{
  "id": "keplers-laws",
  "title": "Kepler’s Laws",
  "lang": "en",
  "source_lang": "en",
  "summary": "Three laws describing planetary orbits, their changing speeds, and the relationship between orbital period and size.",
  "category": "science",
  "subcategory": "astronomy",
  "type": "concept",
  "aliases": [
    "Kepler’s laws of planetary motion",
    "Keplerian laws"
  ],
  "generated_at": "2026-11-02",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/keplers-laws/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/keplers-laws/",
      "json": "https://aiwiki.page/zh/keplers-laws.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Kepler’s laws are three mathematical descriptions of planetary motion formulated by Johannes Kepler in the early seventeenth century. The first states that each planet follows an ellipse with the Sun at one focus. The second states that the line joining a planet to the Sun sweeps out equal areas in equal times. The third states that the square of a planet’s orbital period is proportional to the cube of its orbit’s semimajor axis. The first two appeared in 1609; the third followed in 1619.\n\nOriginally established from astronomical observations, especially those of Mars, the laws replaced the assumption that celestial motion must be constructed from uniform circular motions. They became a foundation of astronomy and helped make possible Isaac Newton’s explanation of planetary motion through universal gravitation. In Newtonian mechanics, they are exact consequences of an idealized, isolated two-body system, with an appropriate correction for the masses of both bodies.\n\nReal planetary systems contain many interacting bodies, so their trajectories are not permanently fixed ellipses. Nevertheless, Keplerian motion remains the basic approximation used to describe planets, moons, comets, binary stars, and artificial satellites. The laws connect orbital geometry with elapsed time and gravitating mass, making them useful both for predicting positions and for interpreting astronomical measurements."
    },
    {
      "heading": "Historical development and observational foundations",
      "text": "Before Kepler, mathematical astronomy generally represented planetary motion through combinations of circles. The geocentric system associated with Ptolemy used deferents, epicycles, and an equant to reproduce observed changes in planetary speed. Nicolaus Copernicus placed the planets around the Sun in De revolutionibus orbium coelestium, published in 1543, but retained circular constructions. Heliocentrism changed the arrangement of the planetary system without immediately supplying the elliptical paths and precise timing rules now associated with it.\n\nThe decisive observational material came from Tycho Brahe, whose large instruments measured celestial positions before the astronomical telescope became available. Kepler joined Tycho in Prague in 1600 and subsequently gained access to his observations. Mars was particularly important: its appreciable orbital eccentricity made departures from circular models detectable, while its changing position relative to Earth supplied information about both planets’ motions. Reconstructing an orbit from such observations required separating actual planetary movement from the effects of observing it from a moving Earth.\n\nKepler initially tried combinations of circular motion. In Astronomia nova, published in 1609, he described an approximately eight-arcminute discrepancy that he refused to dismiss as observational error. This was small compared with the apparent diameter of the Moon, but significant relative to the precision of Tycho’s best measurements. The discrepancy helped drive the rejection of a circular account of Mars’s orbit. The resulting investigation established the area rule and the elliptical orbit, although their order of discovery did not match their later numbering as the second and first laws.\n\nThe third law appeared in Harmonices mundi in 1619. It expressed a common mathematical relationship linking the sizes and periods of different planetary orbits, rather than describing motion along a single orbit. Kepler’s search for this relationship formed part of a broader investigation of mathematical harmony in nature. The durable astronomical result was the proportionality between squared periods and cubed semimajor axes.\n\nKepler’s achievement involved both empirical discipline and physical speculation. He sought a solar cause for planetary motion, but did not possess Newton’s concepts of inertia and universal gravitation. His Rudolphine Tables, published in 1627, combined Tycho’s observations with the new orbital theory to improve predictions of planetary positions. Acceptance of the laws developed through their predictive success and their later integration into dynamics; they did not instantly displace every competing astronomical model."
