Case Study: Designing a Molniya Orbit for High-Latitude Coverage

"A geostationary satellite is useless above about sixty degrees of latitude — it sits too low on the horizon. The Soviets, with most of their territory in the far north, needed a different orbit. They let Kepler's second law do the work."

Executive Summary

Geostationary satellites hover over the equator, which is wonderful if you live in the tropics and useless if you live near the pole, where a satellite on the equator sits so low on the horizon that hills and buildings block it. The classic answer, flown by the Soviet Union from the 1960s and still in use, is the Molniya orbit: a highly eccentric, half-sidereal-day orbit whose apogee hangs over the northern hemisphere. Because a body near apogee crawls (Kepler's second law), a Molniya satellite spends the overwhelming majority of each orbit loitering high over the north, appearing nearly stationary to a ground station there — and a small constellation of them provides continuous coverage no geostationary satellite can match at high latitude.

In this case study we design one from scratch. We set the period from the repeat-ground-track requirement and recover the semi-major axis (Kepler's third law); set the perigee from a launch constraint and recover the eccentricity; choose the two orientation angles that make the design work; and then compute the payoff quantitatively — how many hours of each orbit the satellite spends loitering near apogee — using the eccentric-anomaly relations of §8.6. That last calculation is the whole reason the orbit exists, and it is Kepler's second law cashed out in hours.

Design targets are illustrative but realistic (Tier 2/3). We use $\mu_\oplus = 3.986\times10^{5}\ \text{km}^3/\text{s}^2$ and $R_\oplus = 6{,}371\ \text{km}$.

Skills applied

• Sizing an orbit from a period requirement via Kepler's third law (§8.5).
• Setting eccentricity from a perigee constraint via the apsidal relations (§8.3).
• Choosing inclination and argument of perigee for a mission purpose (§8.4).
• Turning Kepler's second law into a dwell time by integrating the anomalies over an arc (§8.6).

Background: the design brief

• Purpose: continuous communications coverage of a high-latitude service region (say, northern Russia, Canada, or the Arctic) that geostationary satellites serve poorly.
• Repeat requirement: the ground track must repeat daily, so the period should be a simple fraction of a sidereal day. We choose a 12-hour orbit — exactly half a sidereal day, $T = 86{,}164/2 = 43{,}082\ \text{s}$ — so the satellite retraces its track every two revolutions per day.
• Launch constraint: a perigee altitude of $500\ \text{km}$ (low enough for an efficient launch, high enough to avoid significant drag).
• Deliverable: a full six-element set plus a quantified apogee-dwell time.

Phase 1: Size the orbit from the period (Kepler III)

Invert Kepler's third law to get the semi-major axis from the required period: $$a = \left(\frac{\mu\,T^{2}}{4\pi^{2}}\right)^{1/3} = \left(\frac{(3.986\times10^{5})(43{,}082)^{2}}{39.478}\right)^{1/3} = \left(\frac{(3.986\times10^{5})(1.8561\times10^{9})}{39.478}\right)^{1/3}.$$ The numerator is $7.398\times10^{14}$, divided by $39.478$ gives $1.874\times10^{13}\ \text{km}^3$, and the cube root is $$a = (1.874\times10^{13})^{1/3} = 26{,}562\ \text{km}.$$ Sanity check: this is smaller than the geostationary $42{,}164\ \text{km}$, as it must be — a 12-hour orbit is lower than a 24-hour one — and $26{,}562\ \text{km}$ is indeed the textbook Molniya semi-major axis.

Phase 2: Set the shape from the perigee (the apsidal relations)

A perigee altitude of $500\ \text{km}$ means $r_p = 6{,}371 + 500 = 6{,}871\ \text{km}$. From $r_p = a(1-e)$, $$e = 1 - \frac{r_p}{a} = 1 - \frac{6{,}871}{26{,}562} = 1 - 0.2587 = 0.741.$$ This is a very eccentric orbit — nothing like the near-circular ISS. The apogee radius is $$r_a = a(1+e) = 26{,}562\,(1.741) = 46{,}244\ \text{km},$$ an apogee altitude of $46{,}244 - 6{,}371 = 39{,}873\ \text{km}$ — slightly higher than geostationary altitude. So the satellite plunges to $500\ \text{km}$ at perigee and climbs to nearly $40{,}000\ \text{km}$ at apogee twice a day. By vis-viva (Chapter 6) its speed ranges from $v_p = 10.05\ \text{km/s}$ down to $v_a = 1.49\ \text{km/s}$ — a factor of nearly seven, which is precisely the lever that makes it loiter at the top.

