Chapter 9: Orbit Types and Their Uses

"An 'artificial satellite' at the correct distance from the earth would make one revolution every 24 hours; i.e., it would remain stationary above the same spot and would be within optical range of nearly half the earth's surface." — Arthur C. Clarke, "Extra-Terrestrial Relays," Wireless World (1945)

Overview

There is no such thing as the orbit. When a beginner pictures "a satellite going around the Earth," they usually imagine one generic circle. But the space around our planet is not a single lane; it is a vast, layered filing cabinet, and mission designers have learned to put each kind of spacecraft in its own precise drawer. A weather satellite that must stare at one hemisphere without blinking lives in a completely different orbit from a GPS satellite that must be seen from everywhere, which lives in a different orbit again from an imaging satellite that must photograph the whole Earth under identical morning light. Each of these is a different altitude, a different shape, a different tilt — and every one of those choices is an answer to a question the mission asked first.

That is the real subject of this chapter. In Chapter 8 you learned that any orbit is pinned down by six numbers — its orbital elements — and that its size $a$ fixes its period through Kepler's third law. Here we put those numbers to work. We will tour the orbits that missions actually use — low Earth orbit, medium Earth orbit, geostationary orbit, highly elliptical and Molniya orbits, and the sun-synchronous and polar orbits — and for each one we will ask the only question that matters in engineering: why does this orbit exist? What does its particular altitude and inclination buy, and what does it cost? By the end you will not memorize a list of orbits; you will be able to derive which orbit a mission needs from what the mission is trying to do. That skill — turning a requirement into an orbit — is the first genuine act of mission design, and it is where the beauty of orbital mechanics stops being abstract and starts steering real hardware.

This chapter leans harder than any so far on the theme that orbital mechanics is beautiful: the same handful of relations from Chapter 6 and Chapter 8 — circular velocity, vis-viva, Kepler's third law — explain the entire catalog. One set of equations, applied with a designer's judgment, lays out every drawer in the filing cabinet.

In this chapter, you will learn to:

• Place any Earth satellite in the right regime — LEO, MEO, GEO, HEO, or SSO — from its altitude and job.
• Compute a circular orbit's period and speed, and show why "geostationary" forces one exact altitude.
• Explain the clever orbits — Molniya (loiter high over the north) and sun-synchronous (constant lighting).
• Read the trade behind every orbit: coverage versus resolution, latency versus reach, cheap versus radiation-safe.
• Select and defend the orbit for your own mission, recording its altitude, inclination, and rationale.

Learning Paths

🚀 Space Enthusiast: Read 9.1, 9.3, and 9.4 for the three most iconic regimes (LEO, geostationary, and the beautiful Molniya loiter). Skim the arithmetic; keep the pictures. The Clarke history in 9.3 and the "why the poles need a different orbit" story in 9.4 are the ideas that will change how you read a satellite headline.

📐 Engineering Student: Read everything and do every worked example yourself before reading ours — they are nothing but Chapter 6 vis-viva and Chapter 8 period, reused. The ⭐⭐/⭐⭐⭐ exercises on coverage, Molniya geometry, and sun-synchronous inclination are the foundation for the maneuvers of Chapter 10.

🎮 KSP Player: You have flown most of these orbits. Focus on 9.3 (getting the geostationary altitude exactly right, the thing the game punishes you for missing) and 9.5 (why real Earth-observation sats are tilted just past 90°). The orbit catalog in the Checkpoint is the cheat-sheet you wish the game gave you.

🛰️ Industry Prep: This whole chapter is the vocabulary of a mission concept review. Section 9.6 and the Mission Design Checkpoint are the deliverable — by the end you will have chosen your mission's orbit and be able to justify altitude and inclination the way a systems engineer must in the first design meeting.

9.1 Low Earth Orbit

Start at the bottom, closest to home, where almost everything humans have ever launched actually lives.

Definition (low Earth orbit, LEO). Low Earth orbit is the regime of orbits with altitudes roughly from $160\ \text{km}$ up to about $2{,}000\ \text{km}$. Below about $160\ \text{km}$ the residual atmosphere drags a satellite down within days; above about $2{,}000\ \text{km}$ you enter the inner radiation belt and the character of the orbit changes. A typical LEO — the International Space Station, most imaging satellites, the Starlink constellation — sits near $400$–$800\ \text{km}$, circles the Earth in about $90$–$100$ minutes, and moves at roughly $7.5$–$7.8\ \text{km/s}$.

We already met the physics in Chapter 6: the circular speed at radius $r$ is $v_{\text{circ}} = \sqrt{\mu/r}$, and for a $400\ \text{km}$ orbit ($r = R_\oplus + h = 6{,}371 + 400 = 6{,}771\ \text{km}$) that came out to $7.67\ \text{km/s}$. Kepler's third law from Chapter 8, $T = 2\pi\sqrt{a^3/\mu}$, gives its period. Let us compute that period explicitly, because the number is one every space engineer carries in their head.

Worked Example: how long is a lap in LEO? For the $400\ \text{km}$ circular orbit, $a = 6{,}771\ \text{km}$ and Earth's $\mu_\oplus = 3.986\times10^{5}\ \text{km}^3/\text{s}^2$: $$T = 2\pi\sqrt{\frac{a^3}{\mu}} = 2\pi\sqrt{\frac{(6{,}771)^3}{3.986\times10^{5}}} > = 2\pi\sqrt{7.79\times10^{5}\ \text{s}^2} = 2\pi\,(882.5\ \text{s}) = 5{,}544\ \text{s}.$$ That is $92.4\ \text{minutes}$ — call it an hour and a half. The units check: $\text{km}^3 / (\text{km}^3/ \text{s}^2) = \text{s}^2$, whose square root is seconds. A satellite in LEO therefore completes about $1{,}440 / 92.4 \approx 15.6$ orbits every day. Sanity check: the ISS is famously said to give its crew "sixteen sunrises a day," and $15.6$ orbits is exactly that. The physics and the folklore agree.

