Greenhouse Gas Inventories
56
Route Length and Frequency Calculations
To estimate route lengths and frequencies, MAPC downloaded General Transit Feed Specification (GTFS) schedule data archived
on the MBTA’s website.
1
The MBTA publishes a GTFS feed every time there is a known change in service, which means there are
multiple versions for a single year. For example, in 2017, the MBTA published 48 GTFS feeds.
Each of the MBTA’s GTFS feeds consist of 22 tables, of which MAPC utilized six for our calculations:
• The Calendar table defines dates when service is available for particular routes. This file specifies start and end dates, as
well as the days of the week when service is available.
• The Stops table defines the locations of stops on the network by latitude and longitude
• The Stop Times table defines the times that a vehicle arrives at and departs from stops for each trip.
• The Trips table defines trips for each route. A trip is a sequence of two or more stops that occur during a specific time
period.
• The Shapes table provides a geospatial representation of the path followed by the transit vehicles on each route
• The Routes table provides additional descriptive information on each route
The MBTA aggregates GTFS feeds into four seasonal feeds per year (winter, spring, summer, and fall). MAPC combined the data
from each seasonal feed as described below.
MAPC joined the Calendar, Stops, and Stop Times tables by Service ID field, and then joined this table to the Trips table using
the Trip ID field. MAPC calculated the total number of days in the year that each route was in service using the start and end
dates from the Calendar table and the days of week the service is running. We then calculated the total number of annual trips
for each service ID by multiplying the number of trips per Service ID per day from the Trips table with the number of annual
operating days.
Vehicle Miles Traveled Calculation
MAPC used the Shapes tables to calculate the length in miles of each route. However, not every trip on a route covers the
entire length of the route defined in the Shapes table. Many trips cover only partial distances of the route, depending on day of
week and alternate service schedules. MAPC used the Trips, Stops, and Stop Times tables to identify which trips travel the full
length of the route, and which trips travel only a part of the route, by examining the origination stop and ending stop for each
trip. For the trips that do not travel the full route distance, MAPC used the Google Distance Matrix API and Google Directions
API to obtain the transit mode travel distance between the originating stop and the ending stop locations listed in the Trips
table. MAPC also export a shapefile for each of these partial-route trips, in order to allocate them properly across municipal
boundaries.
To obtain annual vehicle miles travelled, MAPC multiplied the annual number of trips for each route by the route length in miles.
This yielded an annual vehicle miles traveled figure for each route which could be used to allocate system-wide consumption.
Commuter Rail
To estimate commuter rail emissions by municipality, MAPC first allocated the system-wide diesel fuel consumption total for
commuter rail for the 2017 calendar year
2
to individual routes by multiplying the fuel consumption total by a ratio of annual
vehicle miles traveled for each commuter rail route to annual vehicle miles traveled for all commuter rail routes. MAPC
then multiplied the consumption total for the commuter rail route by a ratio of vehicle miles traveled for that route in each
municipality and the sum of vehicle miles traveled for the route.
Heavy Rail, Light Rail, and Trackless Trolleys
Heavy rail, light rail, and trackless trolleys are all powered by electricity, so, as a first step in estimating emissions by
municipality, MAPC allocated the system-wide electricity figure for transit (60% of the total system-wide figure) to each
electricity-consuming transit mode. MAPC based this distribution on vehicle miles traveled. In the 2017 calendar year, 78.2
percent of total vehicle miles traveled was from heavy rail (red, blue, and orange lines), 19.8 percent was from light rail (green
lines and the Mattapan Trolley), and 2.0 percent was from trackless trolley lines (bus routes 71, 72, 73, and 77).
3
1 https://cdn.mbta.com/archive/archived_feeds.txt
2 We counted both revenue (ULS Diesel for Commuter Rail) and non-revenue commuter rail diesel (Non-Revenue Diesel for Commuter Rail)
towards commuter rail consumption.
3 This calculation assumes that electricity usage per vehicle mile traveled is roughly similar between heavy rail, light rail, trackless trolley, and
Silver Line. Ideally, this value would be further weighted by an electricity efficiency factor.