Corridor model accuracy scorecard

Simulated AM-peak intersection volumes vs City of Toronto turning-movement counts. Config: outputs/battery/baseline, 3 seeds, 632 in-area counted intersections.

7.9%

GEH < 5 FAIL
target ≥ 85%

0.456

R² modeled vs obs FAIL
target ≥ 0.88

0.444

total flow ratio FAIL
target 0.95 to 1.05

19.9

median GEH

78.9%

%RMSE

GEH<55-10>10 dashed line = perfect match

Largest remaining discrepancies

Intersection	Observed	Simulated	GEH
Lake Shore Blvd W / Remembrance Dr / New Brunswick Way	5007	1203	68.26
Lake Shore Blvd W / Newfoundland Rd / Ontario Place Blvd	4835	1214	65.83
Lake Shore Blvd W / Remembrance Dr / Ontario Dr	4535	1203	62.2
Queen's Park / Charles St W	2397	176	61.9
York St / Lake Shore Blvd W	3032	523	59.51
Gardiner Expy E Spadina Ave Ramp / Lake Shore Blvd W / Lower	3698	946	57.1
Lake Shore Blvd W / Stadium Rd	4121	1299	54.2
Lake Shore Blvd E / John M Tinsley St	2432	411	53.59
Bay St / Wellington St W	1827	149	53.38
Lake Shore Blvd W / Rees St	3106	783	52.67
Avenue Rd / Cumberland St	1559	67	52.31
Jarvis St / Lower Jarvis St / Front St E	2464	505	50.82

How to read this

This model is calibrated to the best achievable with the public data available. The street network is curated to current conditions (2026): where Toronto has put in bike lanes or streetcar right-of-way, car capacity is modeled as reduced, so the model reproduces streetcar and bike corridors well. Demand is scaled to the City counts. The residual gap is structural, not a bug: the model simulates the downtown core with boundary-gate demand, so it under-represents the regional through-traffic that a wide-area origin-destination survey would capture, and it is validated here against City counts spanning 1984 to 2026, many predating recent road changes. Closing that gap to the professional bars (GEH<5 at 85%, R²≥0.88) requires a commissioned OD matrix (e.g. the Transportation Tomorrow Survey), which is the institutional-tier calibration. What is shown here is honest, standards-based disclosure of exactly where the model stands.

Standards: GEH<5 for >=85% (DMRB/FHWA); R2>=0.88 (FHWA); total flow within +/-5% (FHWA). Sites outside the modeled downtown box (5723) are excluded, they carry no modeled demand.