Suitability Modeling

Understanding of Suitability Modeling through Finding Optimal habitats for Bald Eagle in the San Bernardino National Forest

Problem and Objective

Suitability modeling can be used in various applications. It involves finding where something is expected to be or identifying a new location for a phenomenon. This analysis is centered around a clear definition of a goal and criteria. The purpose of this analysis is to identify three optimal habitats for the bald eagle in the San Bernardino National Forest.

Analysis Procedures

Strategies: The strategies adopted for this analysis are explained as follows. I used ArcGIS Pro (Version 2.9) software. I used the following Modules/tools for the analysis: the raster calculator tool, the distance accumulation tool, the aspect tool, and the suitability modeler. Data and data types – The data include (a) raster data of NLCD land cover (b) a polygon feature of major lakes, (c) raster data of DEM, and (d) raster data of canopy cover. Data sources – Esri Academy through GIS 520 Fall Semester, NCSU.

Methods: I opened the project and reviewed all the source data. I set the parameters in the environment dialog box, under raster analysis to NLCD land cover data. I used the distance accumulation tool to derive a surface for the distance from the major lakes. I used the raster calculator tool with a map algebra expression to extract the developed areas from the NLCD land cover data and every other class was classified as no data. Then, I used the distance accumulation tool to compute a surface for the distance from the developed areas. I used the aspect tool to derive an aspect surface raster from the DEM.

Next, I created a suitability model using the suitability modeler. I added all four derived data to the model. I configured the model environment (in the suitability modeler pane) using the lake mask raster data as a mask since lakes are considered unsuitable for eagle nests. By checking each derived raster data in the suitability modeler, the transformation pane automatically opens and their respective transformed raster layer will be added to the content pane. For the lake surface distance in the transformation pane (continuous functions), I selected the small option for the function, 1609.35 meters as the midpoint and 1 as the point spread. Also, I set every other parameter as specified in the instruction. I transformed the derived developed land raster by selecting the large option as the function and 0.5 as the point spread in the transformation pane. I transformed the canopy cover raster by selecting the near option as the function, 45 as the mid-point, and 0.001 as the point spread. I transformed the aspect raster by selecting the symmetric-linear option for the function, with 22.5 degrees as the minimum value and 67.5 as the maximum value. For each of these transformed data, I clicked the calculate button to effect the changes in the parameters. Also, I compared these transformed data to their respective source data.

I selected percent as the weight by option under the setting table in the suitability modeler pane. I assigned a weight to the individual transformed raster under the suitability tab in the percent column as follows: 40% for lake surface distance, 30% for developed distance, 20% for canopy cover, and 10% for aspect surface, and then I clicked the calculate button to effect the change. I changed the cell size of the raster analysis to 60. I also changed the number of regions to 3 under the locate tab in the suitability modeler pane. I also selected ellipse as the region shape, square miles as area units, 1.2 as the region maximum area, and 3.4 as the total area. Additionally, miles was selected as the distance units with 1 for the minimum distance between regions.

Workflow diagram of the analysis

Result

The map above represents the three suitable locations for bald eagle habitats in the San Bernardino Nationa Frest based on the following criteria: (a) Close distance to lakes, (b) Far distance from developed areas, (c) Not too densely or sparsely covered by forest, and (d) Located on northeast-facing aspects.

Application & Reflection

Problem Description: The feature of the land selected for farming can lead to higher productivity. Factors such as access to water, road for fuel efficiency, source, large space, market, secure location, feed, power supply, etc. The purpose of this analysis is to find three optimal locations for the establishment of poultry farms with about 200,000 square feet of area in Wake County.

Data Needed: The data needed include (a) NLCD land cover raster data (b) Stream/river network feature (c) line feature of road networks (d) raster data of DEM, (e) point feature of the location of poultry feeds stores, and (f) raster canopy cover

Analysis Procedures: Upon adding the data to the map and configuring the environment parameters as required, I will derive a surface for the distance from the major streams, roads, and poultry-feed location using the distance accumulation tool. I will use the raster calculator tool with a map algebra expression to extract the developed areas from the NLCD land cover data and every other class was classified as no data. Then, I will use the distance accumulation tool to compute a surface for the distance from the developed areas. I used the slope tool to derive a slope surface raster from the DEM.

Next, I will create a suitability model using the suitability modeler. I will add all six derived data to the model. I will transform each of the derived data in the suitability pane as follows: For the stream and road surface distances in the transformation pane (continuous functions), I will select the small option for the function, 3218.7 meters (2 miles) as the midpoint and 1 as the point spread. For the poultry-feed surface distance, I will also select the small option for the function, 8046.75 meters (5 miles) as the midpoint and 1 as the point spread. As a result of the smell of the poultry farm, it will be located very far from the developed areas. I will select the large option as the function and 0.5 as the point spread in the transformation pane. I will select the canopy cover raster by selecting the near option as the function, 35 as the mid-point, and 0.001 as the point spread. I will transform the slope raster by selecting the symmetric-linear option for the function, with 20 degrees as the minimum value and 40 as the maximum value. I will compare these transformed data to their respective source data.

I will assign a weight to the individually transformed raster under the suitability tab in the percent column as follows: 10% for the stream surface distance, 25% for the road distance, 10% for the developed distance, 20% for canopy cover, 20% for the poultry-feeds location raster and 15% for slope surface. I will also change the number of regions to 3 under the locate tab in the suitability modeler pane. I will select ellipse as the region shape and square miles as area units. Additionally, I will select miles as the distance units with 1 for the minimum distance between regions.

Page updated