Authors: Dustin Smith*, San Diego State University, Adam Russnogle, San Diego State University, Kolbe Kulda, San Diego State University, Atsushi Nara, San Diego State University
Topics: Urban and Regional Planning, Temporal GIS, Economic Geography
Keywords: space-time clustering; spatio-temporal clustering; python; postgresql; tslearn; smart growth; transit; residential; city planning; urban planning; california
Session Type: Poster
Start / End Time: 11:10 AM / 12:25 PM
Room: Virtual Track 2
Presentation File: Download
Building enough dense residential housing units near existing transit is essential to meeting California’s aggressive climate goals. Localities must understand where best to greenlight these dense developments so that sufficient construction actually takes place. Our project utilized space-time clustering and database management to analyze property price trends on the individual parcel level in San Diego, California. To our knowledge, there have been no other spatio-temporal analyses like ours on the assessor’s parcel level. The project was conducted by students and staff of San Diego State University in partnership with the City of La Mesa. We developed a program using Python and PostgreSQL to group properties into k space-time clusters, which demonstrate significant similarities in their price level over the period of study. From there, the appropriate parties could factor a property’s price trajectory into their assessment calculus. Land and improved price components of parcels were analyzed separately, as well as the combined price. We cleaned and constrained the available parcel data to yield the most accurate and relevant results possible. As an added externality, the clusters reveal undervalued properties beyond those under initial consideration. Our analysis was limited to the City of La Mesa due to time and computational resource constraints, but the program can be extended to a large county level. We hope our program will lead to better investments in smart residential growth in the City of La Mesa, and will provide rich contextual information for future decision making.