

– Input = blk10pop_fs
– Field = VALUE
– Output = blk10reclass

20.) Edit table for blk10reclass. Open attribute table and add a new field named GRIZZ_WT, Float format.

21.) Open Editor toolbar, start editing blk10reclass. Replace values of 1,2,3,4,5 with the weighting values provided in Singleton et al. (2002)…..When done, stop editing and save edits.

22.) Resample population raster to 90m pixels (Raster > Raster Processing > Resample tool).

ROAD DENSITY
23.) Road density data is available from…

RASTER CALCULATOR
.) The travel cost raster equation is shown below. Make sure all input rasters are 90m pixels and in Float format. Double check your syntax in raster calculator.

Travel Cost (Permeability) Equation:
Slope x Elevation x Road Density x Population Density x Habitat Type

Syntax in Raster Calculator:
Lookup(“inputraster1″,”field”)*Lookup(“inputraster2″,”field)*Lookup(“inputraster3″,”field”)…

.) Create start and destination points or polygons. Use either existing files or create new ones with Draw tools and Convert Graphics to Features….

.) Cost Distance tool…

.) Cost Backlink tool…

.) Cost Path tool…

.) Convert path to polyline (Conversion tools > Raster to Polyline tool)…

 

Comments are closed.