Escolar Documentos
Profissional Documentos
Cultura Documentos
Research Article
Classical and Relativistic Orbital Motions around
a Mass-Varying Body
L. Iorio
INFN-Sezione di Pisa, Viale Unità di Italia, 68, 70125 Bari, Italy
I work out the Newtonian and general relativistic effects due to an isotropic mass loss Ṁ/M of a body on the orbital motion of a
test particle around it; the present analysis is also valid for a variation Ġ/G of the Newtonian constant of gravitation. Concerning
the Newtonian case, I use the Gauss equations for the variation of the elements and obtain negative secular rates for the osculating
semimajor axis a, the eccentricity e, and the mean anomaly M, while the argument of pericenter ω does not experience secular
precession; the longitude of the ascending node Ω and the inclination i remain unchanged as well. The true orbit, instead, expands,
as shown by a numerical integration of the equations of motion with MATHEMATICA; in fact, this is in agreement with the
seemingly counter-intuitive decreasing of a and e because they refer to the osculating Keplerian ellipses which approximate the
trajectory at each instant. A comparison with the results obtained with different approaches by other researchers is made. General
relativity induces positive secular rates of the semimajor axis and the eccentricity completely negligible in the present and future
evolution of the solar system.
particle moving in the gravitational field of a central mass with μ̇ ≡ μ̇|t=t0 . μ̇ will be assumed constant throughout the
experiencing a variation of its GM linear in time. Then, I will temporal interval of interest Δt = t − t0 , as it is, for example,
clarify the meaning of the results obtained by performing a the case for most of the remaining lifetime of the Sun as a
numerical integration of the equations of motion in order to Main Sequence (MS) star [1]. Note that μ̇ can, in principle,
visualize the true trajectory followed by the planet. Concern- be due to a variation of both the Newtonian gravitational
ing the method adopted, I will use the Gauss perturbation constant G and the mass M of the central body, so that
equations [19, 20], which are valid for generic disturbing
accelerations depending on position, velocity, and time, the μ̇ Ġ Ṁ
= + . (3)
“standard” Keplerian orbital elements (the Type I according μ G M
to, e.g., [16]) with the eccentric anomaly E as “fast” angular
variable. Other approaches and angular variables like, for Moreover, while the orbital angular momentum is conserved,
example, the Lagrange perturbation equations [19, 20], the this does not happen for the energy.
Type II orbital elements [16], and the mean anomaly M By limiting ourselves to realistic astronomical scenarios
could be used, but, in my opinion, at a price of major concep- like our solar system, it is quite realistic to assume that
tual and computational difficulties. (Think, e.g., about the
cumbersome expansions in terms of the mean anomaly and μ̇
(t − t0 ) 1 (4)
the Hansen coefficients, the subtleties concerning the choice μ
of the independent variable in the Lagrange equations for
the semimajor axis and the eccentricity [19].) With respect over most of its remaining lifetime: indeed, since Ṁ/M is of
to possible connections with realistic situations, it should be the order of (about 80% of such a mass-loss is due to the
noted that, after all, the Type I orbital elements are usually core nuclear burning, while the remaining 20% is due to
determined or improved in standard data reduction analyses average solar wind) 10−14 yr−1 for the Sun [1], the condition
of the motion of planets and (natural and artificial) satellites. (4) is satisfied for the remaining (the age of the present-
My approach should, hopefully, appear more transparent day MS Sun is 4.58 Gyr, counted from its zero-age MS star
and easy to interpret, although, at first sight, some counter- model [1]) ≈ 7.58 Gyr before the Sun will approach the RGB
intuitive results concerning the semimajor axis and the tip in the Hertzsprung-Russell Diagram (HRD). Thus, I can
eccentricity will be obtained; moreover, for the chosen time treat it perturbatively with the standard methods of celestial
variation of the mass of the primary, no approximations mechanics.
