GIS 4048 Module 8: Location Decisions - Homing in on Alachua County, Florida

Figure 1: Layout depicting the four variables that are the basis for choosing a housing location.

Figure 2: The results of two Weighted Overlay tool operations using the four variables from Figure 1 and different weightings.

The assignment for this module saw us working through the process of selecting areas for buying a house based on four variables that our hypothetical clients provided. These variables were proximity to two workplaces, percentage of the population between 40 and 49 years old, and percentage of homeowners versus renters. Our first map layout (Figure 1) depicts the four variables separately while the second layout (Figure 2) depicts the results of using the Weighted Overlay tool to combine the four variables and categorize areas based on different variable priorities.

The first step in preparing the data for our variables was to use the Euclidean Distance tool to arrive at a classification of Alachua County, FL based on distance from the two work places. Because all variables were to be compared and combined in the Weighted Overlay tool at the end, we then used the Reclassify tool to break the proximity categories into simple, single-digit values (reversing the values made the higher values closer and the lower values further from the workplaces).

The variable of population percentage between 40 and 49 years of age required a new field be added to the census tract layer's attribute table in which this calculation could be run. The field calculator was used in the new field with a formula to divide the number of people aged between 40 and 49 years by the total population with the result multiplied by 100 to get the percentage. This field could then be used as the basis of a choropleth map. The next step was to convert this vector layer into a raster (in order to use it with the proximity rasters and weighted overlay tool) with the Feature to Raster tool and to reclassify it as described for the proximity layers above.

The method for processing the percentage of homeowners variable was the same as for the age variable. A new field was added to the census tracts attribute table and the field calculator was used to populate the field with the results of dividing the number of homeowners by the number of households and multiplying the result by 100. The Feature to Raster and Reclassify tools were then used to finalize the processing.

The final task was to use the Weighted Overlay tool with the four rasters representing the variables described above as inputs. Because the Weighted Overlay tool would be run more than once, a model was created to streamline the process. The first run of the Weighted Overlay tool gave each variable equal weight (25%); the results are seen in the top map of Figure 2. For the second run, we were to consider the fact that our hypothetical clients were not happy about the traffic of the region and wished to prioritize workplace proximity. Thus, for the second run the weights for the two proximity rasters were set to 40% while the other two variables were set to 10%. Additionally, the scale values for the lowest three proximity categories were set to restricted, effectively eliminating such areas from consideration. The results can be seen in the second map of Figure 2. The second weighted overlay opens up areas between the two workplaces as being most highly rated, giving the hypothetical clients more options to consider in their pursuit of a home location.