I agree with the composite theory. The girder and white area above it
are too artificial to be original for a start. If you look at Paul
Terry's picture the first canopy supoort is just about level with the
front of the train, about two or three feet back from the end of the
pipe, where the red machine is.
I don't know if this is significant, but the righthand trainstop lights
don't have a shadow on them, whilst the other two sets do. I think the
wall is definately Wimbledon, the clincher is the number 47 just before
the end of the pipe in both pictures. If you examine that in both
pictures, the 47 in the Canadian one seems to be larger than in Paul's
picture. The viewpoint in the Canadian picture is further away however,
suggersting the foreground features could have been dropped in at a
smaller scale in front of the wall.

Neill