Not perfect, but nothing ever is. Approximately one annual geotagged tweet within 300 feet of an intersection per 100 people crossing the intersection each day. R=0.80699. A 270-foot radius would have had a slightly better correlation, it turns out (R=0.8094).
Or perhaps it is actually 109, with an ambient level of 15 tweets that are posted even when there are no pedestrians.
Pedestrian counts thanks to the Toronto Traffic Safety Unit, Traffic Management Centre. Tweet data from the Twitter streaming API.
Previously: one geotagged tweet per 22 daily transit riders at elevated rail stations, suggesting that people must spend 5 times as long waiting for the train as they do on the street.