GIS Analysis Flowcharts

The assignment of treatment prescriptions required several attributes that are calculated through ArcGIS. This page provides a graphical representation of the steps required to calculate required attributes and to generate the comma separated file (>CSV file) used by the assignment program.

The first step is to describe the desired treatment area. This can be as simple as a rectangle bounding box, or it can be an arbitrarily complex collection of polygons. For example one test case for this model is the John Martin State Wildlife Area and adjecent Corps of Engineers land. The basic idea of this step is to create a boundary layer for our treatment presecriptions. Using Analysis Tools -> Extracts -> Clip, clip the tamarisk data to the treatment area. For display purposes, roads, water bodies, and cities are clipped to a regional map such as the county which contains the treatment area.

Flow to create Tamarisk to Treat layer file

One factor influencing the cost of a tamarisk control project is the remoteness of the treatment area. This is described by a remoteness layer incorporating roads and off-road travel. The first step to create a remoteness layer is to create the grid used to calculate travel cost. This is a multi-step process: we need a grid for road travel cost and a grid for off-road travel cost. These grids are then merged into an overall travel cost grid.

To create the road cost grid, we merge the three road layers available from CDOT: local, collector, and highway road layers. We then assign a speed limit attribute to each road section based on the road surface type:
Road Type
 Speed limit
10 Primtive 2
20 Unimproved 5
30 Graded & Drained 15
40 Soil, Gravel or Stone 25
51-53 Bituminous
 35
61-62 Flexible
 45
71-72 Rigid
 50
80 Other
 55

Create a COST column with type short integer. Calculate the cost of movement along the road as 66 – SPEEDLIM. This produces a cost range from 1 for highways with a speed limit of 65, to 64 for dirt roads with a speed limit of 2. We then convert the road polylines to a grid using the Polyline to Raster tool. The value field of the grid is the movement cost along the roads.

Off-road travel is much slower than travel by roads. Though not part of this model, the environmental impacts of off-road travel are much greater than travel by road. The model minimizes off-road travel by ensuring the off-road_travel_cost layer is always higher cost than the road_travel_cost layer. A slope layer (created from a DEM of the treatment area and surrounding areas) is processed by a custom model created through model builder. The model performs the following calculations:
  1. Reclassify the slope model in two values. Less than 20 degrees gets 0, greater than 20 degrees gets 1000. This reflects the difficulty of travel on slopes greater than 20 degrees. The temporary file is called SlopeReclas20.
  2. Multiply slope times 100, then add 1000 to the result
  3. Add the two rasters together for a final slope travel cost grid.

This gives the cost based on the slope, ranging from 1000 as a base off-road cost to a high of over 10,000 for steep areas. The model as shown in Arc's model builder looks like this:
Model Builder Image
Once we have the road travel cost and the off road travel cost, we can merge them into a total travel cost layer using the Spatial Analyst Tools->Map Algebra->Single Output Map Algebra tool. Road travel is selected where it is defined (on the roads) otherwise the off road travel cost is selected. The output is the overall travel cost layer.
con(isNull([bentroadscost]), [slopecost], [bentroadsCost])


Flow to create cost layer

The remoteness layer is generating by calculating the cost-distance for every point in the treatment area to the nearest town. Use the Spatial Analyst Tools->Distance->Cost Distance tool to generate a cost distance grid.

Flowchart to create cost-distance layer
There are two attribute types needed to assign treatments: slope and cost (or remoteness) data. For both types of data we create an external flat text file (.CSV file) then join this file with the tamarisk polygon data table using the FID field. The Spacial Analyst Tools->Zonal->Zonal Statistics as Table tool for this process.

The process to add slope attributes:

Flow to add slope data to tamarisk polygons

And the process to add cost attributes:

Flow to add cost data to tamarisk polygons

Once we have joined the required attributes to the table of tamarisk polygons, we need to export this data to a CSV file so the program to assign treatments can read the data. Some columns need to be renamed to use human-readable labels.

Flow to export data to a CSV file


To group polygons, we need to know which polygons are adjacent. This can be calculated with Hawth's Tools to get a matrix of distances between the centroid points from each polygon to every other polygon. The centriod points are created with the Spacial Analyst Tools->Zonal->Zonal Geometry As Table tool. We then add the zonal geometry table to the map and Display XY Data. This provides a view of the centroid points. With Hawth’s Tools->Analysis Tools->Distance Between Points (Within Layer) we then generate a NxN matrix of the distance between each polygon centriod.

FLow to create centroid matrix

The above steps have generated two .CSV files:

  1. The file of tamarisk polygons, including slope and remoteness cost information
  2. The file of distances between polygon centriods

These layers are the input for the custom program to compute treatment prescriptions.