Lab 1 - Applying the Spatial Analyst Extension

Lab Overview

Objective: To use ArcGIS to find the best location for a school in Stowe, Vermont and identify the least costly access route to the new school site.

In Lab 1, ArcGIS Model Builder and various Spatial Analyst tools are employed to identify the best place for a new school in Stowe, Vermont based on various factors including slope, distance from waterways and wetlands, and distance from recreational activities and existing schools. Lab 1 is separated into four exercises: (1) preparing for analysis, (2) accessing the Spatial Analyst extension and exploring the data, (3) finding a site for the new school, and (4) finding an alternate access route. The following report details the methods used and the outputs of the tools applied to solve this problem.

1.0 Preparing for analysis: understanding the ArcGIS Spatial Analyst extension

Exercise 1 introduces us to the Spatial Analyst extension, the data used to solve current problem, and the workspaces used to organize it.

1.1 What is the Spatial Analyst extension?

The Spatial Analyst extension contains a variety of tool-sets designed to support spatial analysis of raster and vector data. In the current lab, the Conditional, Distance, Overlay, and Reclass tool-sets were applied to identify the optimal location for a new school in Stowe, and to identify the least-cost path from the given destination point to the school to support, for example, the construction of a road for school access. A description of the tools used in Lab 1 is presented below. Note that these are taken from ArcGIS.com (2016a) Spatial Analyst Toolbox.

Hillshade: creates a shaded relief from a surface raster using illumination source angle and shadows.

Euclidean (straight-line) Distance: measures the straight-line distance (in degrees) from each cell to the closest source, measured from cell centre to cell centre.

Slope: identifies the gradient, or rate of maximum change in z-value (slope) from each cell of a raster surface

Weighted Overlay: a method of comparing several rasters and assigning value fields weights according to their importance, using a common measurement scale

Cost-weighted Distance: modifies Euclidean distance by accounting for accumulated travel cost. Uses the Cost Back Link tool, which identifies the route to take from any cell along the least-cost path, back to the nearest source.

Cost Path: calculates the least-cost path (in degrees) from a user-defined source to a user-defined destination.

Con tool: performs a conditional if/else evaluation based on a SQL command, on each input cell in an input raster

Majority Filter tool: replaces cells in a raster based on the majority of their contiguous neighboring cells

1.2 Data

We were provided with the following datasets to solve the problem:

Elevation
Landuse
Roads
Recreational Sites
Schools
Road construction starting point (Destination)

1.3 Workspaces

Workspaces are the directories that hold datasets and geodatabases in many file formats for use in ArcGIS programs. They are useful for organizing, documenting, and cataloging geographic datasets (ArcGIS Desktop, 2016). In Lab 1 the Stowe.gdb and the Scratch.gdb workspaces were created to store the project files.

Figure 1

The Scratch.gdb and Stowe.gdb workspaces

2.0 Accessing the ArcGIS Spatial Analyst extension and exploring the data

Exercise 2 explores several Spatial Analyst tools, including the Hillshade tool, categorization of landuse values using the Symbology Tab, identifying features by attribute and field using the Attribute Table, and displaying a histogram based on a specific field.

2.1 The Hillshade Tool

After enabling the Spatial Analyst extension, a hillshade of the Stowe terrain is created using the Hillshade tool. This layer is later used to demonstrate the changes in elevation and slope under subsequent map layers. The Z values (elevation values) are converted to meters by setting the Z factor to 0.3048, essentially multiplying the original values (in feet) by 0.3048. Figure 2 below shows the output of the Hillshade tool.

2.2 Land-use Map: value field catagorization and layer-specific histograms

Under Layer Properties, the Symbology tab allows users to tweak how particular value fields are displayed in ArcMap (ArcGIS Desktop, 2016b). In this case, the Unique Value option is chosen, allowing us to assign appropriate colors for each individual land-use type on the provided land-use map. The transparency of the land-use map is adjusted to 30%, and it is superimposed on top of the hillshade map to produce Figure 3, a map of the land-use categories on the impression of Stowe, Vermont terrain.

The Histogram tool on the Spatial Analyst toolbar is activated with the land-use layer selected, producing a histogram with the y-axis representing the number of cells covered by each type of land use on the given map. This can be used to determine the approximate area covered by each land-use category relative to the other categories. The histogram is shown in Figure 4 below.