Why would you want to be this low? Because low is close, and closeness is worth a great deal. Three advantages fall straight out of the small radius:

• It is the cheapest place to reach. From the delta-v map of Chapter 3, simply getting to LEO is the single most expensive line item of any mission (~$9.4\ \text{km/s}$). Once you are there you have paid the hardest toll in spaceflight; going no higher means spending nothing more. Every kilogram is cheapest to put in LEO, which is why the overwhelming majority of all satellites are here.
• You are close to the ground. For a camera, resolution scales with proximity: a lens $400\ \text{km}$ away resolves detail a hundred times finer than the same lens at GEO, $35{,}786\ \text{km}$ away. For a radio link, the round-trip time is milliseconds, not the quarter-second lag of a geostationary hop — which is precisely why the new broadband constellations chose LEO.
• You are (mostly) below the radiation belts. A low orbit spends its life inside the protective cocoon of Earth's magnetic field, beneath the Van Allen belts previewed in Chapter 1. That means less shielding mass and longer-lived electronics than an orbit that must plow through the belts.

The price of being close is that you cannot see much at once. From $400\ \text{km}$ the geometric horizon is only about $2{,}200\ \text{km}$ away (a satellite sees a circle roughly the size of the continental United States), and because it is racing along at $7.7\ \text{km/s}$ that footprint sweeps past any point on the ground in a few minutes and does not return for hours. A single LEO satellite is a searchlight, not a floodlight. To get continuous coverage of the whole planet from LEO, you need not one satellite but a whole constellation of them — dozens, hundreds, or in Starlink's case thousands — an idea we develop in Chapter 33.

💡 Intuition: searchlight versus floodlight. Altitude is, more than anything, a knob that trades resolution against coverage. Fly low and you see a small patch in exquisite detail; fly high and you see a whole hemisphere but blur the fine print. There is no free lunch: the same geometry that makes a low satellite a superb camera makes it a poor broadcaster, and vice versa. Almost every orbit choice in this chapter is a position taken on that one trade.

🔧 Engineering Reality: LEO is not empty, and it does not last forever. Even at $400\ \text{km}$ there is enough wisp of atmosphere to slowly drain an orbit's energy. Left alone, the ISS would lose altitude and re-enter within a year or two; it must be reboosted periodically to stay up. This is not a nuisance to wish away — it is a feature we will lean on in Chapter 35, because it means low satellites eventually clean themselves out of orbit rather than becoming permanent debris. The same drag that we will fight in Chapter 12 is what keeps LEO from filling up forever. Low orbits are self-limiting; high ones are not.

🔄 Check Your Understanding 1. Two satellites carry identical cameras. One is at $400\ \text{km}$, the other at $800\ \text{km}$. Which resolves finer detail on the ground, and which sees a larger area at once? 2. Why does a single LEO satellite fail to provide continuous communications to a fixed city, when a single geostationary satellite can?

Answers

1. The $400\ \text{km}$ satellite resolves finer detail (it is closer, so a given angular resolution maps to a smaller ground distance); the $800\ \text{km}$ satellite sees a larger area at once (its horizon is farther away). This is the resolution-versus-coverage trade in miniature. 2. A LEO satellite moves at $7.7\ \text{km/s}$ relative to the ground and is over any given city for only a few minutes per pass, with gaps of hours between passes; to a fixed antenna it rises and sets like a fast-moving star. A geostationary satellite (Section 9.3) hangs motionless over one longitude, so a dish can point at it once and never move again.

9.2 Medium Earth Orbit and navigation

Climb higher — past a couple of thousand kilometers, but not yet to the great heights of the geostationary belt — and you enter the middle floor.

Definition (medium Earth orbit, MEO). Medium Earth orbit is the broad band of orbits between LEO and GEO, from roughly $2{,}000\ \text{km}$ altitude up to just below $35{,}786\ \text{km}$. It is less crowded than either neighbor, and it is dominated by one killer application: satellite navigation. The Global Positioning System (GPS), Europe's Galileo, Russia's GLONASS, and China's BeiDou all fly in MEO.

The signature MEO is the GPS orbit, at an altitude of about $20{,}200\ \text{km}$ ($r \approx 26{,}560\ \text{km}$). What makes this altitude special is its period. Run the numbers.

Worked Example: the twelve-hour orbit. For a GPS satellite at $r = 26{,}560\ \text{km}$: $$v_{\text{circ}} = \sqrt{\frac{\mu}{r}} = \sqrt{\frac{3.986\times10^{5}}{26{,}560}} = \sqrt{15.0\ \text{km}^2/\text{s}^2} > = 3.87\ \text{km/s},$$ $$T = 2\pi\sqrt{\frac{(26{,}560)^3}{3.986\times10^{5}}} = 2\pi\,(6{,}856\ \text{s}) = 43{,}077\ \text{s} > = 11\ \text{h}\ 58\ \text{min}.$$ That period is not an accident. A day measured against the stars — a sidereal day — is $23\ \text{h}\ 56\ \text{min}$, and $11\ \text{h}\ 58\ \text{min}$ is exactly half of it. A GPS satellite therefore circles the Earth precisely twice per day and retraces the same ground track every day. Engineers call this a semi-synchronous orbit, and the repeating ground track makes the constellation's geometry predictable and easy to manage.

Why is MEO the right home for navigation? Because GPS works by trilateration: your receiver measures its distance to several satellites at once (from the travel time of their signals) and solves for the one point in space consistent with all those distances. That job has two requirements that pull in opposite directions, and MEO is the altitude where they balance.