    },
    {
      "heading": "The first law: elliptical orbital geometry",
      "text": "The first law states that a planet’s orbit is an ellipse, with the Sun at one focus. An ellipse is a closed curve for which the sum of the distances to two fixed points, its foci, is constant. Its longest diameter is the major axis; half that diameter is the semimajor axis, \\(a\\). The semiminor axis, \\(b\\), measures half its width in the perpendicular direction. If the distance from the center to either focus is \\(c\\), the eccentricity is \\(e=c/a\\), and\n\n\\[b=a\\sqrt{1-e^{2}}.\\]\n\nFor a noncircular ellipse, \\(0<e<1\\). A circle is the limiting case \\(e=0\\), in which the two foci coincide. Thus the first law does not exclude circular orbits; it identifies them as a special case. The Sun is not at the ellipse’s geometric center, except in this circular limit. The second focus has a geometrical role but need not contain any physical object.\n\nWith the occupied focus as the origin, the orbit can be written\n\n\\[r=\\frac{a(1-e^{2})}{1+e\\cos f},\\]\n\nwhere \\(r\\) is the planet–Sun distance and \\(f\\), called the true anomaly, is the angle measured from the direction of closest approach. The quantity \\(p=a(1-e^{2})\\) is the semilatus rectum. This polar equation expresses orbital shape independently of the timing of motion: it specifies where points lie, but not when the planet reaches them.\n\nThe nearest and farthest distances from the Sun are the perihelion and aphelion distances,\n\n\\[r_{\\mathrm p}=a(1-e), \\qquad r_{\\mathrm a}=a(1+e).\\]\n\nConsequently, \\(a=(r_{\\mathrm p}+r_{\\mathrm a})/2\\). This explains one limited sense in which the semimajor axis is an “average distance.” It is not generally the average distance obtained by sampling equal intervals of time. For a Keplerian ellipse, that time average is \\(a(1+e^{2}/2)\\), because the body spends more time in the more distant parts of its orbit.\n\nEarth’s orbital eccentricity is approximately 0.0167, so its orbit is nearly circular. Mars has an eccentricity of about 0.093, while Mercury’s is about 0.206. These values vary slowly under planetary perturbations. Highly elongated diagrams can therefore misrepresent the appearance of many planetary orbits if treated as scale drawings.\n\nThe first law describes shape, not orientation. An ellipse can be tilted relative to a reference plane and rotated within its own plane. Describing an actual orbit therefore requires additional information about inclination, the direction of its ascending node, and the orientation of perihelion. These quantities become part of the full system of orbital elements."
    },
    {
      "heading": "The second law: equal areas and changing speed",
      "text": "The second law states that the line connecting a planet to the Sun sweeps out equal areas in equal intervals of time. It does not say that the planet travels equal distances or turns through equal angles in those intervals. A planet moves faster near perihelion and more slowly near aphelion, with precisely the variation needed to keep the rate of swept area constant.\n\nFor a short time interval, the swept region is approximately a triangle. Its area is half the product of the radius and the transverse displacement. In the limit,\n\n\\[\\frac{dA}{dt} =\\frac{1}{2}r^{2}\\frac{df}{dt} =\\frac{h}{2},\\]\n\nwhere \\(h=|\\mathbf r\\times\\mathbf v|\\) is the specific angular momentum, or angular momentum per unit mass in the fixed-center approximation. The area law is therefore equivalent to conservation of \\(h\\). Because \\(r^{2}\\dot f\\) is constant, the angular rate increases strongly as the distance decreases.\n\nAt perihelion and aphelion the velocity is perpendicular to the radius, so the area law gives\n\n\\[r_{\\mathrm p}v_{\\mathrm p} =r_{\\mathrm a}v_{\\mathrm a}, \\qquad \\frac{v_{\\mathrm p}}{v_{\\mathrm a}} =\\frac{1+e}{1-e}.\\]\n\nFor Earth, this speed ratio is about 1.034; its heliocentric speed is approximately 30.3 kilometers per second near perihelion and 29.3 kilometers per second near aphelion. The modest difference reflects the orbit’s low eccentricity. For a strongly elongated cometary orbit, the contrast can be much larger.\n\nThe dynamical reason for the area law is that a force directed along the radius produces no torque about the attracting center. A radial force can change the magnitude and direction of velocity, but it cannot change angular momentum about that center. Importantly, the equal-area law does not by itself establish that the force follows an inverse-square dependence. Any central force conserves angular momentum; the particular combination of elliptical shape and the area law identifies the inverse-square form.\n\nThe area rule also explains why intervals between successive seasonal markers are unequal. Earth moves more rapidly during the part of its orbit near its early-January perihelion than near its early-July aphelion. Seasons themselves arise principally from axial tilt, not this distance variation, but the changing orbital speed affects their durations.\n\nThe second law is measured about the force center, not an arbitrarily chosen point such as the ellipse’s geometric center. Its geometrical statement also remains meaningful for open trajectories: an ideal gravitational flyby sweeps out equal areas in equal times even though the passing body never completes a closed orbit."