Phase 3: Orient the orbit — the two angles that matter

Two of the remaining elements are chosen deliberately; the third ($\Omega$) simply sets the local time of coverage and can be picked to suit.

• Argument of perigee $\omega = 270^\circ$. This places perigee at the southernmost point of the orbit, and therefore apogee at the northernmost — so the slow, high-loitering half of the orbit hangs over the northern hemisphere, exactly where we want coverage. Setting $\omega$ is the single most important design choice after the shape.

• Inclination $i = 63.4^\circ$. This specific value is the critical inclination. At $63.4^\circ$ (and its supplement $116.6^\circ$), the oblateness of the Earth does not rotate the orbit's apogee away from the north over time — the argument of perigee stays frozen at $270^\circ$ instead of drifting. Why $63.4^\circ$ does this is a perturbation result we derive in Chapter 12; for now, take it as the inclination that keeps a Molniya orbit pointing where you aimed it. It also happens to give good coverage of high latitudes.

That completes the six-element specification:

Phase 4: The payoff — how long does it loiter? (Kepler's second law, in hours)

Here is the number that justifies the whole design. The satellite is useful to a northern ground station only while it is high — say, above a radius of $r = 30{,}000\ \text{km}$ (an altitude of about $23{,}600\ \text{km}$), where it appears high in the sky and moves slowly. How much of each 12-hour orbit does it spend up there? Kepler's second law says "a lot," because it crawls near apogee; let us make "a lot" a number.

Step 1 — find the true anomaly where $r = 30{,}000\ \text{km}$. The semi-latus rectum is $p = a(1-e^2) = 26{,}562\,(1 - 0.549) = 11{,}976\ \text{km}$. From the orbit equation $r = p/(1 + e\cos\nu)$, $$\cos\nu = \frac{1}{e}\left(\frac{p}{r} - 1\right) = \frac{1}{0.741}\left(\frac{11{,}976}{30{,}000} - 1\right) = \frac{-0.6008}{0.741} = -0.8108,$$ so $\nu = 144.2^\circ$ (outbound) and, by symmetry about apogee, $\nu = 215.8^\circ$ (inbound). The satellite is above $30{,}000\ \text{km}$ for the arc $144.2^\circ \le \nu \le 215.8^\circ$.

Step 2 — convert that arc to eccentric anomaly. Using $\cos E = (\cos\nu + e)/(1 + e\cos\nu)$ at $\nu = 144.2^\circ$: $$\cos E = \frac{-0.8108 + 0.741}{1 + 0.741(-0.8108)} = \frac{-0.0698}{0.3992} = -0.1749 \ \Rightarrow\ E = 100.1^\circ.$$

Step 3 — convert to mean anomaly (time) with Kepler's equation. $M = E - e\sin E$, with $E = 100.1^\circ = 1.7466\ \text{rad}$ and $\sin E = 0.9848$: $$M = 1.7466 - 0.741(0.9848) = 1.7466 - 0.7297 = 1.0169\ \text{rad} = 58.3^\circ.$$ By symmetry, the inbound crossing is at $M = 360^\circ - 58.3^\circ = 301.7^\circ$.

