A few additional stray thoughts on pair-matching northside and southside.
The key here is that we want to match like-for-like services, so you can use the same equipment and frequencies on both legs. I see the (future) regional rail network broken up into three buckets.
Within-128 Metro Services (Indigo Lines)
These are stretches that we want to create rapid-transit-like service. 15 min headways max, but if we can get below 10 min, all the better. The prototypical example is the
Fairmount Line, with an optional extension to Dedham Corp Center. The
B&A to Riverside is a similar example, although I am skeptical that the short-turn service to Riverside could see headways better than 15 minutes; other stations would receive additional service from Framingham trains.
The northside doesn't have any direct equivalents to the Fairmount Line, but the
Eastern Route to Salem or Peabody comes close (as a complement to an extended Blue Line,
not as a substitute). The
Fitchburg Line to Waltham or Weston/128 is a strong contender for 15 min headways, particularly given Waltham's usefulness as a reverse commute destination. The NH Main Line to Anderson RTC is a weaker contender, as it goes through a pretty significant density gap between Winchester and Woburn, and also because it will see several longer-distance routes layered on top of each other. (See below.)
On the face of it, the Reading Line is an obvious contender for a Metro service. However, Metro service through the NSRL will require electrification, and I believe that infrastructure improvements on the Reading Line such as electrification are better spent on an extension of the Orange Line. Additionally, the bottleneck on the mainline tracks next to the existing Orange Line will hamper frequencies.
So, the southside sees 4-8 tph on the Fairmount, and perhaps 4 to Riverside. On the northside, you have Salem/Peabody and Waltham/Weston, both of which could easily merit 4 tph, neither of which necessarily is screaming for 8 tph... but since we are only talking about two southern legs and two northern legs, you could split the difference, and have 6 tph on each northern branch, splitting up 2-to-1 on the southside weighted toward Fairmount. Or you could run a trio of overlapping services:
- Dedham-Waltham (4 tph)
- Readville-Lynn (4 tph)
- Riverside-Peabody (4 tph)
Giving Readville-Downtown-Lynn headways under 10 minutes, providing a one-seat ride from the North Shore to Back Bay and Longwood, and giving Dorchester riders reverse-commute options to both Lynn and Waltham.
These examples are purely illustrative, and the details would still need to be worked out, but it should be very feasible to pair-match the Metro services.
Suburban Services
These are outside of 128 but still close enough to merit 30-min headways all day, per the Regional Rail model TransitMatters proposes. I think there are some modestly clear pair-matches as well here -- you'll notice some conspicuous absences, which I'll get to below.
The
NH Main Line to Lowell and Haverhill matches pretty well with the
Northeast Corridor to Providence and Stoughton/Taunton/South Coast. Major cities form strong anchors and reverse-commute destinations, with main lines that diverge about 14 miles outside of downtown, and which will see limited stops inside of that merge: Canton Junction, Route 128, Readville, Ruggles on the south (I love TransitMatters, but am very unconvinced about Forest Hills and other intermediate stops), and Wilmington, Anderson/Woburn, Montvale, Winchester, and maybe West Medford. There is some rough parity in travel times -- Lowell is the closest terminal in a major city on the network, and Providence benefits from long straightaways that run pretty much as the crow flies. And while more dispersed, Taunton, New Bedford and Fall River's ~250K residents are roughly equal in number to those in Haverhill, Methuen, Lawrence, and the Andovers; Lawrence is exceptionally dense, but otherwise the SCR cities are roughly as dense as Haverhill and Methuen, again suggesting that it would be relatively easy to pair-match based on demand.
Remaining on the southside are the B&A and Franklin Lines, and on the north we have Fitchburg and Eastern Route. The pair-matching is less ideal here: the Eastern Route and the B&A have roughly similar ridership, but the Fitchburg Line is a bit lower than the Franklin; based on ridership, we might suggest those pairings. On the other hand, the Eastern Route and the Franklin Line share similar profiles with urbanized inner halves and lower density outer halves; Worcester and Fitchburg aren't as similar, but do both have anchor cities at the far end, along with significantly circuitous routes to Boston.
I don't think there's as obvious a conclusion with these four, but I also think we can be reasonably confident that a solution is feasible, and will depend on the particulars of what service looks like at the time. (For example, I can see a strong argument for running Metro services past Dedham into Norwood; at that point, a
Suburban Franklin service would pair match well with
Fitchburg in terms of profile and frequency, which would leave
Fram/Worcester to match with
Rockburyport -- an all-the-more parsimonious match if a Marlboro Branch is added: to use some old analogy notation, Worcester : Marlboro : Framingham : : Newburyport : Rockport : Salem.)
Long-Haul Services
These routes are long enough that I can't ever see them being through-routed across the city. Most of these are proposed extensions at this point. Of the current routes,
South of Providence is a pretty clear candidate for this group, and I would argue that local services from Worcester and Fitchburg would be as well. Proposed extensions to
Buzzard's Bay, Cape Cod, Woonsocket, Milford, Plaistow/north and
Portsmouth would also be unlikely to ever through-run; a Nashua extension might, as potentially could a Manchester, but a Concord extension wouldn't.
Fall River & New Bedford are a little odd. As I recall the travel times, I think they are weaker contenders for through-running. If there are Stoughton or Taunton short-turns available, then probably those would through-run and FR/NB trains would terminate at South Station.
Exceptions
The Needham Line should be converted to rapid transit via extensions of the Orange and Green Lines and removed from the Regional Rail network.
The elephant in the room here is my conspicuous absence of the Old Colony Lines.
One reason is physical: the Old Colony would require its own dedicated portal, adding significant cost to the project. [EDIT: This is not true, as F-Line points out below. The cost of adding a portal for the OCR is
not significant in the overall context of this project.] Another reason is network-based: the Southside has more lines and more ridership than the north, so some lines never will run through. If you subtract the ridership of the OCR Lines, northside and southside ridership is roughly equal. To me, the portal issue then makes it tough to justify. But there's a third piece of the puzzle here...