are used in the calculations which are quite straightforward. The unperturbed Keplerian ellipse at epoch t0 , assumed
However, it is important to stress that such allegedly puzzling coinciding with the time of the passage at perihelion t p , is
features are only seemingly paradoxical because they will characterized by
turn out to be in agreement with numerical integrations of
the equations of motion, as explicitly shown by the numerous r = a(1 − e cos E),
pictures depicted. Anyway, the interested reader is advised to
look also at [16] for a different approach. In Section 3 I will 1 − e cos E
dt = dE,
work within the general relativistic gravitoelectromagnetic n
framework by calculating the gravitoelectric effects on cos E − e (5)
all the Keplerian orbital elements of a freely falling test cos f = ,
1 − e cos E
particle in a nonstationary gravitational field. Section 4 is √
devoted to a discussion of the findings of other researchers 1 − e2 sin E
sin f = ,
and contains some numerical calculations concerning the 1 − e cos E
previously mentioned orbital phenomena of LAGEOS and
the Moon. Section 5 summarizes my results. where a and e are the osculating semimajor axis and the
eccentricity, respectively, which fix the
size and the shape of
the unchanging Keplerian orbit, n = μ/a3 is its unperturbed
2. Analytical Calculation of Keplerian mean motion, f is the true anomaly, reckoned
the Orbital Effects by μ̇/μ from the pericentre, and E is the eccentric anomaly. This
would be the path followed by the particle for any t > t p if
By defining the mass loss would suddenly cease at t p . Instead, the true
path will be different because of the perturbation induced
μ ≡ GM (1) by μ̇ and the orbital parameters of the osculating ellipses
approximating the real trajectory at each instant of time will
slowly change in time.
at a given epoch t0 , the acceleration of a test particle orbiting
a central body experiencing a variation of μ is, to first order
in t − t0 , 2.1. The Semimajor Axis and the Eccentricity. The Gauss
equation for the variation of the semimajor axis a is [19, 20]
μ(t) μ μ̇
A=− r ≈ − 2 1 + (t − t0 ) r, da 2 p
r2 r μ
(2) = √ eAr sin f + Aτ , (6)
dt n 1−e 2 r
SRX Physics 3
where Ar and Aτ are the radial and transverse, that is, or- Moreover, the osculating Keplerian period
thogonal to the direction of r, components, respectively, of
the disturbing acceleration, and p = a(1 − e2 ) is the semilatus a3
Kep 2π
P ≡ = 2π , (15)
rectum. In the present case n μ
μ̇
which, by definition, yields the time elapsed between two
A ≡ Ar = − 2
t − tp , (7)
r consecutive perihelion crossings in absence of perturbation,
that is, there is an entirely radial perturbing acceleration. that is, it is the time required to describe a fixed osculating
For μ̇ < 0, that is, a decrease in the body’s GM, the total Keplerian ellipse, decreases according to
gravitational attraction felt by the test particle, given by (2), is 3/2
reduced with respect to the epoch t p . In order to have the rate dP Kep 3 ȧ 6πeμ̇ a
= P Kep = . (16)
of the semimajor axis averaged over one (Keplerian) orbital dt 2 a (1 − e) μ
revolution, (7) must be inserted into (6), evaluated onto the
unperturbed Keplerian ellipse with (5), and finally integrated As I will show, also such a result is not in contrast with the
over ndt/2π from 0 to 2π because n/2π = 1/P Kep (see below). genuine orbital evolution.