• You need many satellites visible at once — at least four, so the receiver can solve for its three position coordinates and its own clock error (the fourth unknown). More satellites, spread across the sky, give a sharper fix.
• You do not want a swarm of thousands of satellites, because each one is an expensive, precisely clocked, radiation-hardened machine.

From LEO, each satellite sees only a small patch of Earth, so you would need hundreds to guarantee four in view everywhere. From GEO, satellites hang over the equator and give poor geometry near the poles. MEO is the sweet spot: from $20{,}200\ \text{km}$, each satellite is visible across nearly half the planet, so a constellation of just 24 satellites (in six orbital planes, tilted at $55^\circ$ to reach high latitudes) guarantees at least four in view from anywhere on Earth, at almost any time. The altitude is chosen to make the coverage arithmetic close cleanly with the fewest expensive satellites.

🔗 Connection: why four satellites, not three? Three distances would seem enough to fix a point in 3-D space — and they would be, if your receiver had a perfect atomic clock. It does not. The receiver's cheap quartz clock is off by some unknown amount, and since distance is measured as (speed of light) × (signal travel time), a clock error of even a microsecond corrupts every distance by $300\ \text{m}$. The fix is elegant: treat the clock error as a fourth unknown and add a fourth satellite to solve for it. This is why your phone needs to "see" four GPS satellites to get a position — and why the navigation payload's onboard atomic clocks, stable to nanoseconds, are the real heart of the system. We return to precise ranging and timing in Chapter 26.

⚠️ Common Misconception: "Higher orbits must be safer from radiation." The opposite can be true. A MEO satellite spends its entire life among the Van Allen radiation belts, the zones of trapped high-energy particles that a LEO satellite sits safely beneath. That is the hidden tax of the middle altitudes: GPS and Galileo satellites must be built with heavily radiation-hardened electronics and extra shielding, mass they would not need down low. Altitude does not simply mean "less atmosphere and therefore easier" — it means moving your machine into a harsher particle environment. Every orbit is a bargain with the space environment (theme two), and MEO pays for its coverage in radiation.

9.3 Geostationary orbit and GTO

Keep climbing. Somewhere far above the navigation satellites there is one magic altitude — a single, razor-thin ring — where an orbit's period stretches to exactly one day. A satellite placed there turns around the Earth at precisely the rate the Earth turns beneath it, so from the ground it appears to stop: it hangs motionless in the sky, day and night, forever over the same spot on the equator. This is the most commercially valuable orbit ever discovered.

Definition (geostationary orbit, GEO). A geostationary orbit is a circular orbit directly over the equator (inclination $i = 0^\circ$) whose period equals one sidereal day, so the satellite appears fixed in the sky above one longitude. Its altitude is $35{,}786\ \text{km}$ (radius $42{,}164\ \text{km}$) and its speed is $3.07\ \text{km/s}$. The broader family of orbits with a one-day period but any inclination is called geosynchronous; a geostationary orbit is the special geosynchronous orbit that is also equatorial and circular, and therefore truly stationary.

There is nothing arbitrary about the altitude $35{,}786\ \text{km}$. It is forced on us by the demand that the period equal one day. We do not guess it — we solve for it. Invert Kepler's third law: given the period $T$, the semi-major axis must be $a = \left(\mu T^2 / 4\pi^2\right)^{1/3}$.

Worked Example: deriving the geostationary altitude. We want the orbit whose period is one sidereal day, $T = 86{,}164\ \text{s}$ (why sidereal and not $86{,}400\ \text{s}$, see below). Then $$a = \left(\frac{\mu T^2}{4\pi^2}\right)^{1/3} > = \left(\frac{(3.986\times10^{5})(86{,}164)^2}{4\pi^2}\right)^{1/3} > = \left(7.496\times10^{13}\ \text{km}^3\right)^{1/3} = 42{,}164\ \text{km}.$$ Because a geostationary satellite sits over the equator, its altitude is that radius minus Earth's equatorial radius, $6{,}378\ \text{km}$: $$h = 42{,}164 - 6{,}378 = 35{,}786\ \text{km}.$$ And its speed, from circular velocity, is $v = \sqrt{\mu/a} = \sqrt{3.986\times10^{5}/42{,}164} = 3.07\ \text{km/s}$ — less than half the LEO speed, exactly as the "higher is slower" rule of Chapter 6 demands. Sanity check: this is the same $42{,}164\ \text{km}$ and $3.07\ \text{km/s}$ we quoted in Chapter 6, now derived rather than asserted. The magic altitude is nothing but the radius at which a lap takes a day.

🚪 Threshold Concept: one altitude buys you a satellite that never moves. This is the idea that turned orbital mechanics into a global industry. At every other altitude, a satellite streaks across the sky and a ground antenna must chase it. At this one altitude — and only this one — the satellite's orbital period matches Earth's spin, so it appears nailed to a point in the sky. A television dish can be bolted to a wall, aimed once, and never touched again. Three such satellites, spaced $120^\circ$ apart around the equator, can blanket nearly the entire populated world with a signal. Arthur C. Clarke worked this out in a 1945 magazine article, decades before any satellite existed; the geostationary belt is still sometimes called the Clarke orbit in his honor. The lesson is deeper than television: a specific number chosen from physics — here, one particular radius — can unlock an entire capability that no amount of engineering at the wrong altitude could ever provide. Choosing the orbit is the invention.

