User:David Schneider
Transit-Oriented Discoveries: Data Driven Insights into U.S. Urban Landscapes
Introduction
My research investigates the prevalance and characteristics of Transit Oriented Development (TOD) across the United States by leveraging open data sources and analytics. It aims to uncover patterns in land use, urban form, transportation networks, and civic infrastructure within accessible distances from these stations. Utilizing recent datasets from the Federal Transit Administration’s National Transit Database, OpenStreetMap, and the Census Bureau, the research addresses a significant gap in national-level TOD analysis, offering fresh insights since the last comprehensive study a decade ago. The core research questions explore the presence and extent of dense, walkable, and mixed-use developments around rail stations and identify key factors contributing to high levels of TOD. Through a combination of exploratory data analysis, feature creation, and advanced clustering techniques, the study classifies station areas by their TOD characteristics and assesses changes in local demographics and infrastructure.
Methodology
I developed Python scripts to query the OSM Application Programming Interface (Overpy) for areas within a circle with a quarter-mile radius around transit stations in the United States. Scripts include a function that calculates the area covered by selected land features and counts the number of buildings. This method allows for a detailed analysis of land use near transit stations. Specific land uses selected include building footprint, surface parking and parking garages, highways, primary roads, residential, commercial, and retail areas, and a variety of natural features such as parks, woodlands, and water. (A complete list of features is included in the report and code). In addition, I developed scripts to count the number of intersections, number of street segments, and average street segment length and also identified the presence or absence of selected amenities, and the total number of amenities within these station areas (selected amenities included grocery stores, libraries, community centers, theaters, cinemas, pharmacies and other elements of civic infrastructure).
Results
OpenStreetMap data was used to create various indexes and scores that can be used to evaluate the presence of TOD and potential for additional development near transit. They include a walkability score that is based on OSM road typology and street network characteristics, a TOD score based on the number of buildings, percentage of an area that is covered by buildings, walkability, presence or absence of an educational, commercial, or retail land uses, and number of amenities. A TOD potential score is also included and is based on the amount of surface parking, proportion of land in the station area devoted to natural features, and proportion of land that is covered by buildings.
Discussion
This research analyzes data and shares insights on a topic that many people who live in cities and towns feel strongly about: the built environment and our place in it. It references a dataset of approximately 3,900 station areas representing the vast majority of transit fixed guideway stations in the United States. Each station area includes features related to land uses, transportation characteristics, and amenities/civic infrastructure. Stakeholders who live or work in a particular area can take advantage of the raw datasets to delve into TOD characteristics, housing, and demographics in their community. Further research would take advantage of additional OpenStreetMap data and data mining and machine learning capabilities. For further discussion or access to the raw data and Python code, please contact David Schneider at schneiderd41@gmail.com
External Links
A copy of my research paper and Jupyter Notebooks used can be found on my GitHub repository at https://github.com/DavidSchneider47/transit-oriented-discoveries