Note that, from (5), there can be obtained
E − e sin E 2.2. The Pericentre, the Node, and the Inclination. The Gauss
t − tp = . (8) equation for the variation of the pericentre ω is [19, 20]
n
√
As a result, I have dω 1 − e2 r dΩ
= −Ar cos f + Aτ 1 + sin f − cos i ,
2π dt nae p dt
da e μ̇ (E − e sin E) sin E
=− a dE (17)
dt π μ 0 (1 − e cos E)2
(9)
e μ̇ where i and Ω are the the inclination and the longitude of
=2 a. the ascending node, respectively, which fix the orientation of
1−e μ
the osculating ellipse in the inertial space. Since dΩ/dt and
Note that if μ decreases, a gets reduced as well: ȧ < 0. di/dt depend on the normal component Aν of the disturbing
This may be seemingly bizarre and counter-intuitive, but, as acceleration, which is absent in the present case, and A = Ar ,
it will be shown later, it is not in contrast with the true orbital I have
motion. √
2π
dω 1 − e2 μ̇ (E − e sin E)(cos E − e)
The Gauss equation for the variation of the eccentricity = dE = 0 :
is [19, 20] dt 2πe μ 0 (1 − e cos E)2
√
(18)
de 1 − e2 1 r
= Ar sin f + Aτ cos f + 1− . the osculating ellipse does not change its orientation in the
dt na e a
(10) orbital plane, which, incidentally, remains fixed in the inertial
space because Aν = 0 and, thus, dΩ/dt = di/dt = 0.
For A = Ar , it reduces to
de 1 − e2 da 2.3. The Mean Anomaly. The Gauss equation for the mean
= , (11)
dt 2ae dt anomaly M, defined as M = n(t − t p ), [19, 20] is
so that dM 2 r dω dΩ
=n− Ar − 1 − e2 + cos i . (19)
de μ̇ dt na a dt dt
= (1 + e) ; (12)
dt μ It turns out that, since
also the eccentricity gets smaller for μ̇ < 0. 2 r 2μ̇
As a consequence of the found variations of the oscu- − Ar dt = 3 3 (E − e sin E)dE,
na a na
lating semimajor axis and the eccentricity, the osculating (20)
orbital angular momentum per unit mass, defined by L2 = dM μ̇
= n + 2π ;
μa(1 − e2 ), remains constant: indeed, by using (9) and (12), dt μ
it turns out
the mean anomaly changes uniformly in time at a slower rate
dL2 with respect to the unperturbed Keplerian case for μ̇ < 0.
= μȧ 1 − e2 − 2μaeė = 0. (13)
dt
The osculating total energy E = −μ/2a decreases 2.4. Numerical Integration of the Equations of Motion and
according to Explanation of the Seeming Contradiction with the Analytical
Results. At first sight, the results obtained here may be rather
dE μ e μ̇ confusing: if the gravitational attraction of the Sun reduces in
= ȧ = . (14)
dt 2a2 1−e a time because of its mass loss, the orbits of the planets should
4 SRX Physics
with 0.4
sin E E2 + 2e(cos E − 1)
D(E) = e −2(sin E − E cos E) +
1 − e cos E 0.3
Δr (AU)
1 − e2 sin2 E(sin E − e cos E)
− , 0.2
(1 − e cos E)2
1 − e2 0.1
F (E) =
1 − e cos E
× cos E(sin E − E cos E) 0.5 1 1.5 2 2.5 3
t (yr)
1 + e − (1 + e) cos E − E sin E
+ sin E . Figure 2: Difference Δr(t) between the radial distances obtained
1 − e cos E
(26) from the solutions of two numerical integrations with MATHE-
MATICA of the equations of motion over 3 years with and without
From (25) and (26) it turns out that for E > 0 Δr(E) never μ̇/μ; the initial conditions are the same. Just for illustrative purposes
vanishes; after one orbital revolution, that is, after that an a mass loss rate of the order of −10−1 yr−1 has been adopted for the
Sun; for the planet initial conditions corresponding to a = 1 AU,
angular interval of 2π has been swept by the (osculating)
e = 0.01 have been chosen. The cumulative increase of the Sun-
radius vector, a net increase of the radial (osculating) planet distance induced by the mass loss is apparent.
