Table 1.

Granular land-use categories from OpenStreetMap allow for more detailed understandings of traffic disruptions. Compared with the traditional land-use categories shown in (a) that produce an adjusted R² = 0.11, the granular classifications used in (b) increase the adjusted R² to 0.55. Only a small subset of the 40 predictor variables are shown for (b), with all other coefficients shown in Table S3 in the Supplementary Information. Respectively, *, ** and *** indicate p < 0.05, p < 0.01 and p < 0.001.

Variable	Estimate
(a)Meta-categories only
Residential	− 0.09**
Industrial	− 0.18**
Recreational	− 0.10*
Institutional	0.14*
Green space	0.26***
Commercial	0.32***
Observations	6529
Adjusted R2	0.11
(b) Granular model
Residential	0.61***
Farmland	0.56***
School	0.042**
Place of worship	0.009**
…
Apartments	− 0.09**
Observations	6529
Adjusted R2	0.55