My intuition tells me it wouldn't even be close, but let's do the math since it's pretty straightforward.
Caltrain uses cars that are 26 meters long and 3 meters wide[1]. That's 78m^2 of area to catch sunlight. The largest trains have 6 passenger cars and one locomotive. Let's cover all 7 in solar cells, giving 532m^2 of sunlight. At noon on the equator, approximately 1kW/m^2 of energy hits the earth. The bay area is around 37ºN latitude. sin(37º) ≈ 0.6x the energy per unit area. While 600 watts per square meter sounds like a lot of power, we must also account for solar panel efficiency. The best commercially-feasible panels are around 19% efficient. That gives us 114 watts per square meter. Multiply by the total area and we have 60kW of power at noon on a cloudless day. To compare with your car: 60kW is 80 horsepower. The locomotives used by Caltrain put out 3600-4000 horsepower[2].
To sum up: 26m * 3m * 7 * 1kW/m^2 * sin(37º) * 0.19 ≈ 60kW.[3]
On a cloudless day at noon, with no tunnels or shade on the tracks, solar cells could generate 2-3% of the train's needed power. As I thought, not even close.
While you're basically correct, there's one point worth correcting: Caltrain locos need to put out 4000 hp because diesel trains built to US standards are ridiculously heavy, see eg. https://bikeeastbay.org/rail/fra.html. A modern EMU is much lighter and consequently needs a lot less power.
Yes, these suggestions of powering cars / trains with the solar panels stuck to their rooftops is ludicrous. I sometimes imagine that these "great ideas" are actually introduced & elaborated by the dirty energy companies to make the clean energy folks look like a bunch of drifting dreamers. Sadly it's probably well meaning people without any sort of science understanding.
I also think it's also well-meaning people without science understanding, or that didn't do the math. Intuition for those things is hard. While the result for this calculation is roughly where I expected it to be, I know that my intuition fails me wrt. houses - I just can't believe, without doing the math, that a typical solar installation people put up is sufficient to actually power a household, with washing machine, fridge and computer equipment.
Caltrain uses cars that are 26 meters long and 3 meters wide[1]. That's 78m^2 of area to catch sunlight. The largest trains have 6 passenger cars and one locomotive. Let's cover all 7 in solar cells, giving 532m^2 of sunlight. At noon on the equator, approximately 1kW/m^2 of energy hits the earth. The bay area is around 37ºN latitude. sin(37º) ≈ 0.6x the energy per unit area. While 600 watts per square meter sounds like a lot of power, we must also account for solar panel efficiency. The best commercially-feasible panels are around 19% efficient. That gives us 114 watts per square meter. Multiply by the total area and we have 60kW of power at noon on a cloudless day. To compare with your car: 60kW is 80 horsepower. The locomotives used by Caltrain put out 3600-4000 horsepower[2].
To sum up: 26m * 3m * 7 * 1kW/m^2 * sin(37º) * 0.19 ≈ 60kW.[3]
On a cloudless day at noon, with no tunnels or shade on the tracks, solar cells could generate 2-3% of the train's needed power. As I thought, not even close.
1. http://en.wikipedia.org/wiki/Bombardier_BiLevel_Coach
2. http://en.wikipedia.org/wiki/MPI_MPXpress
3. http://www.wolframalpha.com/input/?i=26m+*+3m+*+7+*+1kW%2Fm%...