distance occurs according to (according to (25) and (26),
Δr(0) = 0)
it is possible to obtain from (23) and (30)
2π μ̇
Δr(2π) − Δr(0) = Δr(2π) = − a (1 − e). (27) √
n μ a μ̇ 1 − e2
Δτ(E) =
This analytical result is qualitatively confirmed by the n μ (1 − e cos E) (31)
difference Δr(t) between the radial distances obtained from
× G(E) + H (E) + I(E) + J(E) + K(E) ,
the solutions of two numerical integrations with MATHE-
MATICA of the equations of motion over 3 yr with and with
without μ̇/μ; the initial conditions are the same. (Strictly G(E) = sin E(E cos E − sin E),
speaking, Δr and the quantity plotted in Figure 2 are different
objects, but, as the following discussion will clarify, I can (1 − e cos E)
assume that, in practice, they are the same.) For illustrative H (E) = [(1 + e)(cos E − 1) + E sin E],
e
purposes I used a = 1 AU, e = 0.01, μ̇/μ = −0.1 yr−1 . The
result is depicted in Figure 2. Note also that (25) and (26) tell I(E) = E2 + 2e(cos E − 1),
(32)
us that the shift at the aphelion is
1 − e2 (e cos E − sin E)
J(E) = sin E ,
1 1+e 1 − e cos E
Δr(π) = Δr(2π), (28)
2 1−e
1 − e2 (1 + e)(1 − e cos E) − E sin E
in agreement with Figure 1 where it is 4.5 times larger than K(E) = .
e 1 − e cos E
the shift at the perihelion.
Since Figure 1 tells us that the orbital period gets larger From (31) and (32) it turns out that for E > 0 Δτ(E) never
than the Keplerian one, it means that the true orbit must vanishes; at the (osculating) time of perihelion passage
somehow remain behind with respect to the Keplerian one.
Thus, a negative perturbation Δτ in the transverse direction 4π 2 μ̇ 1+e
Δτ(2π) − Δτ(0) = a < 0. (33)
must occur as well; see Figure 3. n μ 1−e
Let us now analytically compute it. According to [21], it This means that when the Keplerian path has reached the
can be used perihelion, the perturbed orbit is still behind it. Such features
a sin E are qualitatively confirmed by Figure 1.
Δτ = √ Δe + a 1 − e2 ΔE + r(Δω + ΔΩ cos i). From a vectorial point of view, the radial and transverse
1−e 2
(29) perturbations to the Keplerian radius vector r yield a
correction:
By recalling that, in the present case, ΔΩ = 0 and using
√
Δ = Δr r + Δτ τ , (34)
1 − e2 μ̇ 1 + e − (1 + e) cos E − E sin E
Δω = − , so that
ne μ 1 − e cos E
(30) rpert = r + Δ. (35)
6 SRX Physics
y (AU)
1
True orbit
r Δr
Line of the x (AU)
−1 −0.5 0.5 1
rpert Δτ Δ apsides
−0.5
−1
Figure 3: Radial and transverse perturbations Δr and Δτ of the
Keplerian radius vector (in blue); the presence of the transverse
perturbation Δτ makes the real orbit (in red) lagging behind the Figure 4: Black continuous liner represents true trajectory obtained
Keplerian one. by numerically integrating with MATHEMATICA the perturbed
equations of motion in Cartesian coordinates over 2 years; the
disturbing acceleration (2) has been adopted. The planet starts from
a point on the x axis. Just for illustrative purposes, a mass loss rate of
The length of Δ is the order of −10−2 yr−1 has been adopted for the Sun; for the planet
initial conditions corresponding to a = 1 AU, e = 0.0 have been
Δ(E) = Δr(E)2 + Δτ(E)2 . (36) chosen. Red dashed line represents unperturbed Keplerian circle at
t = t0 . Blue dash-dotted line represents osculating Keplerian circle
after the first x axis crossing. As can be noted, its semimajor axis and
Equations (27) and (31) tell us that at perihelion it amounts
eccentricity are equal to those of the initial unperturbed circle. Note
to also that after 2 years the planet has not yet reached the x axis as it
would have done in absence of mass loss.
(1 + e)
Δ(2π) = Δr(2π) 1 + 4π 2 . (37)
(1 − e)3
Δτ(E)
tan ξ(E) = ; (38) δ ≈ Δr(2π). (43)
Δr(E)
at perihelion it is This explains why Figure 2 gives us just Δr.
