Background
One of the things that always shocked me about the New York City subway lines was how much the racial makeup of the passengers varied along each and every subway line. It makes logical sense, since our city is made up of so many little pockets of races and nationalities, and the subway system takes us through them – but the transition of racial distribution in passengers is sometimes so drastic that as few as a few subway stops can take you from a surrounding of white upper-class Upper West Siders (eg. “86th Street” on the D train) to having not a single white person in sight (eg. 5 stops later, “125th Street” on the same D train).

In this report, I wish to overlay geographical data of the subway lines and stops onto a map of racial distribution in NYC to give a visual showcase of where, or, which stops these racial transitions take place. For people who hardly ever leave their borough/neighborhood – that was me during my first years in New York – choose a subway line you often take, see where it starts and ends, and see how the neighborhood demographics change before or after your trip. Hopefully, this map can help you understand the people around you better, maybe even encourage you to explore different neighborhoods of the city.

Materials
Dataset:
Subway Lines – NYC Open Data
Subway Stations – NYC Open Data
Race and Hispanic Origin Variables (Boundaries) – ACS (retrieved from ArcGIS’ Living Atlas)
Tools: ArcGIS (mapping)

Methods
Upon opening ArcGIS, I chose the “Dark Gray Canvas” basemap because it would form the most contrast with the subway lines, which I eventually colored according to the actual NYC subway system. For my base feature layer, I used the latest ACS Race dataset that is available on ArcGIS’ Living Atlas. Two versions are available for this dataset: one represents the data as boundaries, the other as centroids. I chose the boundaries version because this would provide me with a clear dividing line between census tracts. In order to use this information, I first filtered out racial data in Staten Island’s census tracts, because there is no subway there. The remaining census tracts are those of Brooklyn (Kings County), Queens (Queens County), Manhattan (New York County), and Bronx (Bronx County). I also filtered out census tracts in which the total population is at least 500, since I would like to display racial numbers by percentage and a very low population number may bias the result. For this layer, I chose the “Predominant Category” styling.

Then, from NYC Open Data’s website, I downloaded the NYC subway line and station information as GeoJSON files and uploaded them as separate layers to my map. Apart from coloring the subway lines, I also changed the styling on the subway stops to all white, as the map was already very colorful, with the colors of the subway lines and the background racial distribution map. White would make the subway stops stand out.

The final map displays New York’s subway system against the 4 boroughs’ racial distribution. Only Black, Asian, White, and Hispanic populations are given a display color, since other racial categories’ population (eg. 2+ races, other, Pacific Islanders, Native Americans) are not significant enough to show up as the predominant racial category on the map. A brighter color on the census tract indicates a higher concentration of that racial category. When user clicks on a census tract, a pop-up appears with all the racial distribution information – including the ones that were not given a display color – in that census tract; when user clicks on a subway stop, another pop-up appears with the station name and the lines that are serviced at that stop.

To view the full extent of the map, open in ArcGIS’ Map Viewer (not Map Viewer Classic).

Results
Apart from obvious neighborhoods (eg. Chinatown) where a drastic transition is expected to take place, some of the most interesting findings include little pockets of a certain race within a larger neighborhood of a predominant race. One example of this would be small white neighborhoods in the predominantly black- and Hispanic-populated Bronx. When the #6 train goes to “East Tremont Avenue”, the passenger demographics may well suddenly become white, due to the Italian populations that settled there for long.

Other interesting observations include places where a drastic racial transition takes place. For example, on the #7 train , there is a high concentration of Hispanic population around “103rd Street – Corona Plaza”, but this intense concentration sudden switches to Asian after only 3 stops, when the #7 train ends in Flushing – where one finds the largest Asian settlement in all of New York.

Reflection
Some difficulties that arose during this project are technical: I had wanted to add user interactions where one could click on one line and go through the subway stops one by one, but this exceeds my current capabilities and the supported functions of ArcGIS Online. Another issue of this map is the use of color, which can seem too busy. I experimented around it, but eventually found that taking away certain colorful elements – subway line colors or the background racial map, for example – severely compromises the amount of information that is conveyed. In light of this, my next step for this project would be to visualize this racial transition by each line, possibly via an illustration or animation instead of a GIS software.