📜 From History: the orbit that was published before it existed. In October 1945 a Royal Air Force radar officer named Arthur C. Clarke — later far more famous as a science-fiction writer — published a paper in Wireless World titled "Extra-Terrestrial Relays." In it he pointed out that a satellite at $42{,}000\ \text{km}$ would orbit once every 24 hours and thus appear stationary, and that three of them could provide worldwide radio coverage. He had no rocket, no satellite, and no way to build one; Sputnik was twelve years away. Yet the physics was exactly right, and the first geostationary communications satellite, Syncom 3, flew in 1964 and broadcast the Tokyo Olympics across the Pacific. It is one of the purest examples in engineering of theory arriving decades before hardware — the same pattern we saw with Tsiolkovsky and the rocket equation. The equation knows the answer before the machine can be built.

Why a sidereal day and not the ordinary $24$-hour solar day? Because the satellite must keep pace with Earth's rotation relative to the stars, not relative to the Sun. In one orbit the Earth also moves a little along its path around the Sun, so it must turn a bit more than $360^\circ$ for the Sun to return overhead — which is why the solar day ($86{,}400\ \text{s}$) is about four minutes longer than the true rotation period, the sidereal day ($86{,}164\ \text{s}$). A satellite synchronized to $86{,}400\ \text{s}$ would drift about one degree east per day and slowly wander off station. The four-minute distinction is the difference between a satellite that stays put and one that slowly escapes its slot.

Getting there: the geostationary transfer orbit

A launch vehicle almost never delivers a satellite straight into its circular geostationary home. Instead it drops it onto a stretched-out ellipse — low at one end, GEO-high at the other — and lets the satellite finish the climb with its own engine.

Definition (geostationary transfer orbit, GTO). A geostationary transfer orbit is an elliptical orbit with its perigee (low point) in low Earth orbit and its apogee (high point) at geostationary altitude, $35{,}786\ \text{km}$. A satellite is released into GTO by its launcher, coasts up to apogee for free, and there fires its own engine to circularize into GEO. It is the standard on-ramp to the geostationary belt.

This is precisely the transfer ellipse we computed in Chapter 6. Take a GTO with perigee at $250\ \text{km}$ altitude ($r_p = 6{,}628\ \text{km}$) and apogee at GEO ($r_a = 42{,}164\ \text{km}$). Its semi-major axis is $a = (r_p + r_a)/2 = 24{,}396\ \text{km}$, and vis-viva gives the speed at apogee: $$v_a = \sqrt{\mu\left(\frac{2}{r_a} - \frac{1}{a}\right)} = \sqrt{3.986\times10^{5}\left(\frac{2}{42{,}164} - \frac{1}{24{,}396}\right)} = 1.60\ \text{km/s}.$$ To circularize, the satellite must speed up from that $1.60\ \text{km/s}$ to the $3.07\ \text{km/s}$ of a circular GEO — a burn of $\Delta v = 3.07 - 1.60 = 1.47\ \text{km/s}$. That matches the "GTO → GEO $\approx 1.5\ \text{km/s}$" figure from the delta-v map of Chapter 3 almost exactly. The satellite's own fuel load is sized against this number; how to compute such transfer burns in general is the business of Chapter 10.

⚠️ Common Misconception: "Just put everything in GEO — it can see half the Earth." GEO is wonderful and also expensive in three different currencies. First, delta-v: reaching GEO costs several km/s more than LEO, so you launch far less mass for the same rocket. Second, distance: at $35{,}786\ \text{km}$ a camera's resolution is dreadful and a radio signal takes about a quarter-second to make the round trip — a delay you notice on a phone call and cannot tolerate for fast data. Third, real estate: because every geostationary satellite must share one thin equatorial ring and not interfere with its neighbors, the usable slots are a genuinely limited, internationally regulated resource. GEO is superb for broadcasting to a hemisphere and useless for photographing a license plate. There is no universal best orbit — only the best orbit for a job.

Leaving cleanly: the graveyard orbit

A geostationary satellite cannot simply be left dead in its slot; the slot is too valuable and the belt too crowded. Nor can it be de-orbited into the atmosphere — that would cost more delta-v than the satellite has left, because GEO is so high. So at end of life it does the cheap thing: it climbs a few hundred kilometers higher, out of the way.

Definition (graveyard orbit). A graveyard orbit (or disposal orbit) is an orbit a few hundred kilometers above the geostationary belt, into which a retiring GEO satellite is boosted so that it vacates its operational slot permanently. Raising the orbit by about $300\ \text{km}$ requires only a small delta-v — on the order of $11\ \text{m/s}$ — which operators reserve as the satellite's final maneuver.

That $11\ \text{m/s}$ is a tiny fee for keeping the belt usable, and it is now an expected part of responsible operations. We take up the broader problem of not littering space — and the international rules that now require disposal — in Chapter 35.

9.4 Highly elliptical and Molniya orbits

Every orbit so far has been essentially circular. But the most elegant orbits in this chapter are the lopsided ones, and they solve a problem the geostationary belt cannot: how do you serve the far north?

The trouble with GEO is that it sits over the equator. From a city at high latitude — Moscow, Anchorage, northern Canada — a geostationary satellite hangs low on the southern horizon, easily blocked by mountains, buildings, and trees, and at the very poles it disappears below the horizon entirely. A country that lives in the high north needs a satellite that spends its time up high over the north, not parked over the equator. The solution is to abandon the circle.

Definition (highly elliptical orbit, HEO). A highly elliptical orbit is an orbit with large eccentricity — a low perigee and a very high apogee. By Kepler's second law, a spacecraft moves slowly near apogee and quickly near perigee, so an HEO satellite spends the great majority of each orbit loitering near its distant apogee and only a brief interval whipping through perigee. Point that apogee over a region and the satellite hovers there, in effect, for hours at a time.

The masterpiece of this idea is the Molniya orbit, engineered by the Soviet Union in the 1960s for exactly their high-latitude problem.