√ Since the approximate calculations of other researchers
1+e often refer to circular orbits, and in view of the fact that
tan ξ(2π) = −2π , (39)
(1 − e)3/2 when a Sun-like star evolves into a giant tidal interactions
circularize the orbit of a planet [23, 24], it is interesting to
that is, ξ is close to −90 deg; for the Earth it is −81.1 deg. consider also such limiting case in which other nonsingular
Thus, the difference δ between the lengths of the perturbed osculating orbital elements must be adopted. (This fact has
radius vector rpert and the Keplerian one r at a given instant been quantitatively proven by the observation of convective
amounts to about binary stars [22].) The eccentricity and the pericentre lose
their meaning: thus, it is not surprising that (12), although
δ ≈ Δ cos ξ; (40) formally valid for e → 0, yields a meaningless result;
that is, the eccentricity would become negative. Instead, the
if fact, this is precisely the quantity determined over 3 years semimajor axis is still valid and (9) predicts that ȧ = 0
by the numerical integration of Figure 2. At the perihelion I for e → 0. The constancy of the osculating semimajor
have axis is not in contrast with the true trajectory, as clearly
showed by Figure 4. Again, the true orbital period is larger
(1 + e) than the Keplerian one which, contrary to the eccentric case,
δ = Δr(2π) 1 + 4π 2 cos ξ; (41)
(1 − e)3 remains fixed. Since D(E) = 0 for e = 0 and F (2π)|e=0 =
−2π, F (0)|e=0 = 0, the radial shift per revolution is
since for the Earth
(1 + e) 2π μ̇
1 + 4π 2 cos ξ = 1.0037, (42) Δr(2π)|e=0 = − a . (44)
(1 − e)3 n μ
SRX Physics 7
nae sin f
vr = √ ,
0.5 1 1.5 2 2.5 3 1 − e2
t (yr) (49)
na 1 + e cos f
vτ = √ ,
Figure 5: Difference Δr(t) between the radial distances obtained 1 − e2
from the solutions of two numerical integrations with MATHE-
MATICA of the equations of motion over 3 years with and without the radial and transverse components of (48), evaluated onto
μ̇/μ; the initial conditions are the same. Just for illustrative purposes the unperturbed Keplerian orbit, are
a mass loss rate of the order of 10−1 yr−1 has been adopted for the
Sun; for the planet initial conditions corresponding to a = 1 AU, 3μ̇ ne sin E
e = 0.0 have been chosen. The cumulative increase of the Sun-
Ar = − ,
c2 (1 − e cos E)2
planet distance induced by the mass loss is apparent. √ (50)
3μ̇ n 1 − e2
Aτ = − 2 .
c (1 − e cos E)2
Also in this case the secular increase of the radial distance is
present, as qualitatively shown by Figure 5. Concerning Δτ, After lengthy calculations they yield
after 2π it is
da 6μ̇ 2
4π 2 μ̇ =− √ −1 ,
Δτ(2π) = a ; (45) dt c2 1 − e2
n μ (51)
de 3μ̇μ 1
also in this case, the orbital period is larger than the = 2 4 (2 − e) 1 − √ .
dt c ae 1 − e2
unperturbed one.
Contrary to the classical case, now both the osculating
semimajor axis and the eccentricity increase for μ̇ < 0. It
3. The General Relativistic Case turns out that the pericentre and the mean anomaly do
The field equations of general relativity are nonlinear, but in not secularly precess. Also in this case the inclination and
the slow-motion (|β| = |v|/c = 1) and weak-field (U/c2 = 1) the node are not affected The qualitative features of the
approximation they get linearized resembling to the linear motion with the perturbation (48) are depicted in Figure 6 in
equations of the Maxwellian electromagnetism; here v and which the magnitude of the relativistic term has been greatly
U are the magnitudes of the typical velocities and the enhanced for illustrative purposes.
gravitational potential of the problem under consideration.
This scenario is known as gravitoelectromagnetism [25, 26]. 4. Discussion of Other Approaches and
In this case the space-time metric is given by Numerical Calculations
Φ 2 2 4 Φ
ds2 = 1 − 2 c dt + (H · dr)dt − 1 + 2 2 δi j dxi dx j , Here I will briefly review some of the results obtained by
c2 c c
(46) others by comparing with ours.
