I don’t know how familiar you are with California geography, Omniscient, but here are three maps:[ul]
[li]City/road map. [/li][li]Population map.[/li][li]Relief map.[/li][/ul]If you’re building a line from San Francisco to Los Angeles, you have two choices of terrain:[list=a]
[li]Coastal route, mainly following the same route as US Highway 101 and The Southern Pacific Railroad’s Coast Line. The problem is that you have to have lots of curves to get around the mountains, and HSR trains hate curves (tilt trains are an option when one can’t avoid curves, but they are heavier, more expensive, and slower than true TGV-style HSR trains). There’s one unavoidable hilly section at Cuesta (north of San Luis Obispo), but otherwise it’s pretty flat. Still, twisty.[/li][li]Central Valley route, taking advantage of the relatively cheap land and pancake-flatness of the Valley (shown in green on the relief map). There are two sets of hills to traverse – the Diablo Range separating the SF Bay Area from the Central Valley, and the Tehachapi Range between the Valley and the LA Basin. However, TGV-style HSR trains don’t much mind shallow elevation changes, being very light and having a massive power/weight ratio. Also, much of the electricity expended getting them up the hill is regenerated going down the other side. A tunnel near the summit (not cheap, but certainly done all the time in HSR systems) can mean that the train is scarcely slowed down by the mountain range.[/li][/list]The above considerations make a Central Valley routing by far the best choice for the SF<-> LA route.
Okay, moving into finer detail, there are three main areas where choices have to be made:[ol]
[li]SF Bay Area <-> Central Valley. A trans-Bay crossing from downtown SF to Oakland would be too expensive for Phase I, so the line will go down the SF Peninsula using the Caltrain right-of-way (which will be electrified and upgraded, thereby benefitting local commuters as well as HSR riders). There were two options to get through the Diablo Range – either via the Altamont Pass via Livermore (i.e. same as I-680), or via the Pacheco Pass eastwards from Gilroy (i.e. same as Hwy 152). For reasons that are too complex to go into here (but I’ll be happy to expound further on request), the Pacheco Pass has been chosen – this was a very difficult decision, and may yet be overturned.[/li][li]Central Valley. The fastest and cheapest route would just go down the western side of the Valley, next to I-5. However, the major cities of Fresno and Bakersfield are only a few miles east, and are growing by leaps and bounds. Putting HSR through their downtowns won’t add a great deal to the cost or the SF-LA travel time, and brings in lots more riders (and voters!). [/li][li]Central Valley <-> LA Basin. For environmental and geological reasons, the Tehachapi Pass was chosen rather than The Grapevine used by I-5. The easterly routing also lines up well with a stop at Bakersfield, and brings in riders from the Antelope Valley.[/li][/ol]Having several potential city-stops on an HSR route isn’t a disadvantage; such stations would be built with fast “through” tracks so that express trains can pass more local trains without slowing down. So, service would be a mixture of nonstop SF-LA trains, limited-stop expresses, and stopping “locals”. This method serves the maximum number of passengers (and taxpayers!) while still allowing high speeds. So, SF-LA would be 2h38min nonstop, but you’d also be able to get from Fresno (metro pop >1million) and Bakersfield (metro pop 750k) to either SF or LA in less than 90 mins. Since the majority of population growth is happening in the Central Valley, it makes sense to include the Valley cities as HSR stops.
[On the Population map linked to above, one can see the LA megalopolis on the coast in Southern California, with San Diego to the southeast on the Mexican border. The SF Bay Area is two-thirds of the way up the coast with its 7 million people. Going NE from there one finds Sacramento, then in a line to the Southeast are the red blobs of Stockton, Modesto, Fresno, and Bakersfield – millions of customers in a straight line, on flat ground. Too good for an HSR planner to pass up on, and the French would do exactly the same thing with the same choice. If, decades from now, the CA-HSR line is so successful that capacity is stretched, a new West Valley bypass following I-5 might be worth adding on.]
Sacramento branch: Although it might seem logical to build directly from SF to Sac, geography gets in the way again. A tube or bridge crossing the Bay from SF adds a huge amount to the cost, and then there’s still the Diablo Range to traverse, unless one goes along the East Bay shore (like the current Amtrak trains do), but that’s longer and much twistier (bad for HSR). Since the main LA-SF HSR line will already be going through Fresno, it’s much cheaper and quicker to build a spur totally within the Central Valley. In the future – again, if HSR proves to be successful – there will probably be a line from SF to Sac via the Altamont Pass, making the northern part of the network a triangle.
as far as the Anaheim and San Diego extensions are concerned, I think that’s limited by geography (hence track curve radius) and land costs along the Pacific Coast, hence the inland route to San Diego. I’m not as familiar with Southern California as with Northern, however, so I’ll try to look into it some more.
By the time we’re all clamoring for HSR it’ll be too late. It’ll take years to build and be anti-climactic when it arrives because of the hype that will surround it. Other things that might not be as ideal will fill the void and become acceptable as an alternative in the meantime and we’ll be stuck with half-assed solutions to problems we should have seen coming.
