Does the data for LA Metro's A Line show that some stations can reasonably be skipped for an express service from Long Beach to Los Angeles?

     LA Metro's A line runs along the old pacific electric line right of way, with people traveling along the corridor just as they did 100 years ago. In fact, the old pacific electric line from 100 years ago was faster than the current A line. If you look at this old timetable from the Long Beach- Los Angeles pacific electric line you'll see that it only took 48 minutes to get to the pacific electric station in downtown Los Angeles compared to 55 minutes for the A line today.

    Notice that they also had "Limited" trains that skipped several stations on the way to Los Angeles. This shaved the total travel time down to around 42 minutes.

    100 years later, the travel times on the A line are worse than they were on the Pacific Electric Line. Though we do have twice the train frequency that the Pacific Electric Line had and we also serve about twice as many stations, perhaps we should consider whether it is worth running express trains on the A line to get travel times closer to what we had in the early 20th century.

    In order to look into the feasibility of having a "limited" train, I requested some data from LA metro about both the ridership data and the station locations. Using both the station locations and the published time table data for LA metro's A Line, I created a map with power bi and python that shows the average travel speed between stations, highlighting the sections that are slowing down the trip.

 

    From this map we can see that the sections that are slowing down travel times on the A line are the street running sections near Long Beach and Los Angeles. Increased street traffic in these areas in the past 100 years could explain the increased travel time between Long Beach and Los Angeles since these parts of the line are street running. 

    Travel times along the central part of the line from Willow St station to Slauson station are actually quite good. By encouraging the development of housing and job centers along this section of the line, there is the potential for shorter commute times and quicker trips in the region. Additionally, since this section makes up the majority of the A line, any potential improvements to speed up trips only need to occur on the shorter parts of the line. The street running section going into LA from Slauson especially would benefit from grade separation as about 20% of the ridership activity on the line occurs at the 7th / Metro station but the average travel speed to/from the station is much slower than the rest of the line.

    However, without any funding for improvement projects eliminating stops could save a significant amount of travel time for commuters without needing as much money. Lets do a quick calculation on how much time could be saved by skipping a single station. The maximum allowed speed on the A line is 55 mph. In the best case scenario let's assume that for each stop the train is decelerating from 55 mph to 0 mph for each stop and then accelerating back to 55mph. The Kinki Sharyo trains that the A line uses can decelerate at a rate of 3.5 mph/s and can accelerate at a rate of 3 mph/s. Under these conditions it will take 15.7 seconds to decelerate for a single stop and then 18.3 seconds to accelerate back to 55mph. The train ends up taking a total of 34 seconds just decelerating and accelerating at each stop. Compare this to the case where the train does not decelerate and continues traveling at 55 mph. In this case it takes a total of about 17 seconds to travel the same distance. This can be calculated by using the equation of motion to find the distance spent decelerating/accelerating then using this distance to calculate the travel time at 55mph.

    Therefore it will be up to about 17 seconds spent at each stop, not including the dwell time of the train at the station. I requested data from LA Metro on the dwell time for each station, which they confirmed to be 20 seconds at each stop. Thus by skipping a stop on the A line we can shave off up to about 37 seconds of travel time spent at each stop.

    To keep most of the ridership on the line, we should keep the busiest stations on any potential express train. Here is the relative ridership for each station on the A line that I calculated from LA Metro data:

 

    Clearly, any express service should definitely include the Willowbrook / Rosa Parks Station and the 7th st/ Metro Center Station as just these two stations make up over 30% of the line's ridership activity. By keeping just the Downtown Long Beach, Compton, Willowbrook / Rosa Parks, Florence, Pico, and 7th / Metro stations, we can see from the ridership data that this would keep over 50% of the ridership on the line while keeping the least amount of stations. This route would eliminate up about 16 stations from the A line while keeping much of the ridership. 

    By eliminating 16 stations this route could shave up to 9.87 minutes of travel time from the route assuming that the train is accelerating and decelerating from 55mph at each stop. Even ignoring the time spent accelerating and decelerating which will vary based on the actual conditions of the route, just eliminating the 20 second dwell time at 16 stations will save at least 5.3 minutes on the trip from Long Beach to LA. So we could run some express trains on the A line that maintains the majority of the line's ridership and would save commuters between 5-10 minutes in travel time. 

    However, these express trains would fail to serve a significant portion of riders on the line for only a few minutes in travel time saved. But this might be worth it for a lot of riders so maybe the idea should be considered. Another idea to speed up the travel time on the line that would still serve each stop is called a skip-stop train. With a skip-stop train every other station is skipped and the next train serves the stations that were skipped. Under this arrangement every station is still served with the benefit of saving 3.67 to 6.78 minutes of travel time for the riders of the line. But this will result in trains only arriving every 20 minutes instead of every 10 minutes like they do now. Additionally, one of the skip-stop trains would skip the important transfer stations such as Willowbrook / Rosa Parks station and 7th / Metro station that a third of the riders need to get to for transferring to other lines. So a skip stop train system would be pretty uneven in ridership activity. 

    Overall I think that there are a lot of potential improvements that LA metro could make on the A line, some of which could be cheap to implement but could come at a cost to ridership and frequency. The A line has a lot of potential as a decent rapid transit line but it needs the investment from LA metro to improve the travel speeds on the ends of the line and the county needs to encourage residential and commercial development along the central section that already has incredibly quick travel times. LA metro should also look at investing in increased train frequency as that would negate a lot of the downsides of a skip stop system while also improving travel times by about 4-7 minutes.