Definition (Molniya orbit). A Molniya orbit is a specific highly elliptical orbit with a period of half a sidereal day (about $11\ \text{h}\ 58\ \text{min}$), an inclination of $63.4^\circ$, and its apogee placed high over the northern hemisphere (apogee altitude near $40{,}000\ \text{km}$, perigee only a few hundred kilometers up). The satellite dwells over the north for roughly eight hours of each orbit; a constellation of three, spaced eight hours apart, gives a high-latitude region continuous coverage. It is named for the Russian word for "lightning."

Let us build the orbit from its defining requirement, a half-sidereal-day period, and see the loiter emerge.

Worked Example: the anatomy of a Molniya orbit. We want a period of half a sidereal day, $T = 43{,}082\ \text{s}$. Inverting Kepler's third law as we did for GEO, $$a = \left(\frac{\mu T^2}{4\pi^2}\right)^{1/3} = \left(\frac{(3.986\times10^{5})(43{,}082)^2}{4\pi^2}\right)^{1/3} > = 26{,}560\ \text{km}.$$ (This is the same semi-major axis as GPS — of course it is; they share a period, and period depends only on $a$.) Now make it lopsided. Choose a perigee $600\ \text{km}$ up, $r_p = 6{,}971\ \text{km}$. Since $a$ is the average of perigee and apogee radii, the apogee radius is $r_a = 2a - r_p = 2(26{,}560) - 6{,}971 = 46{,}149\ \text{km}$, an apogee altitude of about $39{,}800\ \text{km}$ — higher even than GEO. The eccentricity is $$e = \frac{r_a - r_p}{r_a + r_p} = \frac{46{,}149 - 6{,}971}{46{,}149 + 6{,}971} = 0.74.$$ Now the payoff, from vis-viva. At perigee the satellite screams along at $v_p = \sqrt{\mu(2/r_p - 1/a)} = 9.97\ \text{km/s}$; at apogee it crawls at $v_a = \sqrt{\mu(2/r_a - 1/a)} = 1.51\ \text{km/s}$ — nearly seven times slower. Because it moves so slowly up high, it spends most of the twelve-hour orbit near apogee, hanging over the northern hemisphere, and races through the southern perigee in minutes. That slow, high, northern dwell is the whole point.

💡 Intuition: Kepler's second law is the trick. Recall from Chapter 8 that an orbit sweeps out equal areas in equal times. Near the distant apogee, the thin wedge of area is long and skinny, so covering it takes a long time — the satellite dawdles. Near perigee the wedge is short and fat, so the satellite must hurry. A Molniya orbit weaponizes this: by placing the sluggish apogee over the north, it converts Kepler's centuries-old geometric law into hours of hovering coverage over Russia or Canada. The same law Kepler found in Mars's motion tells a modern engineer where to point an ellipse. That is the beauty of orbital mechanics doing a practical job.

Two subtleties make the Molniya orbit work in practice, and both are worth flagging now even though their full explanation comes later. First, why the oddly specific inclination of $63.4^\circ$? Because Earth is not a perfect sphere — its equatorial bulge would ordinarily cause the orbit's perigee (and hence its apogee) to slowly rotate around the orbit, dragging the carefully placed apogee away from the north over months. At the "magic" inclination of $63.4^\circ$, that rotation cancels exactly, and the apogee stays put over the northern hemisphere. This is a consequence of the J2 perturbation from Earth's oblateness, which we will derive properly in Chapter 12; for now, take $63.4^\circ$ as the inclination that freezes the apogee in place. Second, a close cousin called the Tundra orbit uses a full sidereal-day period (not half) and a gentler eccentricity, tracing a lazy figure-eight over one region — the orbit used by some satellite-radio broadcasters for the northern United States.

📜 From History: coverage for a country that faces the pole. The geostationary belt was tailor-made for nations near the equator, and hopeless for the Soviet Union, most of whose territory lies above $50^\circ$ north — where a GEO satellite skims the horizon or vanishes. Rather than fight that geometry, Soviet engineers embraced it: the first Molniya communications satellite launched in 1965, and for decades the Molniya orbit carried television, telephone, and military communications across the USSR's vast north. It is a clean illustration of a recurring lesson in this book — that the "right" engineering answer depends on where you stand, literally. The United States, launching from Florida toward the equator, built a geostationary network; the Soviet Union, facing the pole, invented a completely different orbit for the same job. Geography wrote the requirement, and the requirement chose the orbit.

🔄 Check Your Understanding 1. A Molniya satellite has a period of about 12 hours but is useful to a northern city for only about 8 of those hours per orbit. Where is it during the other ~4 hours, and why is it not useful then? 2. Why does a highly elliptical orbit provide better high-latitude coverage than a geostationary orbit, even though the geostationary satellite is technically "always up"?

Answers

1. During the other few hours it is racing through perigee low over the southern hemisphere (and moving far too fast to serve any one place), then climbing back up. Only while it loiters near its northern apogee is it high in a northern city's sky. 2. A geostationary satellite sits over the equator, so from a high-latitude city it appears very low on the horizon — easily blocked and, past about $81^\circ$ latitude, below the horizon entirely. A Molniya satellite places its slow apogee directly over the north, so it appears high overhead exactly where GEO cannot reach. "Always up" is worthless if "up" means "on the horizon behind a mountain."

9.5 Sun-synchronous and polar orbits

We return to low altitudes for the last and subtlest family — the orbits designed not around where they look, but around when and under what light. These are the workhorses of Earth observation.

Start with the simple idea. If you want to photograph the entire Earth, an orbit tilted near the equator is useless — it only ever passes over the tropics. You want an orbit that goes over the poles, because then, as the Earth rotates underneath, every strip of longitude eventually slides beneath the satellite's path.

