Suitability Modeling

Finding Optimal Regions for Bald Eagle Habitats in San Bernardino National Forest

Problem and Objective

Scientists with the San Bernardino National Forest are looking to find optimal regions for bald eagle habitats within the forest. To ensure the most optimal regions are selected, the scientists have identified criteria that describe the best regions for bald eagle habitat and would like to create a weighted suitability model and map incorporating the criteria.

Analysis Procedures

To create the weighted suitability model and find the optimal regions of habitat, I used ArcGIS Pro 3.1 and various tools from the Spatial Analyst toolbox and the Suitability Modeler pane.

The data used for this analysis included lake boundaries from the National Hydrography Dataset, raster data indicating land cover classifications from the National Land Cover Database, an elevation raster from the Advanced Spaceborne Thermal Emission and Reflection Radiometer, and a raster indicating percent tree canopy values from the NLCD, all provided by ESRI.

The criteria I focused on was finding areas that were within 2 miles of lakes, areas far from developed areas, areas with around 45% tree cover and located on northeast-facing aspects.

After defining the criteria for our weighted suitability model, the next steps were to derive the source datasets to create rasters that represent our model variables. Before doing any analysis, I first verified the cell size and coordinate system of each raster layer. I then changed the environment variables, first for processing extent which was set to NLCD land cover layer, and then also set the Cell Size, Mask and Snap Raster for this project to the NLCD land cover layer.

To derive the data needed, I first ran the Distance Accumulation tool on the lake boundaries raster, which produced a raster layer with distances to each lake. To determine distance to developed areas, I first selected all of the developed areas in the land cover dataset using the Raster Calculator tool, then ran the Distance Accumulation tool. To derive the slope, I used the Aspect tool on the ASTER DEM raster, which produced a raster indicated the degree of slope for each cell. The percent tree canopy raster already met our criteria goals and deriving a new raster was not necessary.

I then added a binary mask to block out all lakes from the suitability modeler as they do not contain trees and cannot be used for nesting. I then applied transformations for each derived raster to meet the criteria. I then applied percent weights to each layer as specified in the criteria given. I then used the Locate feature of the Suitability Modeler to identify 3 regions that met the model goal of optimal bald eagle habitat.

 

Results

Application & Reflection

Creating a weighted suitability map has applications in a vast array of geoprocessing projects and can be used any time you are trying to find the best place to put something while taking into account multiple criteria.

Problem description

For a hypothetical project, I have envisioned a scenario where a renewable energy company is looking for the best location to place new wind turbines to produce electricity. Areas best suited for wind farms typically have locations with frequent, sustained winds, unpopulated areas and areas with open plains and water.

Data needed

A map of average wind speed by zip code, elevation data to find open plains, hydrography data to locate water sources and population density data from the US Census.

Analysis procedures

I would first consult with professionals to create the weighted criteria needed for creating a suitability map. I would then derive the data from the elevation data, hydrography data and population density data using the Distance Accumulation tool. Then I would use the Suitability Modeler pane to set up my weighted criteria and produce a final weighted suitability map for possible wind farms.