Case Study: Reading the ISS's Orbit from Its Element Set

"Every tracked object in orbit is published as two lines of numbers. Learn to read them and the whole sky becomes legible."

Executive Summary

The United States tracks tens of thousands of objects in orbit and publishes each one's orbit as a two-line element set (TLE) — a terse block of digits that encodes exactly the six orbital elements of §8.4 plus a mean motion. In this case study we take a representative element set for the International Space Station and read it: we extract the six elements, recover the semi-major axis from the mean motion by inverting Kepler's third law, compute the period and the perigee/apogee altitudes, and finally locate the station on its orbit at the epoch by solving Kepler's equation. Every number we get out, we check against something we already know about the ISS — its $51.6^\circ$ inclination, its $\sim!420\ \text{km}$ altitude, its $\sim!93$-minute period, its sixteen-sunrises-a-day. This is the everyday work of flight dynamics: turning a published element set into a physical orbit and a position in the sky.

The element values below are representative of the ISS (Tier 3 — illustrative, rounded; a real TLE changes every few hours as the station is tracked and reboosted), but the method is exactly what a tracking app on your phone runs. We use $\mu_\oplus = 3.986\times10^{5}\ \text{km}^3/\text{s}^2$ and $R_\oplus = 6{,}371\ \text{km}$.

Skills applied

• Identifying the six orbital elements in a real data product (§8.4).
• Inverting Kepler's third law to get $a$ from the mean motion $n$ (§8.5).
• Computing period and apsidal altitudes from $a$ and $e$ (§8.3, §8.5).
• Locating a body via the anomalies and Kepler's equation — and seeing why near-circular orbits make the three anomalies nearly coincide (§8.6).

Background: what a TLE actually contains

Stripped of its formatting and checksums, an ISS element set carries these values (representative):

Notice what is not listed directly: the semi-major axis. TLEs quote the mean motion instead, because it is what tracking most directly measures (how often the object comes around). Our first job is to recover $a$ from it — a pure application of Kepler's third law.

Phase 1: From mean motion to semi-major axis

The mean motion in SI is $$n = 15.50\ \frac{\text{rev}}{\text{day}} \times \frac{2\pi\ \text{rad}}{1\ \text{rev}} \times \frac{1\ \text{day}}{86{,}400\ \text{s}} = 1.1272\times10^{-3}\ \text{rad/s}.$$ Kepler's third law in mean-motion form is $n = \sqrt{\mu/a^3}$, so $a^3 = \mu/n^2$ and $$a = \left(\frac{\mu}{n^{2}}\right)^{1/3} = \left(\frac{3.986\times10^{5}}{(1.1272\times10^{-3})^{2}}\right)^{1/3} = \left(\frac{3.986\times10^{5}}{1.2706\times10^{-6}}\right)^{1/3} = \left(3.137\times10^{11}\right)^{1/3} = 6{,}795\ \text{km}.$$ The altitude of a circular orbit at that radius would be $6{,}795 - 6{,}371 = 424\ \text{km}$ — squarely in the ISS's operating band. Sanity check: the ISS is famously at "about $400$–$420\ \text{km}$," and it is reboosted periodically as drag pulls it down, so $424\ \text{km}$ at this epoch is exactly right.

Phase 2: Period, and the sixteen sunrises

The period is the reciprocal of the mean motion (times $2\pi$): $$T = \frac{2\pi}{n} = \frac{2\pi}{1.1272\times10^{-3}} = 5{,}574\ \text{s} = 92.9\ \text{minutes}.$$ Equivalently, from the rev/day directly, $T = 86{,}400/15.50 = 5{,}574\ \text{s}$ — the same number, a nice internal consistency check. Over a 24-hour day the station completes $86{,}400/5{,}574 = 15.5$ orbits, so its crew sees roughly sixteen sunrises and sunsets every day — the often-quoted figure, now derived. Each orbit, the ISS travels once around a planet whose surface is turning beneath it, which is why its ground track marches westward and it never passes over quite the same place twice in a row.

Phase 3: Perigee, apogee, and how circular "circular" really is

With $a = 6{,}795\ \text{km}$ and $e = 0.0007$, the apsidal radii are $$r_p = a(1-e) = 6{,}795\,(1 - 0.0007) = 6{,}790\ \text{km}, \qquad r_a = a(1+e) = 6{,}795\,(1 + 0.0007) = 6{,}800\ \text{km}.$$ In altitude, the ISS dips to about $419\ \text{km}$ at perigee and rises to about $429\ \text{km}$ at apogee — a variation of only about $10\ \text{km}$ across an orbit spanning the whole planet. That is what $e = 0.0007$ means physically: the orbit is a circle to a fraction of a percent. The station is deliberately kept this round so that its altitude, drag, and lighting are as uniform as possible — a sharp contrast to the geostationary transfer orbit of §8.3, whose $e = 0.723$ swings it from $6{,}771$ to $42{,}164\ \text{km}$.