Definition (polar orbit). A polar orbit is an orbit with an inclination near $90^\circ$, so that the satellite passes over (or close to) both poles on each revolution. Because the Earth turns beneath the orbit, a polar satellite eventually overflies every point on the planet, which makes polar orbits the natural choice for global mapping, weather, and reconnaissance. They are almost always in LEO, for the same closeness-and-resolution reasons as Section 9.1.

A polar orbit gives you global reach. But there is a second, deeper problem for an imaging satellite: light. If you photograph a city at noon today and at dusk tomorrow, the shadows fall differently and the images are hard to compare — and half the time you would be flying over the night side in darkness. What you really want is to cross every location at the same local solar time every single day: always mid-morning, say, with the same slanting light that throws the same crisp shadows. Astonishingly, you can engineer exactly that.

Definition (sun-synchronous orbit, SSO). A sun-synchronous orbit is a near-polar low Earth orbit whose orbital plane rotates (precesses) eastward at exactly the rate the Earth orbits the Sun — about $0.9856^\circ$ per day, or one full turn per year. As a result, the angle between the orbital plane and the direction to the Sun stays fixed, and the satellite crosses any given latitude at the same local solar time on every pass. Typical sun-synchronous orbits sit near $600$–$800\ \text{km}$ altitude with an inclination of about $98^\circ$ — slightly more than polar.

How can an orbit's whole plane be made to turn at precisely one revolution per year? Not with an engine — that would cost impossible amounts of fuel. The trick is to let Earth's own imperfection do the work. Because Earth bulges at the equator, its gravity is not perfectly central, and this bulge exerts a slow, steady torque that makes an orbit's plane precess — a drift of the ascending node called nodal regression. The rate and direction of that drift depend on the orbit's altitude and inclination. Choose them correctly and the bulge-driven precession comes out to exactly $0.9856^\circ$ per day, eastward, matching Earth's march around the Sun.

Worked Example: the rate the plane must turn. For the lighting angle to stay constant, the orbital plane must keep pace with the Earth–Sun line, which sweeps through $360^\circ$ in one year: $$\dot{\Omega}_{\text{required}} = \frac{360^\circ}{365.24\ \text{days}} = 0.9856^\circ \text{ per day, eastward.}$$ That is the target. The bulge of the Earth (the "J2" effect) supplies a precession of the node whose sign depends on inclination: a normal, prograde orbit ($i < 90^\circ$) drifts westward, the wrong way. To get the plane to drift eastward, you must tilt the orbit past $90^\circ$ into a slightly retrograde inclination — which is why sun-synchronous orbits sit near $98^\circ$, not $90^\circ$. The exact inclination that produces $0.9856^\circ$ per day at a given altitude comes from the J2 precession formula we derive in Chapter 12; at about $800\ \text{km}$ the answer is close to $98.6^\circ$.

(Flagged for the careful reader: we assert the $\approx 98^\circ$ result and its eastward/retrograde sense here but do not derive it — the J2 formula that produces it is Chapter 12's. Treat the inclination as a quoted design value, not something computed in this chapter.)

🚪 Threshold Concept: you can engineer the orbit's plane, not just the orbit. Until now, "choosing an orbit" has meant choosing a size (altitude) and a shape (eccentricity) and a tilt (inclination) — all fixed at launch. The sun-synchronous orbit reveals something richer: the orientation of the orbit in space can itself be designed to change at a chosen rate, by deliberately harnessing a perturbation most people would regard as a nuisance. Earth's equatorial bulge would ordinarily be an error to correct; the sun-synchronous designer turns it into a free, fuel-less motor that spins the orbital plane once a year, on schedule, forever. This is a genuine shift in what "designing an orbit" means — from picking a fixed path to commissioning a slowly turning one — and it is the seed of the whole subject of orbit maintenance and perturbation management in Chapter 12.

🔧 Engineering Reality: why constant lighting is worth so much. Nearly every Earth-observation and weather satellite — Landsat, the Sentinels, the American and European polar weather platforms, and the great majority of imaging reconnaissance satellites — flies sun-synchronous, and the reason is operational, not aesthetic. Constant local time means every image of a place is taken under comparable illumination, so changes between images (a flood, a new building, crop growth, troop movements) stand out instead of being drowned in shifting shadows. A special case, the dawn–dusk sun-synchronous orbit, rides the day–night terminator so the satellite is in near-perpetual sunlight — a gift for solar power and thermal stability that radar-imaging satellites in particular exploit. The lighting is not a nicety; it is often the entire reason the mission can do its job, and it is bought purely by the choice of orbit. We will see how that constant sunlight shapes the power and thermal designs in Chapters 24 and 25.

⚠️ Common Misconception: "Polar and sun-synchronous are the same thing." They overlap but are not identical. Polar is about reach — an inclination near $90^\circ$ so the satellite overflies the whole globe. Sun-synchronous is about timing — a plane that precesses once per year so the lighting stays constant. A sun-synchronous orbit is nearly polar (about $98^\circ$), so it happens to give near-global coverage too, which is why the two are often conflated. But you can have a polar orbit that is not sun-synchronous (a plain $90^\circ$ orbit whose lighting drifts), and the distinction matters: one buys you where, the other buys you when.

🔄 Check Your Understanding 1. Why must a sun-synchronous orbit be tilted slightly past $90^\circ$ (retrograde) rather than exactly at the pole? 2. What single operational advantage does "same local time on every pass" give an imaging mission?