    },
    {
      "heading": "The third law: orbital periods, sizes, and masses",
      "text": "The third law compares different orbits around the same dominant body. In Kepler’s original planetary form,\n\n\\[T^{2}\\propto a^{3},\\]\n\nwhere \\(T\\) is the orbital period and \\(a\\) is the semimajor axis. Thus, for two planets whose masses are negligible compared with the Sun’s,\n\n\\[\\left(\\frac{T_1}{T_2}\\right)^2 = \\left(\\frac{a_1}{a_2}\\right)^3.\\]\n\nThe semimajor axis is essential: instantaneous distance, perihelion distance, and the ellipse’s circumference cannot replace it in this relation.\n\nIf periods are expressed in Earth years and semimajor axes in astronomical units, the numerical relationship for solar-system planets is approximately \\(T^{2}=a^{3}\\). Mars, with a semimajor axis of roughly 1.524 astronomical units, consequently has a period of about \\(1.524^{3/2}=1.88\\) years. A hypothetical body with a semimajor axis of four astronomical units would have an approximately eight-year period around the Sun. Such numerical simplifications depend on the chosen units and central mass; they are not universal dimensionless identities.\n\nNewtonian mechanics gives the more general expression\n\n\\[T^{2}=\\frac{4\\pi^{2}}{G(M+m)}a^{3},\\]\n\nwhere \\(G\\) is the gravitational constant, \\(M\\) and \\(m\\) are the two masses, and \\(a\\) is the semimajor axis of their relative orbit. For a planet orbiting the Sun, \\(m\\ll M\\), so neglecting the planetary mass is usually a good first approximation. In a binary star system, both masses may be substantial, and replacing \\(M+m\\) with either mass alone would be incorrect.\n\nAn important consequence is that all isolated two-body ellipses with the same semimajor axis and the same total mass have the same period, regardless of eccentricity. A nearly circular orbit and a highly elongated orbit can therefore take exactly the same time to complete. Their path lengths and speed distributions differ, but these differences compensate in the total travel time.\n\nThe third law establishes a scale dependence characteristic of inverse-square gravity. Increasing every orbital length by a factor \\(\\lambda\\), without changing the masses, increases the period by \\(\\lambda^{3/2}\\). Increasing the total mass while holding the orbit’s semimajor axis fixed shortens the period as \\((M+m)^{-1/2}\\). These dependencies distinguish gravitational orbital motion from, for example, motion in an ideal harmonic restoring force, whose period does not vary with oscillation amplitude.\n\nThe relationship also allows observations of size and period to reveal gravitating mass. For that purpose, the orbital size must be known in physical units: an angular orbit on the sky alone does not determine mass unless the distance to the system is also established."
