Special Topics in GIS Module 6: Scale Effect and Spatial Data Aggregation

This week was the final week of my Special Topics course. In this last lab, we learned how scale and resolution affect vector and raster data on maps (respectively). When working with vector data (e.g., polygons and polyline shapefiles), we must be careful to make sure that all of the data is displayed at the same scale. Vector data is more detailed at a smaller scale; however, resolution and accuracy are negatively affected. When working with raster data (LiDAR DEMs, etc.), the higher the resolution scale, the larger the cell size and the less detailed the data will become. In this week's lab, we worked with 1-meter LiDAR data and changed the resolution up to 90-meters. The raster data at a 90-meter resolution was far less detailed. 

In this week’s lab, we also reviewed US congressional districts to see if the district's lines had been drawn with intentional political bias or gerrymandering. To identify evidence of gerrymandering, I used the Calculate Geometry Attributes geoprocessing tool to identify multi-part features in the attribute table. I identified 15 US congressional districts that included multi-part features. I reviewed the geography of these 15 features and found that 7 of the features were multi-part because the districts included islands. The remaining 8 districts did not appear to have geographic reasons to be multi-part. 

Next, I needed to analyze these 7 features for compactness. Congressional districts that sprawled across the landscape in odd shapes were scrutinized by calculating the Polsby-Popper score. The calculation for this score checks the multi-part polygon for compactness by measuring the total area and perimeter with a circle that is the same size as the polygon's perimeter. Ideally, the more area that fits into the circumference of the circle, the higher the Polsby-Popper score will be. The closer the score is to 1, the more compact the polygon is deemed. I used the calculation (4 * Pi [3.14]) * (Polygon Area / Polygon Perimeter Length ^2). 

I discovered that the single worst offender of gerrymandering was US Congressional District 8 in Maryland, with a score of 0.082. The boundary of the district lies against Washington D.C and covers an oddly shaped area northwest of the Capital. The screen capture above displays the offending district.