3D Income Analysis in Tacoma, Washington

Fig. 1: 3D modeling of the topography of Old Town Tacoma, using slope data in ArcScene.

N

One of the most most useful outcomes of doing a GIS anlysis is the visualization of patterns, such as social, political, environmental, economic, etc.

The use of ArcScene allows a viewer to see these patterns three dimensionally. Some examples might be a high value seen as a high peak, or the topography of region appearing as it is in reality, or close to it (see Fig. 1).

This analysis will focus on child density in Tacoma, WA and how those values may (or may not) correlate with median income values, both analyses utilizing census block group data.

Skills Used:

- ArcScene

- 3D Analyst

- 3D modelling of elevation data

- 3D modelling of demographic data

- Downloading Census Demographic (American Community Survey Files)

1. Interpolation (IDW - Raster Analysis)

Census block group data from

the American Fact Finder

(see Fig.2) was used to create a

raster file.

IDW (Interpolation) analysis

was run to create a continuous

gradient of high density of

children or median income,

respectively, symbolized with

Esri's Slope gradient (white to

pink to green). See Fig. 3 and 4.

Fig. 2: Shapefile of census block groups as points.

^

N

Fig. 3: Interpolated raster of "Kid Density" in Tacoma. Raster output symbolized with standard deviation method using 1/2 standard deviation between 7 classes.

Fig. 4: Interpolated raster of "Median Income" in Tacoma. Raster output symbolized with standard deviation method using 1/2 standard deviation between 7 classes.

2. Density of Children Ages 10 and Under

Fig. 5: Interpolated raster of child density with 2 standard deviation between classes.

3. Median Income

Point Defiance Park

Port of Tacoma

I-5/Hwy 16 Interchange

Data from the American Community Survey allows for census block group analysis that includes ages of residents.

Fig. 5 was created to display how choosing a different standard deviation between the classes (representing low to high density of children by block group) creates two very different maps.

If not familiar with the area, it may be helpful to nate the large solid swaths of magenta (denoting low child density). the Northwest corner is a large forested park, the dead center is where a couple freeways merge and the central eastside is the Port of Tacoma - all areas where no one actually lives, but is withine city limits and was included in the interpolation step of analysis.

Fig. 6, also an interpolated raster like the previous maps, displays median income in Tacoma.

Again, because large uninhabitated areas were included in the interpolation process (Point Defiance, the intersection if I-5 and Hwy 16, the Port of Tacoma), it appears as though here are block groups there.

In another analysis it may be helpful to not include areas such as parks and large commerical districts.

The map viewer may beginto make assumptions about where higher and lower income areas might be, especially if already familiar with Tacoma.

Fig. 6: Interpolated raster of median income levels with 2 standard deviation between classes.

4. ArcScene Analysis

ArcScene™ is apowerful program that displays spatial data three dimensionally along X, Y, and Z axes. The displayed maps may be thematic or realistic, showing actual slope or using peaks to respresent hgh and low values (as shown in the following figures).

To create the 3D scene, the map layer including child density data (and, repecticly, income level data) was essentiall "draped" over a layer that dispalyed high and low values along a Z axis as high and low peaks.

In both maps, high, green peaks represent high values of density of hcildren or income level, and low, red areas represent low values. A street shapefile was also added.

Density of Children Ages 10 and Under

Fig. 7: ArcScene of density of children ages 10 and under.

Median Income

Fig. 8: ArcScene of median income levels.

5. Conclusion

The hypothesis that there may be higher densities of young children in neighborhoods of lower median income (nad vice versa - lower densities of children in higher income neighborhoods) proves to be a fair assessment. Though correlation does not mean causation, furthur analysis including high school graduation rates, number of persons per houshold, etc may lead to a richer analysis.

Figure 9 sybolizes both child density and income level data in one scene. The deep magenta to forest green gradient represents areas of low to high median income. This layer has been draped over a peaked child density layer: High peaks are areas with high density of young children, and lower valley area areas of low density of young children.

Note in the following figure the high, deep magenta peaks and the absence of many forest green peaks.

N

Fig. 9 ArcScene of median income levels and density of young children.

Lauren McKenna

lmckenna@uw.edu, April 2016

University of Washington Tacoma, Geospatial Technologies

ArcMap 10.3.1

Projection: NAD 1983 HARN State Plane Washington South FIPS 4602 Feet

Sources: US Census, American Fact Finder, UWT GIS 414 data.