Reproducible methods for extracting spatio-temporal data about urban fabrics on a large scale

Keywords: openstreetmap, pyrosm, open data, open software, open science, reproducibility
Session Type: Virtual Paper Abstract
Day: Saturday
Session Start / End Time: 2/26/2022 09:40 AM (Eastern Time (US & Canada)) - 2/26/2022 11:00 AM (Eastern Time (US & Canada))
Room: Virtual 34

Abstract

OpenStreetMap (OSM) has established itself as a central data source for geographical analysis, used extensively both in academic research and business. As a free, crowdsourced digital map and database of the world, it provides crucial geographical data that are commonly used for providing useful services for people, such as navigation, and analyzing various urban phenomena.

In this talk, we will introduce a new open source Python library called pyrosm that enables extracting various datasets from OSM and convert that data into geopandas GeoDataFrames. Geopandas is the core library for modern geospatial analysis in Python with a broad ecosystem of spatial analysis libraries built on top of it. Pyrosm has been optimized for performance allowing large-scale analysis based OSM data extracts covering city regions or even countries. The library also supports extracting historical data from OSM (OSH.PBF format), which enables conducting spatio-temporal analysis, such as detect changes in urban areas. The historical data extraction functionality also allows full reproducibility in scientific papers, because the authors can pinpoint to a specific moment in OSM history and extract the data layers as a snapshot of given moment in time. Hence, it can be a useful package for urban researchers and practitioners alike. The tool is distributed via PyPi and conda-forge, and the documentation is available at: https://pyrosm.readthedocs.io/