Almost certainly this was the scenario. A visual approach cleared to land on 28R and ahead are two lit strips the left one must be 28L so the right hand one is 28R got manky lighting on it though... and are those lights work in progress on the runway? Ask tower to confirm cleared for 28R

See this incident at Gatwick :
"(i) Runway 26R was clearly visible throughout the approach but the pilots looked for and selected a pattern of lights to the right of it because they assumed erroneously that 26R was in fact 26L and they knew that the designated runway had to be to the right of this. "

https://aviation-safety.net/wikibase/wiki.php?id=147090

It seems that there is a recurrent human factors problem when a clearance is to parallel runways the crew will attempt to identify both (all) runways and therefore which one to land on. If the airport for whatever reason has lights off or dimmed on the other runway but visible brighter lighting on the parallel taxiway then crews who do not fly regularly to that airport may become confused.
Unfortunately, this has become more apparent with LED lighting where blue and green lights can appear white at a distance at high brilliance. (As in the case of a Delta 767 at ATL).

It is easy to dismiss these errors as inattention or stupidity but faced with what is effectively an optical illusion many crews have made mistakes. It is still happening so perhaps more needs to be done. From a report 10 years ago:

"As of August 23, 2007, 267 such events have occurred at 110 airports in the United States."
From: http://www.airtech.tc.faa.gov/safety...ds/TN07-54.pdf
DOT/FAA/AR-TN07/54
IDENTIFICATION TECHNIQUES TO REDUCE CONFUSION BETWEEN TAXIWAYS AND ADJACENT RUNWAYS

Nope, they confirmed it..

AC759: "Tower, just want to confirm. This is Air Canada 759. We see lights on the runway there. Across the runway. Can you confirm are we cleared to land?"

On my first solo there was a runway incursion while I was on short final. I saw it, queried my landing clearance with the tower (I was indeed cleared), and decided to go around. I'm betting that, even if the tower didn't issue a go-around order, the pilots who have a little more than my 12-14 dual hours at the time would have decided to do the same.

Granted an A320 moves a little faster than the Katana I was in. :)

Do we know what height the GA was initiated at?

Wider view showing where the ACA intercepted the taxiway extended centreline:

http://www.avgen.com/ACA759(2).jpg

San Mateo bridge in bottom RH corner.

Seems obvious that they were flying the visual and got confused, but queried and all ended well.
Last flight of the day, how many legs had they done that day?
Phew !

Here is an edited clip of the tower audio:

https://soundcloud.com/user-66001055...esy-liveatcnet

They were cleared for the dreaded FMS Bridge Visual approach to 28R.

Wind was given as 270/8 when AC 759 was cleared to land, nothing unusual.

Ignoring the AC crew noticing their error, say the taxiway was empty ... at what stage would they have noticed their error?

I'm more interested in your statement i put in bold. In light of the articles posted above, where this has happened at night before, at a large international airport no different to SFO ... what evidence are you using to back your claim that this can never happen? I'm genuinely interested

A visual approach cleared to land on 28R and ahead are two lit strips the left one must be 28L so the right hand one is 28R

Just thinking about approach/landing briefs & TEM. Is 28C a landing runway? If that's the case, and it's included in the brief, then identifying 28L means the next one over is 28R doesn't fit the briefing. Surely 28C would have been mentioned.
Also, from a crew behaviour point of view I'm assuming they would not have briefed a visual approach from TOD and not set up an FMC approach. I'm assuming SFO has a preferential takeoff & landing combination; therefore if 28L is preferential takeoff rwy, it follows on that 28R is preferred landing rwy, and the FMC would have been programmed and briefed before TOD. Being short cut to a visual still leaves the MAP programmed for 28R.
I'm not familiar with SFO or A320, so please correct my assumptions as necessary. Someone also asked about the status of ILS on 28R. If it was on, and I assumed tuned, it would have showed some deflection. Were both heads outside all the time? PM duties etc?
Regarding the lights of a/c at night: in EU strobes are normally only used on the runway not holding. Thus perhaps only taxi lights & ACB's were showing. Amidst all the runway & taxiway lights a/c become hidden, especially from above. Remember the accident at LAX when a B737 landed on top of a Metroliner. When NTSB flew the approach they could not see the Metroliner hidden within the runway lighting. It's not that easy, seeing a/c, and you can believe your eyes are playing tricks. The brain has been programmed by ATC that is clear and you believe you're where you are supposed to be. I'm sure a GA would have been made by the crew eventually, but probably at a low height when it all suddenly became abundantly clear with a huge "WTF Go Around" call.