Hadjidemetriou in [14] uses a tangential perturbing
where, far from the source, the dominant contributions to acceleration proportional to the test particle’s velocity v,
the gravitoelectromagnetic potentials can be expressed as
1 μ̇
μ GJ×r A=− v, (52)
Φ= , H= . (47) 2 μ(t)
r c r3
Here J is the proper angular momentum of the central body and a different perturbative approach by finding that, for a
of mass M and r is so that r GM/c2 and r J/(Mc); c is generic mass loss, the semimajor axis secularly increases and
the speed of light in vacuum. For a nonstationary source the the eccentricity remains constant. In fact, with the approach
geodesic equations of motion yield [27], among other terms, followed here it would be possible to show that, to first order
−βi (3 − β2 )Φ,0 , i = 1, 2, 3 which, to order O(c−2 ), reduces to in (μ̇/μ)(t − t0 ), ȧ = −(μ̇/μ)a and ė = 0 and that the
true orbit is expanding, although in a different way with
μ̇ v GM Ġ Ṁ v respect to (2) as depicted by Figure 7 in which the magnitude
A = −3 = −3 2 + . (48)
c2 r c G M r of the mass-loss has been exaggerated for better showing its
8 SRX Physics
Concerning the observationally determined increase of As a suggestion to other researchers, it would be very
the Astronomical Unit, more recent estimates from process- important to complement my analytical two-body calcu-
ing of huge planetary data sets by Pitjeva point towards a rate lation by performing simultaneous long-term numerical
of the order of 10−2 m yr−1 [29, 30]. It may be noted that integrations of the equations of motion of all the major
my result for the secular variation of the terrestrial radial bodies of the solar system by including a mass-loss term
position on the line of the apsides would agree with such in the dynamical force models as well to see if the N-body
a figure by either assuming a mass loss by the Sun of just interactions in presence of such an effect may substantially
−9 × 10−14 yr−1 or a decrease of the Newtonian gravitational change the picture outlined here. It would be important
constant Ġ/G ≈ −1 × 10−13 yr−1 . Such a value for the especially in the RGB phase in which the inner regions of the
temporal variation of G is in agreement with recent upper solar system should dramatically change.
limits from Lunar Laser Ranging [31] Ġ/G = (2 ± 7) × 10−13
yr−1 . This possibility is envisaged in [32] whose authors
Acknowledgment
use ȧ/a = −Ġ/G by speaking about a small radial drift of
−(6 ± 13) × 10−2 m yr−1 in an orbit at 1 AU. I, now, apply The authors thanks Professor K.V. Kholshevnikov, St. Peters-
my analysis to the daily decrement of the semimajor axis of burg State University, for useful comments and references.
LAGEOS [8] whose geocentric relevant orbital parameters
are a = 12270 km, e = 0.0045. From (9) it turns out
that a secular decrease of a of the order of ≈ 1 mm d−1 References
could only be induced by μ̇⊕ /μ⊕ = −3 × 10−6 yr−1 . Clearly,
[1] K.-P. Schröder and R. C. Smith, “Distant future of the Sun
it cannot be due to a variation of the Earth’s mass M; if and Earth revisited,” Monthly Notices of the Royal Astronomical
it had to be attributed to a variation of G, it would be Society, vol. 386, no. 1, pp. 155–163, 2008.
orders of magnitude too large with respect to the bounds [2] G. A. Krasinsky and V. A. Brumberg, “Secular increase of
in [31, 32]. Adopting (27) does not substantially alter the astronomical unit from analysis of the major planet motions,
situation because the required μ̇⊕ /μ⊕ would be only two and its interpretation,” Celestial Mechanics and Dynamical
orders of magnitude smaller. Now I look at the anomalous Astronomy, vol. 90, no. 3-4, pp. 267–288, 2004.
increase of the eccentricity of the lunar orbit amounting to [3] E. M. Standish, “The Astronomical Unit now,” in Transits of
ėMoon = (0.9 ± 0.3) × 10−11 yr−1 [9]. Such a figure and (12) Venus: New Views of the Solar System and Galaxy, D. W. Kurtz,
yield μ̇⊕ /μ⊕ = 8.5 × 10−12 yr−1 ; it is one order of magnitude Ed., Proceedings of the IAU Colloquium, no. 196, pp. 163–179,
larger than the present-day bounds on Ġ/G [31, 32]. Cambridge University Press, Cambridge, UK, 2005.