Step 4 — the dwell time. Mean anomaly advances uniformly with time, so the fraction of the period spent above $30{,}000\ \text{km}$ is the fraction of $360^\circ$ between the two crossings: $$\frac{\Delta M}{360^\circ} = \frac{301.7^\circ - 58.3^\circ}{360^\circ} = \frac{243.4^\circ}{360^\circ} = 0.676.$$ $$t_{\text{loiter}} = 0.676 \times T = 0.676 \times 43{,}082\ \text{s} = 29{,}100\ \text{s} = 8.1\ \text{hours}.$$

The satellite spends about 8.1 of its 12 hours above $30{,}000\ \text{km}$ — two-thirds of every orbit loitering high over the north — even though that high region is less than a third of the orbit by distance. That gap between "two-thirds of the time" and "a third of the space" is Kepler's second law made economic: near apogee the satellite is barely moving, so it dwells there. It races through the low southern perigee in the remaining $3.9$ hours, unseen and unneeded.

🔗 Connection: three satellites, continuous coverage. Because one satellite is useful for about $8\ \text{hours}$ per $12$-hour orbit, three Molniya satellites, spaced evenly in time, keep at least one loitering over the north at all times — a continuous high-latitude service from just three spacecraft. This is the constellation logic we formalize in Chapter 9; the dwell time we just computed is the number that sets how many satellites the constellation needs.

Phase 5: Reality check and the constant that made it work

Every step leaned on one of the chapter's results: Kepler's third law sized the orbit, the apsidal relations shaped it, and Kepler's equation turned an angular arc into a dwell time. The design closes cleanly — a $26{,}562\ \text{km}$, $e = 0.741$, $63.4^\circ$ orbit with apogee pinned over the north, loitering $8\ \text{hours}$ per pass. Two caveats a real program would add: the $63.4^\circ$ critical inclination only approximately freezes the apogee (higher-order perturbations still require occasional station-keeping, Chapter 12), and the deep perigee passes drive the satellite repeatedly through the inner Van Allen radiation belt, which stresses its electronics — a real cost the Soviets accepted for the coverage. But the orbital design itself is nothing but Kepler, applied with intent.

Discussion Questions

1. The satellite spends two-thirds of its time in the high third of its orbit by distance. State the single law responsible, and explain the mechanism in one sentence.
2. Why is the period chosen as half a sidereal day rather than a full one? What would a full-sidereal-day eccentric orbit (a "Tundra" orbit) buy or cost compared with the 12-hour Molniya?
3. If you set $\omega = 90^\circ$ instead of $270^\circ$, where would apogee hang, and for whom would the satellite then be useful?
4. The design fixed $a$, $e$, $i$, and $\omega$ from mission requirements but left $\Omega$ free. What does choosing $\Omega$ actually control, and why is it the natural "free" element in a constellation?

Your Turn: Extensions

• Option A (design). Redesign for a perigee altitude of $1{,}000\ \text{km}$ instead of $500\ \text{km}$, keeping the 12-hour period. Recompute $a$ (unchanged — why?), $e$, the apogee altitude, and the new $8$-hour-ish dwell above $30{,}000\ \text{km}$. Does raising perigee help or hurt the loiter time?
• Option B (computation). Write dwell_fraction(a, e, r_threshold, mu) that returns the fraction of the period spent above a given radius, following Phase 4 (find $\nu$, then $E$, then $M$, then $\Delta M/2\pi$). Reproduce the $0.676$ figure. Do not run it — hand-trace and add # Expected output:.
• Option C (your mission). If your mission (Track A/B/C/D) has any coverage or dwell requirement — time over a ground site, time in sunlight, time above a minimum altitude — set a radius or true-anomaly threshold and compute the dwell fraction with the Phase-4 method. Record it in your MDR.

Key Takeaways

1. A Molniya orbit is Kepler's second law turned into a service. High eccentricity plus apogee over the north makes the satellite loiter where it is needed, because it moves slowest when highest.
2. The design is a chain of chapter results: Kepler III sizes $a$ from the period; the apsidal relations set $e$ from the perigee; $\omega = 270^\circ$ and $i = 63.4^\circ$ orient and freeze it.
3. Dwell time comes from the anomalies. Converting a radius threshold to $\nu$, then $E$, then $M$ turns "it loiters near apogee" into "$8.1$ hours of every $12$" — the number that sizes the constellation.
4. Every orbit is a design choice. The same two-body physics that gives a boring circle gives this deliberately lopsided ellipse; which one you fly is dictated by the mission, which is the theme of Chapter 9.