Methodology: Proximal Resident Population

To estimate the possible frequent use population incident on a station, I'm using data pulled from 2010 Census Block Groups within a 2-block radius of the center of the station. Take for example a proposed station centered at Leavenworth and Sutter streets in the TenderNob neighborhood of Downtown San Francisco.

First, locate the centerpoint of the station.

Next, draw a circle two blocks in radius (using streets not alleyways) along the major axis of the metro line at this point in its run (two blocks north and south along Leavenworth in this case).

Sum the population of all census block groups partially or completely encompassed by the circle.

In the case of this station, the following block groups are incident on the circle:
111bg2 - population 2,084
119.01bg1 - population 898
111bg3 - population 914
119.01bg2 - population 1,510
120bg1 - population 1,983
121bg1 - population 2,725
120bg2 - population 1,850
122.01bg1 - population 2,699
123.02bg1 - population 1,763
123.02bg1 - population 1,310
Population of proximal block groups — 17,736.

For the purposes of this comparison, an exact science is not what I'm looking for; after all it has been proven time and time again that "ridership estimations" are sorely lacking exactitude. Take, for example the BART SFO extension, originally estimated at 17,000 daily trips but only providing 20-30% of that amount in daily ridership during its first year of operation. Rather than try to divine how many people will use each station, I'm only identifying how many people are within an easily walkable distance (under seven blocks in this case) and could conceivably make use of the station with regularity.