[4] P. D. Noerdlinger, “Solar mass loss, the Astronomical Unit, and
the scale of the solar system,” http://arxiv.org/abs/0801.3807.
5. Conclusions [5] S. A. Klioner, “Relativistic scaling of astronomical quantities
and the system of astronomical units,” Astronomy and Astro-
I started in the framework of the two-body Newtonian physics, vol. 478, no. 3, pp. 951–958, 2008.
dynamics by using a radial perturbing acceleration linear [6] W. Jin, P. Imants, and M. Perryman, Eds., A Giant Step: From
in time and straightforwardly treated it with the standard Milli-to Micro-Arcsecond Astrometry, Proceedings of the IAU
Gaussian scheme. I found that the osculating semimajor Symposium 248, Cambridge University Press, Cambridge,
axis a, the eccentricity e and the mean anomaly M secu- UK, 2008.
larly decrease while the argument of pericentre ω remains [7] W. H. Oskay, S. A. Diddams, E. A. Donley, et al., “Single-atom
unchanged; the longitude of the ascending node Ω and the optical clock with high accuracy,” Physical Review Letters, vol.
inclination i are not affected by the phenomenon considered. 97, no. 2, Article ID 020801, 2006.
[8] D. P. Rubincam, “On the secular decrease in the semimajor
The radial distance from the central body, taken on the fixed
axis of Lageos’s orbit,” Celestial Mechanics, vol. 26, no. 4, pp.
line of the apsides, experiences a secular increase Δr. For 361–382, 1982.
the Earth, such an effect amounts to about 1.3 cm yr−1 . By [9] J. D. Anderson and M. M. Nieto, “Astrometric solar-system
numerically integrating the equations of motion in Cartesian anomalies,” in Relativity in Fundamental Astronomy: Dynam-
coordinates I found that the real orbital path expands after ics, Reference Frames, and Data Analysis, S. Klioner, P. K.
every revolution, the line of the apsides does not change, and Seidelmann, and M. Soffel, Eds., Proceedings of the IAU
the apsidal period is larger than the unperturbed Keplerian Symposium, no. 261, 2009.
one. I have also clarified that such results are not in contrast [10] E. Strömgren, “ber die Bedeutung kleiner Massennderun-
with those analytically obtained for the Keplerian orbital gen fr die Newtonsche Centralbewegung,” Astronomische
elements which, indeed, refer to the osculating ellipses Nachrichten, vol. 163, no. 9, pp. 129–136, 1903.
approximating the true trajectory at each instant. I also [11] J. H. Jeans, “Cosmogonic problems associated with a secular
computed the orbital effects of a secular variation of the decrease of mass,” Monthly Notices of the Royal Astronomy
Society, vol. 85, pp. 2–11, 1924.
Sun’s mass in the framework of the general relativistic lin-
[12] J. H. Jeans, Astronomy and Cosmogony, Dover, New York, NY,
earized gravitoelectromagnetism which predicts a perturbing USA, 1961.
gravitoelectric tangential force proportional to v/r. I found [13] G. Armellini, “The variation of the eccentricity in a binary
that both the semimajor axis and the eccentricity secularly system of decrasing mass,” The Observatory, vol. 58, pp. 158–
increase; the other Keplerian elements remain constant. Such 159, 1935.
effects are completely negligible in the present and future [14] J. D. Hadjidemetriou, “Two-body problem with variable mass:
evolution of, for example, the solar system. a new approach,” Icarus, vol. 2, pp. 440–451, 1963.
10 SRX Physics