In addition to Jim's questions,
Apart from getting a precisely accurate overlay of the 1913 topo, in what other ways is its usability impaired by the fact that the photo was taken at an angle?
I cannot attest to much about the usability of that 1913 topo. Can you? If you are the one who wants to use it, then shouldn't you be the one verifying it's usefulness?
And lining it up is a pretty big issue, isn't it? Another issue is readability. You did a great job on extracting the lines for the first third, but there are places where I cannot quite make out the lines elsewhere. There has been so much writing and erasing on the map it makes it difficult to tell what is what in places.
Take for example the area where the 7th greens are marked. I can make out the 170 contour line, and part of the 175 contour line, and even a bit of what seems to be the 180 contour line, but cannot make it all out, and cannot tell if there is any circular top of ridge contour in all the eraser and pencil marks. Can you?
As I recall your sample cross check gave a variety of differences. Are you positing that the entire map is off by a uniform 10 feet (or some other number around 10 feet) or is it randomly different?
Obviously those are not the only two options. It could be skewed in the same direction with the errors in an approximate range.
You seem to want to pin me down on the accuracy of this 1913 topo, and I wouldn't even know where to begin. If you want to use it, that should be your job to determine the reliability, shouldn't you?
Generally, the errors seem to all go the same direction and the magnitude seems to be
roughly around 10 ft, but it very difficult to check, mostly because of limitations on the 1913 map and lack of information about how it was created, but also because of limitations on the USGS maps. We are dealing with contours lines of five and 10 feet, so telling the exact magnitude of the difference at any one point isn't feasible at this point.
What does my exercise with the 1/9 NED data along the 1st fairway and green tell you about the precision (within 5 feet, say) of the 1/9 data on the PV site?
I addressed this above in an post to you and a post to Jim. Generally, it confirmed what I expected from looking at the rasterized image-- you chose to do your sampling where there appears to be gaps in the data, and so it isn't a big surprise that your result was not very precise for this particular sample area. Given that we could both see this from looking at the rasterized image, I am still left wondering why you would chose such a spot? My guess is that if you repeated your experiment in a higher resolution area, you'd have much different results (provided you used a smaller contour interval.)
Are you able to actually look at the raw data for individual collection points? If so it might give us some idea on what is happening.