    },
    {
      "heading": "Newtonian derivation and the two-body problem",
      "text": "Newton’s Philosophiæ naturalis principia mathematica, published in 1687, supplied a dynamical framework for Kepler’s empirical relationships. Combining Newton’s laws of motion with universal gravitation gives an attractive force whose magnitude is \\(GMm/r^{2}\\). The same framework describes falling objects near Earth and astronomical orbits, making planetary motion part of a general classical mechanics rather than a separate category of celestial behavior.\n\nFor an isolated pair, define the relative position as \\(\\mathbf r=\\mathbf r_m-\\mathbf r_M\\). Its equation of motion is\n\n\\[\\ddot{\\mathbf r} =-\\frac{\\mu}{r^{3}}\\mathbf r, \\qquad \\mu=G(M+m).\\]\n\nMeanwhile, the center of mass moves uniformly in an inertial reference frame. Separating these two motions reduces the original two-body problem to a single relative-coordinate equation. Neither body must be assumed absolutely stationary.\n\nThe central direction of acceleration conserves specific angular momentum, immediately yielding the area law. Solving the radial dynamics, for example by substituting \\(u=1/r\\) and using orbital angle as the independent variable, gives\n\n\\[\\frac{d^{2}u}{df^{2}}+u=\\frac{\\mu}{h^{2}}.\\]\n\nIts solution is a conic section,\n\n\\[r=\\frac{h^{2}/\\mu}{1+e\\cos f}.\\]\n\nFor bound, noncollision motion, \\(0\\leq e<1\\), so the conic is an ellipse. Comparing this equation with the ellipse’s geometrical form gives \\(h^{2}=\\mu a(1-e^{2})\\).\n\nThe area of an ellipse is \\(\\pi ab\\), and its constant areal velocity is \\(h/2\\). Therefore,\n\n\\[T=\\frac{2\\pi ab}{h}.\\]\n\nSubstituting \\(b=a\\sqrt{1-e^{2}}\\) and the expression for \\(h\\) yields \\(T^{2}=4\\pi^{2}a^{3}/\\mu\\). The third law thus follows from the same dynamics as the first two, rather than requiring an independent assumption.\n\nThe barycenter, or center of mass, clarifies the meaning of “orbiting the Sun.” Both bodies trace ellipses about their shared barycenter. The relative separation traces an ellipse with one body’s position serving as the coordinate origin; this relative ellipse is the one used with \\(G(M+m)\\). The two barycentric semimajor axes satisfy \\(a_M+a_m=a\\) and \\(Ma_M=ma_m\\). These distinctions are minor for a low-mass planet but central to interpreting binary stars.\n\nThe derivation also identifies the conditions under which the laws hold exactly: Newtonian point-mass gravity, or equivalent spherical mass distributions outside their surfaces, an isolated pair, and no additional forces. Real systems can approximate these conditions very closely without satisfying them perfectly."
    },
    {
      "heading": "Calculating positions: anomalies and orbital elements",
      "text": "The first law gives an orbit’s shape, while the second provides a clock along that shape. Turning them into a practical position prediction requires relating elapsed time to orbital angle. The central relation is Kepler’s equation, which, for an ellipse, is\n\n\\[\\mathcal M=E-e\\sin E.\\]\n\nHere \\(E\\) is the eccentric anomaly, a geometrical parameter associated with an auxiliary circle, and \\(\\mathcal M\\) is the mean anomaly. Neither should be confused with the true anomaly \\(f\\), the physical angle at the occupied focus.\n\nMean anomaly advances uniformly:\n\n\\[\\mathcal M=n(t-\\tau), \\qquad n=\\frac{2\\pi}{T} =\\sqrt{\\frac{\\mu}{a^{3}}},\\]\n\nwhere \\(n\\) is the mean motion and \\(\\tau\\) is a time of periapsis passage. The reason Kepler’s equation contains \\(E-e\\sin E\\) is geometrical: the area swept from periapsis is \\(ab(E-e\\sin E)/2\\). Equating the fraction of swept area to the fraction of elapsed period converts the area law into a computable time relation.\n\nOnce \\(E\\) has been found, the position in the orbital plane can be written\n\n\\[x=a(\\cos E-e),\\qquad y=a\\sqrt{1-e^{2}}\\sin E.\\]\n\nThese coordinates place the occupied focus at the origin and the periapsis direction along the positive \\(x\\)-axis. The distance is \\(r=a(1-e\\cos E)\\). If the true anomaly is needed, it follows from\n\n\\[\\tan\\frac{f}{2} = \\sqrt{\\frac{1+e}{1-e}}\\tan\\frac{E}{2},\\]\n\nwith appropriate attention to the angular quadrant.\n\nKepler’s equation generally requires numerical solution rather than inversion by elementary functions. A common iteration is\n\n\\[E_{k+1} = E_k- \\frac{E_k-e\\sin E_k-\\mathcal M} {1-e\\cos E_k}.\\]\n\nFor modest eccentricity, choosing the initial estimate \\(E_0=\\mathcal M\\) often works efficiently. Near \\(e=1\\), especially around periapsis, more careful starting values or safeguarded numerical methods are useful. At small eccentricity, the approximation \\(E\\approx\\mathcal M+e\\sin\\mathcal M\\) illustrates how the nonuniform motion departs from a uniform angular clock.\n\nA complete spatial orbit is commonly specified by six orbital elements: semimajor axis, eccentricity, inclination, longitude of the ascending node, argument of periapsis, and a phase variable at a stated epoch. Rotations determined by the three orientation angles transform the orbital-plane coordinates into a chosen reference frame. Circular or equatorial orbits make some conventional angles undefined, so alternative element sets are often used to avoid these coordinate singularities."