Answers

1. The plane must precess eastward at $0.9856^\circ$/day to follow the Sun. Earth's bulge makes prograde orbits ($i<90^\circ$) precess the wrong way (westward); only a slightly retrograde orbit ($i>90^\circ$, near $98^\circ$) precesses eastward at the needed rate. 2. Comparable illumination in every image of a given place: the same sun angle and shadow directions day after day, so genuine changes on the ground are easy to detect and the satellite is never flying over darkness when it should be imaging.

9.6 Choosing an orbit is mission design

We have toured the filing cabinet. Now step back and see the pattern, because the pattern is the real lesson of the chapter. Every orbit we met was an answer to a requirement, and you can run the logic in reverse: tell me what a mission must do, and its orbit falls out almost deterministically.

• Must you photograph fine detail, or move lots of data with low delay? Go low — LEO — and accept that you will need many satellites for continuous coverage.
• Must you be seen from everywhere at once by cheap receivers, with only a modest fleet? Go to the navigation altitude — MEO — and pay the radiation tax.
• Must you hover over one region so a fixed antenna never moves? Go to the one magic ring — GEO — and accept the distance, the delay, and the crowded slots.
• Must you serve the far north, where the equatorial belt sinks below the horizon? Tilt an ellipse and loiter high over the pole — Molniya / HEO.
• Must you map the whole Earth under identical light? Fly near-polar and let the bulge precess your plane — sun-synchronous.

🧩 Productive Struggle. Before reading on, invent the requirement for each of these real missions and guess its orbit: (a) a satellite that must relay live television to all of Europe from a fixed dish; (b) a satellite that must measure Arctic sea-ice thickness under consistent lighting; (c) a satellite that must let a hiker's receiver anywhere on Earth find its position; (d) a satellite that must photograph a specific battlefield at the highest possible resolution. Match each to LEO, MEO, GEO, or SSO, and say why in one sentence — the "why" is the whole skill.

Answers

(a) GEO — a fixed dish demands a satellite that never moves in the sky, and Europe is at moderate enough latitude for GEO to work. (b) Sun-synchronous — "consistent lighting" is the literal definition of SSO, and near-polar reaches the Arctic. (c) MEO — global visibility to cheap receivers with a small constellation is the navigation problem GPS solved in MEO. (d) LEO — highest resolution means closest, and a single tasked pass is acceptable for a specific target. If you reasoned from the requirement to the orbit, you are now doing mission design.

Notice what this section is really claiming. Orbit selection is not a lookup in a table of memorized facts; it is a trade study, the first one your mission will run. Altitude and inclination are design variables — dials you turn — and every setting has consequences you can compute with the equations of Chapters 6 and 8:

There is no row in that table marked "best." A communications broadcaster and a spy satellite and a GPS spacecraft are all correct — for different missions. This is what it means to say that choosing an orbit is the first act of mission design: you are not looking up the right answer, you are negotiating among coverage, resolution, latency, cost, and survivability until a single altitude and inclination satisfy the mission's real requirement. Every later choice in your Mission Design Review — how much delta-v you need (Chapter 10), which launch vehicle can deliver you (Chapter 30), how you point and power and cool the spacecraft — inherits from the orbit you pick here. Get the orbit right and the rest of the design has a fighting chance; get it wrong and no amount of clever engineering downstream can rescue the mission. The whole discipline of mission design, which we formalize in Chapter 29, begins with this one decision.

Mission Design Checkpoint: select and justify your mission's orbit

This is a major checkpoint — arguably the first one where you make a real engineering decision rather than record a physical fact. Every chapter from here forward will assume you know which orbit your mission flies. Open your Mission Design Review (MDR) document and add an Orbit Selection block.

The decision. Choose the orbit for the mission you picked back in Chapter 1, and justify it in one or two sentences that trace from requirement to orbit. Record three things — altitude (or apogee/perigee, if elliptical), inclination, and the one-line rationale:

• Track A — Communications satellite → geostationary orbit (GEO). Altitude $35{,}786\ \text{km}$, inclination $0^\circ$ (equatorial), circular. Rationale: a fixed ground antenna needs a satellite that hangs motionless over one longitude. Also record the GTO your launcher will drop you into and the $\approx 1.5\ \text{km/s}$ circularization burn, plus an $\approx 11\ \text{m/s}$ end-of-life budget to the graveyard orbit.
• Track B — Lunar lander → low lunar orbit (LLO). A circular orbit $\approx 100\ \text{km}$ above the Moon, from which you will descend. Using the Moon's $\mu_{\text{Moon}} = 4{,}903\ \text{km}^3/\text{s}^2$ and radius $1{,}737\ \text{km}$: $v_{\text{circ}} = \sqrt{4{,}903 / 1{,}837} = 1.63\ \text{km/s}$ and period $\approx 118\ \text{min}$ (about two hours — Apollo's lunar parking orbit). Inclination is chosen to pass over your landing site.
• Track C — Mars orbiter → low Mars science orbit. A near-polar circular orbit $\approx 400\ \text{km}$ above Mars for global mapping. Using $\mu_{\text{Mars}} = 4.283\times10^{4}\ \text{km}^3/\text{s}^2$ and radius $3{,}390\ \text{km}$: $v_{\text{circ}} = \sqrt{4.283\times10^{4}/3{,}790} = 3.36\ \text{km/s}$ and period $\approx 118\ \text{min}$. (Many real Mars orbiters are sun-synchronous about Mars for the same constant-lighting reason as Earth imagers.)
• Track D — Asteroid rendezvous → a heliocentric matching orbit. You do not "orbit" a small asteroid in the usual sense — its escape velocity is centimeters per second — so you match its orbit around the Sun and fly in close formation. Record the target's heliocentric orbit (semi-major axis near $1.1$–$1.2\ \text{AU}$ for a typical accessible near-Earth asteroid); the interplanetary transfer that gets you there is designed in Chapter 11.