Figure 2: Hillshade

Figure 2 shows the output of the Hillshade Tool, applied to Stowe, VT. Changes in elevation and slope can be determined visually.

Figure 3: Landuse

Figure 3 shows the landuse map with a transparency of 30%, superimposed, over the hillshade map.

Figure 4: Histogram of Landuse

Figure 4 shows a histogram of the landuse categories. The y-axis represents the amount of pixels covered by each land use category. This can be used to determine the approximate area covered by each landuse category relative to the other categories

3.0 Identifying the optimal site for a new school: creating a suitability model

In Exercise 3 we manipulate our data to solve the problem at hand - identifying the optimal location for the new Stowe school. Model Builder is used to create the final maps and store the tools and data employed to create them. The results of this exercise are displayed below.

3.1 Slope, Euclidean Distance, and Reclassify

First, we use the elevation, school, and recreational site data-sets to create three base-maps. The model used to create these maps is shown in Figure 5. The Slope tool takes elevation as an input and produces an output (Figure 6). The Euclidean Distance tool is used with both school and recreational site data to produce distance rasters showing desirable and undesirable locations based on distance from the source data, ie: near recreational sites (to support easy access by students, Figure 8), and far from existing schools (to maximize the impact of the new school, Figure 7). In order to used these variables to meaningfully determine suitability, they must be adjusted to a common measurement scale. To this end, each of these these are reclassified using the Reclassify tool into 10 categories, the first of which (1) is least suitable for building, and the last of which (10) is most suitable. For example the school must be constructed on relatively flat land, so areas with the smallest slope value (assigned a value of 10) are considered most suitable for the new school site (Figure 9). The reclassified outputs from the Euclidean Distance tools are shown in Figure 10 (schools) and Figure 11 (recreational sites).

Figure 5: Reclassification model for elevation, schools, and recreational sites

Figure 6: Terrain of Stowe, Vermont, elevation in meters

Figure 7: Euclidean distance from schools, Stowe, VT

Figure 8: Euclidean distance from recreational sites, Stowe, VT

Figure 9: Reclassified Terrain, Stowe, VT

Figure 10: Reclassified Euclidean distance from schools, Stowe, VT

Figure 11: Reclassified Euclidean distance from recreational sites, Stowe, VT

3.2 Weighing and combining datasets

The next step is to weigh and combine the reclassified datasets in order to find areas suitable for our school site. Figure 12 shows the model used to create the suitability map. Landuse values are added here, in order to exclude wetlands and water. Slope values less than 4 are considered too steep, and are restricted as well. Using the Weighted Overlay tool, the datasets are weighed with regards to their relative importance:

Reclassed distance to recreational sites: 50%
Reclassed distance to schools: 25%
Reclassed slope: 13%
Landuse: 12%

The output of the Weighted Overlay tool is presented in Figure 13, a raster of the suitable areas for the school. These regions are further refined using the Con tool, which extracts those sites with the highest suitability value (9) only. Finally, these results are further filtered using the Majority Filter tool, which removes areas too small to be suitable for school construction. The final output is presented in Figure 14.

Figure 12: Model Builder for optimal school site identification.

Figure 13: Suitable areas for the new Stowe School

Figure 14: Optimal areas for the school, filtered with the Con and Majority Filter tools

4.0 Identifying the least-cost access route

In the final exercise, the least cost-access route from the given destination point to the optimal school site was calculated using the Weighted Overlay tools, the Cost Distance tool, and the Cost Path tool. In this case, the least costly path was defined as the one with the fewest steep slopes, as it was decided that these would be the most expensive in terms of labour and materials for building a road to the school from the destination point. Figure 15 shows the model used to define the least-cost map, and Figure 16 is the map itself.

Figure 15: Model for generation of least cost path analysis

Done for Advanced GIS for Natural Resource Management, in the McGill University Department of Natural Resource Sciences, Professor Jeffrey Cardille

References:

ArcGIS for Desktop. 2016a. ArcCatalog Basics. Retrieved from: http://desktop.arcgis.com/en/arcmap/10.3/manage-data/using-arccatalog/arcgis-data-is-organized-in-workspaces-and-geodatabases.htm

ArcGIS for Desktop. 2016b. Spatial Analyst: Tutorial. Retrieved from: https://desktop.arcgis.com/en/arcmap/10.4/extensions/spatial-analyst/tutorial/exercise-2-accessing-spatial-analyst-and-data-exploration.htm

Google Sites

Report abuse