    },
    {
      "heading": "Energy, escape, and open trajectories",
      "text": "Although Kepler formulated his laws for planets, Newtonian gravitation places elliptical orbits within a larger family of conic trajectories. The controlling quantity is the specific orbital energy of relative motion,\n\n\\[\\varepsilon=\\frac{v^{2}}{2}-\\frac{\\mu}{r}.\\]\n\nThe first term is the specific kinetic energy and the second is the gravitational potential energy per unit reduced mass, with zero potential chosen at infinite separation. For an ellipse,\n\n\\[\\varepsilon=-\\frac{\\mu}{2a}.\\]\n\nNegative energy corresponds to bound motion; zero and positive energy permit escape.\n\nCombining these expressions gives the vis-viva equation,\n\n\\[v^{2}=\\mu\\left(\\frac{2}{r}-\\frac{1}{a}\\right).\\]\n\nThis relation specifies speed at any point on an elliptical orbit once \\(a\\), \\(r\\), and \\(\\mu\\) are known. It includes the circular-orbit result \\(v=\\sqrt{\\mu/r}\\). The equation complements the area law: the area law constrains transverse motion, whereas vis-viva gives total speed, including the radial component away from the apsides.\n\nEccentricity links energy and angular momentum through\n\n\\[e^{2}=1+\\frac{2\\varepsilon h^{2}}{\\mu^{2}}.\\]\n\nFor nonzero angular momentum, \\(\\varepsilon<0\\) gives an ellipse, \\(\\varepsilon=0\\) gives a parabola, and \\(\\varepsilon>0\\) gives a hyperbola. These open trajectories still obey the equal-area rule and the inverse-square equation of motion. They do not have a finite orbital period, however, so the third law cannot be applied to them as a period–size relationship.\n\nAt a given distance, the parabolic escape speed is\n\n\\[v_{\\mathrm{esc}}=\\sqrt{\\frac{2\\mu}{r}}.\\]\n\nIn the usual approximation of a negligible-mass object near a spherical body, \\(\\mu\\approx GM\\), giving the familiar surface escape speed when \\(r\\) is the body’s radius. Escape here means reaching arbitrarily large separation with no further propulsion in the ideal two-body model. An object can escape one body while remaining bound to a larger system.\n\nComets illustrate the practical importance of these distinctions. Some have long-period, highly eccentric elliptical orbits; others follow unbound paths through the Solar System. An eccentricity slightly above one in a fitted heliocentric orbit does not, by itself, settle an object’s distant origin, because planetary encounters and the chosen reference frame matter. The relevant classification uses the dynamical state in an appropriate model, often including the system’s barycenter and perturbations by major planets."
    },
    {
      "heading": "Perturbations and limits of Keplerian motion",
      "text": "The Solar System contains many mutually attracting bodies. A planet therefore experiences not only solar attraction but also forces from other planets, moons, and smaller objects. Its actual orbit is not a permanently closed ellipse. Keplerian motion remains useful because the Sun supplies the dominant acceleration for most planets, allowing the remaining effects to be treated through perturbation theory or direct numerical integration.\n\nOne standard description assigns an osculating orbit at each instant: the Keplerian conic having the same position and velocity as the real body at that time. Its orbital elements then change under the perturbing forces. Some changes oscillate over orbital cycles; others accumulate as slow precession or long-term variation. The resulting trajectory can be represented as a continuously changing ellipse, provided that the limitations of this representation are recognized.\n\nGravitational interactions can become especially important near an orbital resonance, where orbital frequencies have a relationship involving small integers. Jupiter’s repeated perturbations help create the Kirkwood gaps in the asteroid belt. Neptune and Pluto occupy a 3:2 mean-motion resonance: Pluto completes roughly two orbits during three of Neptune’s, and the resonant geometry prevents close encounters despite overlap in their ranges of heliocentric distance. Such behavior depends on interactions absent from an isolated two-body model.\n\nA central body’s departure from spherical symmetry is another source of non-Keplerian motion. Earth’s equatorial bulge causes the orbital planes and apsidal directions of satellites to precess. Atmospheric drag lowers the energy of satellites in low Earth orbit, while solar radiation pressure affects high-area-to-mass objects. Cometary outgassing can also produce measurable nongravitational acceleration. These mechanisms alter different orbital quantities and cannot all be represented by merely changing the central mass.\n\nGeneral relativity predicts additional departures from Newtonian ellipses. In the weak-field, nearly isolated case, the leading advance of periapsis per revolution is\n\n\\[\\Delta\\omega = \\frac{6\\pi G(M+m)} {a(1-e^{2})c^{2}},\\]\n\nwhere \\(c\\) is the speed of light. For Mercury, the relativistic contribution is approximately 43 arcseconds per century beyond the principal Newtonian perturbations. Albert Einstein accounted for this residual in 1915.\n\nNear compact objects or at very high precision, a fixed Keplerian ellipse becomes inadequate. Relativistic precession, gravitational time effects, and in close compact binaries the emission of gravitational waves can be essential. These are departures from the assumptions underlying the laws, not inconsistencies within their Newtonian two-body domain."
