NavList:

Don, you wrote:
"My estimated line of position was just a bit more then halfway from Frank's MA line to the NH line, putting it within 2 miles of my viewing site. Unfortunately, I screwed up on trying to record the time."

Glad you saw it and were able to get navigational information from it. I put the word out far and wide on this one. My best report back so far was from a friend who was bartending at a nice restaurant in Newport. She took the half-dozen people who were sitting at her bar plus the restaurant's piano player outside to watch. Naturally their interest was non-navigational, but it gets people thinking!

Regarding the timing of these satellite passes, for an object like the space station in low Earth orbit, motion along the ground track is very close to four nautical miles per second, so that's clearly a limiting factor for "visual satellite navigation" (the number makes sense at 16x faster than the Sun's motion along the equator since a LEO satellite completes sixteen orbits in one day). Also, the primary uncertainty in the orbital elements for satellites is the "mean longitude" which is equivalent to an error of some seconds in the passage time. Given the timing difficulty and this uncertainty in the satellite's position, I prefer to think of these observations in terms of short lines of position. As in your case, an observation gives us a line of position more or less parallel to the centerline ground track. For a satellite like the ISS whose orbit is tracked and updated frequently, the length of the LOP is could be as small as five or ten nautical miles. For other satellites, it might be better to assume fifty nautical miles.

We really need two satellite passes with crossing LOPs to get a proper fix from visual satellite observations. On some, relatively rare, nights, you could do this with two separate high passes of the ISS, but since weeks can go by with no ISS passes at all, it's clearly better to use multiple satellites. I actually spotted another satellite Saturday night at 18:46:24 just as it passed almost exactly over zeta Cygni (the end of the wingtip of Cygnus in the direction away from Vega). That was Cosmos 1626, a defunct former-Soviet satellite on a polar orbit. Given that convenient pass right on top of an identifiable star, and since this satellite was travelling north-south, this observation determined my longitude correctly within a mile. If I trust the timing, the latitude from that satellite along would have been about two miles north of my actual position. Or, crossing with the LOP from the ISS observation, I get my position accurate within one mile in both latitude and longitude. 

Given the recent iPhone photos from Norm Goldblatt and Peter Monta, it's not hard to imagine an app that could be set up to take photos once a minute for twenty minutes during late twilight that could get a position fix automatically, and its net accuracy would be potentially ten or even twenty times better than these simple "eyeball" observations. The app would require relatively current orbital elements for the satellites, but for many objects the elements could be as much as two weeks out of date with no significant additional inaccuracy.

The next step that I see in constructing this system of visual satellite navigation is to get people to try it out. Just how well can minimally trained observers track the motion of a satellite among the stars? How well do they need to know the constellations? Also I am trying to collect a list of "reliable" satellites. There are some satellites whose orbits are not checked frequently, and there are other satellites that maneuver fairly often. We need satellites that behave.  

-FER