Every factor, source, and assumption behind the numbers, out in the open.
When you mark a ride as a Commute and tell us what transport you replaced, Riding Places credits the CO2 you avoided by riding instead. The calculation is deliberately simple and auditable:
CO2 avoided (kg) = distance (km) × emission factor (kg CO2 per km)
The emission factors come from standardized government datasets and follow the GHG Protocol for employee commuting (Scope 3, Category 7). We publish them here in full, drawn from the same source of truth the app calculates with - so this page can never silently disagree with your rides.
us-v1Factors for United States (kg CO2 per km). Each row lists the exact published figure it derives from.
| Transport replaced | kg CO2/km | Published source |
|---|---|---|
| 🚘 Car or truck (not known) | 0.219 | A distance-weighted blend of EPA "Passenger Car" and "Light-Duty Truck" (0.21920 kg/km), using EPA’s own published vehicle-miles as the weights. |
| 🚗 Car | 0.185 | EPA "Passenger Car" (0.184547 kg/km) - every car under 8,500 lbs, so this is the US average car (EPA publishes no separate small-car figure). |
| 🚙 SUV or truck | 0.245 | EPA "Light-Duty Truck" - SUVs, minivans, and pickups. |
| 🚌 Bus | 0.041 | EPA "Bus", per passenger-mile (converted to per-km). |
| 🚆 Subway / transit rail | 0.058 | EPA "Transit Rail" (subway / light rail), per passenger-mile.EPA lists three rail types and we map to a single option. If you actually ride a suburban commuter railroad (for example Metro-North, LIRR, or Caltrain), this credits you about a third less than EPA’s separate commuter-rail figure - we chose the more conservative row. |
| 🏍️ Motorcycle | 0.11 | Motorcycle, per mile.We do not use EPA’s current published motorcycle row. That figure carries a known upstream data artifact that makes it implausibly high; we use the corrected value, which is also the more conservative choice. |
When you choose the transport yourself, we use the factor for that mode in your home market. Riding Places is not location-aware for this on purpose: "what would you otherwise have driven?" almost always means your vehicle, which is at home - and asking your phone for its location fails exactly when you are traveling.
When a ride is classified as a commute automatically (because it passed through a workplace GeoZone your group set up), we never got to ask which vehicle you would have driven. So we use a car of unknown size - the figure the data publishers provide for exactly that situation. It is an honest "we don't know your car", not an assumption dressed up as precision. These rides use the market of the group that owns the workplace zone, since that zone is a fixed place. You can always open the ride and set the real mode yourself; once you do, we keep your choice.
Each ride is stamped with the exact methodology version used at the time (us-v1 today), and we never go back and recalculate old rides. If we adopt a newer dataset, only rides recorded afterwards use it. Most tools quietly restate history when their numbers change; we don't. The honest trade-off: after a version change, a long-running group total can blend two methodologies - which we would rather disclose than hide behind a silently-restated number.
Wherever we had to choose between a higher and a lower figure, we took the more conservative one - so if anything, these numbers understate rather than overstate what you saved.
Source: EPA GHG Emission Factors Hub, January 2025, Table 10. Methodology version us-v1, source last reviewed 16 July 2026. We pin the exact edition of each dataset rather than tracking "latest", and re-verify it before adopting a new one.