- This terminology needs clarification, otherwise there will be many ways to compute it.
Loosely speaking, it is an indicator of how often a satellite revisits a given site on Earth. While it is self-evident that it depends on the orbit of the satellite, less obvious is that it depends on the latitude of the observing site.
Furthermore, it is non-ambiguous only when you pin down what “often” and “visit” mean. For example, you may wish to clarify that “how often” means “how many times in a day” (note that "in 48h", "in a week", are perfectly plausible options too). Likewise a visit means that the satellite appears above the horizon, at least above 10° elevation (or 50° above).
Still, another less obvious aspect is the implicit assumption that the events are more or less periodic. In other words, there would be a quasi-constant “period” between visits.
- Let’s examine an example. Say we want to know how often, PER DAY, we have the ISS (ZARYA) appearing above New-York (~ 40.74°N,~ 73.95°W) horizon, at least above 10° elevation.
Referring to heavens-above:
17Sep= 7 visits; 18Sep= 5; 19Sep= 6; 20Sep= 6; 21Sep= 7; 22Sep= 6; 23Sep= 6; 24Sep= 6;
You may hastily make the “short-cut” by generalizing that “revisit rate” is 6 per day, on average, equivalently, one visit every 4 hours (about). But this is not correct, the different visits are irregularly spaced (look at the rise times data in details), although the number of visits per day is fairly constant.
Let’s change our definition of “visit” by requiring that the satellite is at least 50° above the horizon. The number of visits per day then becomes (look at the “Highest points” data):
17Sep= 1; 18Sep=2; 19Sep=1; 20Sep=1; 21Sep=1; 22Sep=2; 23Sep=2; 24 Sep=0.
Now, change the observation site’s latitude to 51°N. The result for 50° above horizon:
17Sep= 3; 18Sep=3; 19Sep=3; 20Sep=2; 21Sep=3; 22Sep=3; 23Sep=2; 24 Sep=2.
This shows the strong dependency to the latitude of the observing site.
From these examples, for a single satellite, there is in general insufficient “periodicity” to warrant the use of “revisit rate” and/or “revisit period”.
A particular case for a regular visit pattern is when the orbit is equatorial (or near-equatorial) AND the observing site is near the equator. Another particular case leading to regular visits is when the orbit is polar (90° inclination), or near-polar, and the observing point is near-polar.
Nevertheless, with a constellation of many satellites, it is possible to make the visiting pattern (visits from several satellites) almost regular for almost any points on Earth, while minimizing the number of satellites. This is however a complex optimization problem, requiring sophisticated genetic algorithms.