Title: Acquisition and analysis of crowd-sourced traffic data

Abstract: Crowd-sourced traffic data offer great promise in environmental modeling. However, archives of such traffic data are typically not made available for research; instead, the data must be acquired in real time. The objective of this paper is to present methods we developed for acquiring and analyzing time series of real-time crowd-sourced traffic data. We present scripts, which can be run in Unix/Linux like computational environments, to automatically download tiles of crowd-sourced Google traffic congestion maps for a user-specifiable region of interest. Broad and international applicability of our method is demonstrated for Manhattan in New York City and Mexico City. We also demonstrate that Google traffic data can be used to quantify decreases in traffic congestion due to social distancing policies implemented to curb the COVID-19 pandemic in the South Bronx in New York City.