So the main tool that I use to evaluate service priority is an inverse square or gravity model. To calculate, you take the product of the populations and divide them by the distance squared. We'll take the Lowell Commuter rail as our example because it would be the terminus that would extend to Nashua, Manchester, and maybe Concord.
We'll use this as our baseline. The units here are gibberish (people^2/sqmi) but we can proportionally compare these numbers to get an approximation of the relative demand. Now, lets add Nashua and Manchester.
Distance to Boston: 50 miles
Now, when we're proportionally comparing these numbers, you can think of them as ratios of estimated trips. For each 160 trips to Lowell, you might reasonably expect 53 to Nashua and 31 to Manchester. That said, we would be potentially looking at a 55% (88/160) increase in riders compared to Lowell.
55% is a lot, but we do have to factor in increased costs. Certain costs are per mile. These per mile costs would increase by 127% (28 new track miles / 22 current track miles) Other costs are fixed or variable in different ways, and I don't have great tools to evaluate them. What I can say is that if this were to be a practical reality, the federal government could also be called on to help as it would be an interstate transportation project to offset some of those costs as well.
But, just with what we have here, lets compare with each other terminal stop on the network.
Rockport & Newburyport: 5&11 ***
The direct comparison here shows in large part why such a service would be consistent with other commuter rail lines. Many of the gravity scores here are lower than what would be expected of a Nashua Connector, and several of these service areas are comparable lengths as well. New Bedford, Worcester, Providence, and Fitchburg all make journeys of close to 50 miles in under two hours on their schedules, matching the distance to Manchester.
There is another thing that the combination of Nashua and Manchester highlights which also is clear in the abnormally low scores for Rockport and Newburyport. On each line, population centers will have mutually reinforcing gravity scores. Beverly, Salem, Swampscott, Lynn and Chelsea all generate trips to Boston but also generate trips to each other, giving the following table.
Now, some of these numbers look a little absurd. How is the 1515 lynn-swampscott score to be read in context? Well, Lynn center and Swampscott center are only a mile away from each other because swampscott is a continuous suburb of Lynn. Therefore, the economic interaction between these two populations is expected to be exponentially higher because of their proximity. A statistically significant number of someones in swampscott aren’t going into boston on a weeknight dinner, but they are going to lynn. When they travel from one to the other, they do it in cars, buses, bikes, and on foot, because the distance is short enough to make all those transit modes viable. Trains realistically make up the smallest part of that 1515 because other modes are more convenient. We don’t need to try to change this. Even setting reductions in driver miles as a goal, buses will solve the problem better than the commuter rail could as a static piece of heavy rail infrastructure. It doesn’t have the dynamic range buses do. The 1515 represents an estimated proportion of economic interactivity that drives and can be used to measure transit demand. So does Rockport’s 5. Even if everyone in rockport leaves rockport on the commuter rail to boston every day, that’s still only 7k people compared to lynn’s 101k and those people are commuting much farther, incentivizing switching to a job with a shorter commute. Nobody likes their 90min commute.
Now, even if we can’t count 100% of those gravity scores as demand for more trains (because more convenient mode shares outweigh trains in practice over small distances), the train still benefits from high trip demand between cities that aren’t the terminus, and many of these city pairs are in fact excellent train connections. Salem-Lynn and Beverly-Lynn score comparably to Boston-Salem and Boston-Beverly, and are in a distance range where walking, biking and buses are less practical, so the train reasonably serves many more people. It’s entirely reasonable to assume that these network effects will also boost ridership for a manchester lowell connection.
When we look at the manchester and Nashua numbers, we get this chart:
These numbers aren’t as high as the north shore, which is to be expected based on the distances, but they’re reasonable compared to the rest of the network shown below.**** The main exception here is Concord. The only score that Concord has above 10 is with Manchester, and it would be the terminus of the line. It’s 20 extra miles of track, almost as long as he initial segment to lowell in the first place. Why would we extend it all that way?
Well, Concord is the capital of New Hampshire. Capital cities drive business traffic in excess of their populations because certain business can only be done there. The gravity model only takes in population and distance, so it can’t reliably assess a variable like that. I like it because it’s fast to calculate and easily comprehensible, but the gravity model is not the best or most sophisticated tool for assessing transit demand. Additionally, if Massachusetts wants to connect population centers in New Hampshire to MA, they would most likely ask New Hamshire to help, and a railroad connecting the capital to its most populous cities would be a project that would mostly benefit New Hampshire. Having them pay for it would be more than reasonable. Then, the actual terminus of the line would be decided by representatives of the actual riders rather than someone in boston. To be clear, I think either terminating in manchester or concord would be reasonable with clear advantages and disadvantages, it just seems like the sort of thing that should be decided democratically by NH itself.
*Geographical distance is substantially shorter than the actual by-car routes, so error in gravity scores should skew actual figures down.
**The abnormally low Gravity score here is partially explained by a large financial contribution from Wachusett mountain resort, which funded an extension of the line to connect to its bus shuttles
***This abnormally low gravity score is because service is extended to Rockport and Newburyport from Salem and the rest of the populous North Shore. Salem alone has a score of 151.
****I’m going to clean up the data because I grabbed numbers fairly sloppily and publish my spreadsheets with consistent population and distance measurements when I get the chance. These are extremely rough numbers with a spur of the moment methodology but the point that they make is worth investigating further. Once I have more consistent numbers I’ll update the numbers in the original post as well.
areudirty
turnnoffyourmind
izzy806
andulkaphoto
gamechangershow
vaporwave
amandla
tangentwithtanner
