Suitability Analysis and Weighted Overlay

11/6/14

Problem:

Park visitors and bears in Great Smoky Mountain National Park have a tendency to interact, particularly around roads and trails in the park. Biologists maintain a robust relocation program , but these encounters still occur and recur. As a result, the biologists would like to identify areas of the park that will support the bears' basic needs while minimizing the encounters. Research efforts have indicated that five primary parameters influence the likelihood that bears will return to trouble spots after relocation. These parameters include distance to roads, trails and streams, vegetation cover, and slope of the terrain. The data used for these analyses were acquired from Dr. Perver Baran at NCSU. The data included four shapefiles and a raster describing roads, trails, streams, vegetion cover, and elevation information for Great Smoky Mountain National Park. The coordinate system used for these data was UTM, NAD27, zone 17, meters. Using these information, the general strategy is to combine the datasets to identify areas that meet the requirements of all five parameters described above. All parameters are to be parsed into three categories of Most Favorable, Favorable, and Least Favorable: Catetories for distance to roads and trails are greater than 1, 0.51 to 1, and 0 to 0.5, respectively; distances to streams are 0 to 0.5, 0.51 to 1, and greater than 1, respectively; slopes are 0 to 30, 30.1 to 60, greater than 60, respectively; and vegetation types are Grape Thickets, Cove/Mesic/Northern Hardwood, and all others, respectively.

Analysis Procedure:

The general method used to perform these analyses involved first generating a folder in my system to contain the new map, toolbox, etc. using Windows. To provide the necessary location for the model, the next step was to produce a Toolbox, performed using ArcCatalog, followed by producing a blank model. The model generation was initiated by importing the vegetation layer to new ArcMap and setting the environment accordingly (processing extent = vegetation layer, cell size to 30 meters, etc.). The following were performed in the model. Using the Elevation raster, analysis was performed with the Slope Tool, and Euclidean Distance operations were performed on the Roads, Trails, and Streams layers. All layers were then Reclassified following the parameters described above. Upon reclassification, a Weighted Overlay Analysis and the resulting raster were generated.

Figure 1. Workflow diagram for Suitability Analysis and Weighted Overlay

Results:

The model generated to perform the Weighted Overlay operations (Figure 2) converts features to rasters followed by performing Euclidean Distance operations, Reclassification, and Weighted Overlay. The Reclassify Distance Tool converts the Euclidean Distance values for each of the prescribed layers of roads, streams, trails, vegetation, and elevation to a similar scale so weighting functions can be performed. After reclassification, the weighted overlay procedure can be performed. From these analyses, a number of areas that are candidates for bear relocation were identified (Figure 3).

Figure 2. Model for Weighted overlay

Figure 3. Land Suitability for Bear Habitat in the Great Smoky Mountains National Park

Application and Reflection:

The weighted overlay method of data analysis allows one to link layers of different units by putting them into a common scale thereby providing a way to determine the most optimal conditions. In statistical terms (in my limited understanding), it seems akin to principle components analysis. Returning to the mushroom hunting project my daughters and I are working on (see Our Mushroom Hunting), moisture content may be the most heavily weighted condition determining which mushrooms will grow where. But, other factors are also involved, including temperature, soil type, nutrition, prevalence of detritus, and others, most of which will have different units of measurement and varying levels of significance. If we are trying to predict, for example, that an Eastern Cauliflower mushroom will be growing in a particular area of Eno River State Park, we may be able to do so by performing the weighted overlay analysis. The data I would like to have for these analyses are soil moisture content, light level, temperature, relative humidity, barometric pressure, and GPS coordinates collected at the time of the mushroom sightings. I would also like to know the trail segment location, slope, aspect, elevation, and the information produced through linear referencing. Each mushroom variety, it seems, will have different requirements, so further information about optimal environments for the different genera may be useful. I suspect a local mushroom hunting club might have such data. To start, however, I will likely perform statistical analyses (see Spatial Statistics) on the data to identify which parameters are more likely to influence the mushroom sightings, then use these in the weighted overlay.