The code. Add a small helper, astrotools/orbit_catalog.py, that tabulates the standard Earth regimes and returns a circular orbit's period so you can sanity-check any altitude you consider. (This is a convenience helper for orbit selection, not one of the canonical modules; keep your Chapter 6 orbits.py functions as the real workhorses.)

import math
MU_EARTH = 3.986e5   # km^3/s^2
R_EARTH  = 6371.0    # km (mean)

# regime: (typical altitude km, typical inclination deg, one-line use)
CATALOG = {
    "LEO": (400,   "28-98", "imaging, ISS, broadband; cheapest to reach"),
    "SSO": (800,   "98",    "Earth observation at constant local lighting"),
    "MEO": (20200, "55",    "navigation: GPS / Galileo / GLONASS"),
    "GEO": (35786, "0",     "comms & weather; fixed over one longitude"),
}

def circular_period_min(alt_km):
    a = R_EARTH + alt_km
    return 2 * math.pi * math.sqrt(a**3 / MU_EARTH) / 60.0

for name, (alt, inc, use) in CATALOG.items():
    print(name, alt, "km", round(circular_period_min(alt)), "min")
# Expected output:
# LEO 400 km 92 min
# SSO 800 km 101 min
# MEO 20200 km 718 min
# GEO 35786 km 1436 min

Each period is a lap time you can now read at a glance: LEO's $92$ minutes, GEO's $1{,}436$ minutes (one sidereal day), and the MEO and SSO values in between. Write your chosen orbit's altitude, inclination, and period into your MDR. In Chapter 10 you will compute the delta-v to reach that orbit from where your launcher drops you; by the capstone, this single decision will have shaped your entire vehicle.

(Sanity note on the code: SSO at $800\ \text{km}$ gives $a = 7{,}171\ \text{km}$ and $T = 101\ \text{min}$; the GEO row uses the mean Earth radius, so its altitude-to-period round-trip lands on $1{,}436\ \text{min}$ as it should. The helper is deliberately approximate — for the exact geostationary altitude, use the equatorial radius as in Section 9.3.)

Summary

Every orbit is a purposeful trade of coverage, resolution, latency, delta-v, and radiation. Carry this table forward as a reference:

Key equations, both reused from earlier chapters:

Numbers worth memorizing: GEO altitude $35{,}786\ \text{km}$ (radius $42{,}164\ \text{km}$, speed $3.07\ \text{km/s}$, period one sidereal day $86{,}164\ \text{s}$); GPS altitude $\approx 20{,}200\ \text{km}$ (period $\approx 12\ \text{h}$); sun-synchronous inclination $\approx 98^\circ$; Molniya inclination $63.4^\circ$, eccentricity $\approx 0.74$; LEO period $\approx 90$–$100\ \text{min}$, $\approx 15.6$ orbits per day; solar day $86{,}400\ \text{s}$ versus sidereal day $86{,}164\ \text{s}$.

Spaced Review

Retrieval strengthens memory. Answer from memory before checking, then look back at the cited section. This chapter revisits Chapter 6 (energy and vis-viva) and Chapter 8 (Kepler's laws and the orbital elements).

1. (§9.3, Ch. 6) The geostationary orbit is higher than LEO yet its satellites move slower ($3.07$ versus $7.67\ \text{km/s}$). State the Chapter 6 principle that requires this, in one sentence.
2. (§9.3, Ch. 8) We found the geostationary radius by inverting Kepler's third law. Write the relation between period $T$ and semi-major axis $a$, and explain in words why a longer desired period forces a larger orbit.
3. (§9.4, Ch. 8) A Molniya satellite dwells over the north because it moves slowly near apogee. Which of Kepler's three laws guarantees that a satellite moves slowest at its farthest point, and what quantity does that law hold constant?
4. (§9.1, Ch. 6) Using vis-viva, is the speed of a satellite passing through $800\ \text{km}$ altitude on a Molniya orbit greater than, less than, or equal to the speed of a satellite in a circular orbit at $800\ \text{km}$? Why?

Answers

1. "A higher orbit is a slower orbit": raising an orbit increases its total specific energy but pours that energy (and more) into potential energy, so the kinetic term — the speed — falls. 2. $T = 2\pi\sqrt{a^3/\mu}$; since $T$ grows with $a^{3/2}$, demanding a longer period forces a larger semi-major axis (bigger orbit) — equivalently $a = (\mu T^2/4\pi^2)^{1/3}$. 3. Kepler's second law (equal areas in equal times), which holds the orbit's angular momentum constant; with the same angular momentum, the satellite must move slower when it is farther out. 4. Greater. By vis-viva $v = \sqrt{\mu(2/r - 1/a)}$, the Molniya orbit has a much larger semi-major axis $a$ ($26{,}560\ \text{km}$ versus $7{,}171\ \text{km}$), so $-1/a$ is closer to zero and $v$ is larger at the same $r$ — a body merely passing through that altitude on a big elliptical orbit is moving faster there than one that circularly orbits at that altitude.

What's Next

We now own the catalog: we can name every important orbit, compute its period and speed, and — most importantly — reason from a mission's requirement to the orbit it needs. But knowing which orbit you want is only half of mission design. The other half is getting from the orbit your launcher drops you into to the orbit you actually want — and, once there, moving around at will: raising, lowering, tilting, and rendezvousing. Every one of those moves has a price measured in delta-v, and pricing them is the next tool we need. In Chapter 10 we learn to change orbits: we derive the Hohmann transfer that took our satellite from GTO up to GEO, discover why changing an orbit's plane is shockingly expensive, and price out the delicate dance of rendezvous. Now that we know the destinations, we learn to travel between them.