    },
    {
      "heading": "Astronomical measurement and spacecraft applications",
      "text": "Kepler’s laws make orbital motion a method of measuring otherwise inaccessible masses. For a resolved binary star with a known distance, observations can establish the physical semimajor axis of the relative orbit and its period. The third law then gives the sum of the stellar masses. Measuring how the two stars move about the barycenter supplies their mass ratio, allowing the individual masses to be separated.\n\nFor an exoplanet system, the period can often be measured more directly than orbital size. Repeated transits establish a cycle, while stellar mass estimates and the third law convert that period into a semimajor axis. Radial-velocity observations use spectroscopy and the Doppler effect to measure the star’s line-of-sight motion. A Keplerian fit can determine period, eccentricity, orientation parameters, and velocity amplitude. Without an independent orbital inclination, the companion mass is generally constrained through a minimum mass rather than uniquely determined.\n\nTransit observations contain an additional Keplerian connection. The duration of a transit depends on orbital speed, stellar size, and the chord crossed over the stellar disk. Under a circular-orbit approximation, the measured ratio \\(a/R_\\star\\), combined with the period, constrains stellar mean density:\n\n\\[\\rho_\\star \\approx \\frac{3\\pi}{GT^{2}} \\left(\\frac{a}{R_\\star}\\right)^{3},\\]\n\nwhen the planet’s mass is negligible. Eccentricity changes the speed and star–planet distance during transit, so interpreting this relation requires the appropriate orbital model.\n\nFor artificial satellites, the third law determines the distance needed for a chosen period. A nearly circular Earth orbit with a period equal to one sidereal day has a geocentric radius of about 42,164 kilometers, corresponding to an altitude of approximately 35,786 kilometers above the equator. It is geostationary only if it is also equatorial and prograde; matching the period alone produces a geosynchronous orbit, not necessarily a satellite fixed over one terrestrial longitude.\n\nSpacecraft transfer calculations also begin with Keplerian arcs. A Hohmann transfer between coplanar circular orbits uses an ellipse tangent to both. If the circular radii are \\(r_1\\) and \\(r_2\\), the transfer semimajor axis is \\((r_1+r_2)/2\\), and the coast time between tangency points is half the corresponding orbital period. The vis-viva equation gives the velocity changes needed at departure and arrival.\n\nOperational navigation extends these ideal calculations with tracking data, estimated maneuvers, planetary perturbations, and detailed force models. Keplerian elements remain widely used because they provide a compact, physically interpretable description of a trajectory at a specified epoch, even when its subsequent propagation requires a more complete dynamical model."
    }
  ],
  "links": [
    "johannes-kepler",
    "sun",
    "astronomy",
    "isaac-newton",
    "gravity",
    "ptolemy",
    "nicolaus-copernicus",
    "tycho-brahe",
    "telescope",
    "earth",
    "ellipse",
    "angular-momentum",
    "torque",
    "astronomical-unit",
    "newtons-laws-of-motion",
    "classical-mechanics",
    "inertial-reference-frame",
    "center-of-mass",
    "energy",
    "kinetic-energy",
    "potential-energy",
    "solar-system",
    "perturbation-theory",
    "general-relativity",
    "speed-of-light",
    "albert-einstein",
    "gravitational-wave",
    "binary-star",
    "spectroscopy",
    "doppler-effect"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}