Apropos the above: from the report on the 1988 incident referred to in post #13, showing the final positions of the two aircraft involved:

http://www.avgen.com/aar%202-1989.jpg

The One-Eleven (on the left) having come to a stop at the end of its ~950m landing roll on the taxiway and the 737, which had been taxying for takeoff, with both mains bogged down in the grass in an attempt to get out of the path of the oncoming aircraft.

Somehow, I don't think they mentioned 28C in the brief. At least I hope they didn't...

Dumb SLF here so apologies. If my car can know if it is wandering out of the lane I'm driving on, there must be a technical solution to an aircraft being say, 50m left or right of the track it should be on for the runway rather than the taxiway? Why does aviation seem so slow in adopting better technical solutions to the problems in front of it? Coloured lights (imperfect, as pointed out further up this chain) do seem a bit yesterday?

Not true, nolimitholdem
 

A similar attitude led to 35 people dead, 29 seriously injured on February 1, 1991 at LAX. In this case, the "occupied strip of pavement" was runway 24L. The "occupant" was a Skywest Fairchild Metroliner, and the aircraft landing on the "occupied strip of pavement" was a US Air Boeing 737.

The key difference in that case was ATC cleared the 737 to land on 24L while the Fairchild was also cleared to position and hold on 24L. But yes, a professionally piloted aircraft landed on an "occupied strip of pavement" with tragic results. Thank goodness for the go around instruction by ATC this week at SFO.

There is no runway 28C at SFO, there isn't enough room for 28L and R on IFR approaches, which is why SFO gets to be really slow in the summer when the fog is in.

scary rather than a ho hum go around, i would say.

a story went around back in the day that a certain Captain in our regional airline landed on one end of a taxiway in Atl one night, picked up the mike and asked the tower why another aircraft was on the other end of the active runway. then the other pilot said: "I wasnt going to say anything Southern, but you just landed on the taxiway." the story sounded like something the Sou pilot would have said,(in character). if it happened. dont think it made hard copy. not sure.

otoh, I will never forget seeing two stretch 8's in exactly the same attitude on the same runway in the midst of a moderate rain shower, one landing, one taking off, both otg with the same pitch up.

it was eerie. no one said a word.


also I was on a runway holding for TO at a SFla. airport one VFR day and a C-150 flew fight over me and landed down the runway.

I might suggest inexperience with SFO is a major factor. There are many AC pilots who read a route Brief on the airport but it doesn't give you all the answers or information. Also the Airbus managed descent for the arrival usually leaves the aircraft high and many pilots are distracted in getting down and slowing down usually getting stabilized between 1000 and 500 ft.

After the Asiana incident I thought the ops in SFO would have the landing runway ILS operating. Hard to ignore the localizer, or may be not.

Whatever is written or said, what really matters is was whether this was a normal end to a flight. Yes...Fugetaboutit.
No...press and media are going to town.

I dunno what scares me more these days, the utter incompetence of pilots entrusted with the lives of their passengers, or their colleagues tendency to attempt to gloss over incidents like this as minor trivialities

They weren't merely not quite lined up, they'd been lined up for a while

And they didn't initiate a go around with an abundance of caution, they were instructed to go around to save from landing on several large fully fuelled aircraft

The Gatwick landing on the taxyway (actually the relief taxyway alongside the regular one) actually happened twice, the BIA One-Eleven described above, and then later I believe it was an Air Malta 737.

In all these cases there are multiple closely parallel runway/taxyway combinations (three at Gatwick, four it seems at SFO), it happens at night, and heading for the runway on one side, they head for a taxyway further over on that side. Had the SFO runway lights over on 28L been switched off for any reason ? This was what had happened at Gatwick. If you know there are two parallel runways, you are landing on the right hand one, you see from several miles out just two parallel lines of lights, actually the right hand runway and its associated taxiway, you can start to see it.

If I recall SFO correctly, the two closely parallel runways are commonly both used together for simultaneous landings, and also takeoffs, in the "use everything together" style common in the USA but not so much elsewhere. So you might expect two lit runways.

I'm sure that an airport like SFO has superb lighting. In fact the general taxyway illumination is quite likely more brilliant than the runway lighting at some other points the crew encounter on the network, approach lights apart, where the runway is just passable and the taxyway pretty thin viewed from the air.

Looking at the various videos, (both day and night), of the 28R approach available on line, if the full approach lighting system was serviceable and switched on, it's astonishing that this could have happened.
This incident will obviously act as a 'heads-up' for this potential confusion. If both of the 28 approach light systems are routinely displayed simultaneously, a line of eastward-facing HI reds, across the seawall, adjacent to where the building is, while not a high-tech solution, might prevent a similar occurrence.