🔧 Engineering Reality: why not exactly $e = 0$? No real orbit has $e$ identically zero. Atmospheric drag, the slightly non-spherical Earth (the $J_2$ bulge — Chapter 12), and every reboost burn nudge the eccentricity to small nonzero values. Flight dynamics teams manage eccentricity toward zero rather than achieving it; a "circular" operational orbit is really a very low-eccentricity ellipse whose shape is actively maintained. The elements in a TLE are a snapshot of a living, drifting orbit, valid for days, not forever.

Phase 4: Where is the station right now? (Kepler's equation)

The mean anomaly at the epoch is $M = 325.0^\circ = 5.6723\ \text{rad}$. To find the physical position we solve Kepler's equation $M = E - e\sin E$ for the eccentric anomaly $E$, then convert to the true anomaly $\nu$. Start from $E_0 = M + e\sin M$: $$E_0 = 5.6723 + 0.0007\,\sin(325.0^\circ) = 5.6723 + 0.0007(-0.5736) = 5.6719\ \text{rad} = 324.98^\circ.$$ One Newton step barely moves it — because $e$ is so tiny, the correction term $e\sin E$ is under half a milliradian. To four figures, $E = 324.98^\circ$. Converting to the true anomaly with $\cos\nu = (\cos E - e)/(1 - e\cos E)$ gives $\nu = 324.95^\circ$. So at the epoch: $$M = 325.00^\circ, \qquad E = 324.98^\circ, \qquad \nu = 324.95^\circ.$$ All three anomalies agree to within $0.05^\circ$.

The lesson hiding in the near-equality. For a near-circular orbit the three anomalies almost coincide, because the differences between them are driven by eccentricity: $E - M = e\sin E$ and $\nu - E$ is also order $e$. When $e \to 0$, position advances almost uniformly with time, and you can get away with reading the true anomaly straight off the clock — the very shortcut §8.6's "Find the Error" warned against for eccentric orbits. The ISS forgives that sloppiness; the GTO satellite of §8.6 (where $M = 68^\circ$ but $\nu = 147^\circ$) does not. How much the anomalies diverge is a direct readout of how eccentric the orbit is.

To finish placing the station, add the argument of perigee: its angle from the ascending node, measured in the orbital plane, is $\omega + \nu = 130.5^\circ + 324.95^\circ = 455.45^\circ = 95.45^\circ$ (mod $360^\circ$) past the node. Combined with $\Omega$ and $i$, that fixes the ISS's direction in inertial space — the input a ground antenna needs to point at it.

Phase 5: The audit summary

Every number cross-checks against an independently known fact about the station. That is what a good audit looks like: not one heroic calculation, but a web of small ones that all agree.

Discussion Questions

1. The TLE quotes mean motion, not semi-major axis. Why is mean motion the more natural thing for a tracking network to measure, and how does Kepler's third law convert it to $a$?
2. The ISS's three anomalies at the epoch agreed to within $0.05^\circ$. For what kind of orbit would they differ by tens of degrees, and which element controls the size of the disagreement?
3. Inclination $51.64^\circ$ is not an accident. Knowing that a launch site can most cheaply reach orbits whose inclination is at least its own latitude, what does the ISS's inclination tell you about the latitude of the site it was assembled from?
4. The audit found the orbit is only "circular" to $e = 0.0007$. Name two physical effects that keep a real orbit from having exactly $e = 0$.

Your Turn: Extensions

• Option A (analysis). Find a current real ISS TLE online (CelesTrak or Space-Track publish them freely). Pull out $i$, $e$, $\omega$, $M$, and the mean motion, and repeat Phases 1–3 for today's orbit. How much has the altitude changed from our $424\ \text{km}$? (If it has dropped, the station is due for a reboost.)
• Option B (computation). Write tle_to_orbit(n_rev_per_day, e) that returns $(a, T, r_p, r_a)$ in kilometres, inverting Kepler's third law. Reproduce the Phase-1 to Phase-3 numbers. Do not run it — hand-trace and add # Expected output:.
• Option C (your mission). Write your mission's target orbit (Track A/B/C/D) as a six-element set in the same table format, and compute its mean motion in rev/day so it could be published as a TLE. For a GEO comsat, confirm the mean motion comes out to almost exactly $1.00\ \text{rev/day}$.

Key Takeaways

1. A TLE is the six orbital elements in disguise, with the mean motion standing in for the semi-major axis. Reading it is a direct exercise of §8.4–8.6.
2. Kepler's third law inverts to give $a$ from $n$: $a = (\mu/n^2)^{1/3}$. Everything else — period, apogee, perigee — follows from $a$ and $e$.
3. For near-circular orbits the three anomalies nearly coincide, and their divergence is a direct measure of eccentricity. Kepler's equation is "easy" for the ISS and "hard" for a GTO for exactly this reason.
4. A good audit is a web of cross-checks. Every number we computed matched an independently known ISS fact — inclination, altitude, period, sunrise count — which is how you gain confidence that the orbit you read is the orbit that is really up there.