1: %%
2: %% Beginning of file 'ms.tex'
3: %%
4: %% Start 2003/12/08 Version 0.01
5: %% Modified 2005/06/10 Version 1.01.98
6: %%
7: %% Author : JI Jianghui
8: %% Affiliation: Purple Mountain Observorary,Chinese Academy of Sciences
9: %% Address : 2 West Beijing Road, Nanjing 210008, P.R.China
10: %% Phone : +86-25-83332051
11: %% Fax : +86-25-83332050
12: %% Email : jijh@pmo.ac.cn OR jijh@jlonline.com
13: %% This is a sample manuscript marked up using the
14: %% AASTeX v5.x LaTeX 2e macros.
15: \documentclass[12pt,preprint]{aastex}
16: %%\documentclass[10pt,preprint]{aastex}
17: %%\documentclass[manuscript]{aastex}
18: %%\documentclass[10pt,preprint]{emulateapj}
19: %% preprint2 produces a double-column, single-spaced document:
20: %%\documentclass[preprint2]{aastex}
21:
22: \shorttitle{Dynamical structure of the 47 UMa planetary system}
23: \shortauthors{Ji J. et al}
24:
25: %%\slugcomment{submitted to ApJ}
26:
27: \begin{document}
28:
29: \title{Could the 47 UMa Planetary System be a Second Solar System: predicting the Earth-like planets}
30: \author{Jianghui JI\altaffilmark{1,3}, Lin LIU\altaffilmark{2,3},
31: Hiroshi Kinoshita\altaffilmark{4}, Guangyu LI\altaffilmark{1,3}}
32: \email{jijh@pmo.ac.cn}
33:
34: \altaffiltext{1}{Purple Mountain Observatory , Chinese Academy
35: of Sciences, Nanjing 210008, China}
36: \altaffiltext{2}{Department of Astronomy, Nanjing University,
37: Nanjing 210093, China}
38:
39: \altaffiltext{3}{National Astronomical Observatory, Chinese
40: Academy of Sciences,Beijing 100012,China}
41: \altaffiltext{4}{National Astronomical Observatory,
42: Mitaka, Tokyo 181-8588,Japan}
43:
44: \begin{abstract}
45: We numerically investigated the dynamical architecture of 47 UMa
46: with the planetary configuration of the best-fit orbital solutions
47: by Fischer et al. We systematically studied the existence of
48: Earth-like planets in the region 0.05 AU $\leq a \leq 2.0$ AU for
49: 47 UMa with numerical simulations, and we also explored the packed
50: planetary geometry and Trojan planets in the system. In the
51: simulations, we found that "hot Earths" at 0.05 AU $\leq a < $ 0.4
52: AU can dynamically survive at least for 1 Myr. The Earth-like
53: planets can eventually remain in the system for 10 Myr in areas
54: involved in the mean motion resonances (MMR) (e.g., 3:2 MMR ) with
55: the inner companion. Moreover, we showed that the 2:1 and 3:1
56: resonances are on the fringe of stability, while the 5:2 MMR is
57: unstable. Additionally, the 2:1 MMR marks out a remarkable
58: boundary between chaotic and regular motions, inside, most of the
59: orbits can survive, outside, they are mostly lost in the orbital
60: evolution. In a dynamical sense, the most likely candidate for
61: habitable environment is Earth-like planets with orbits in the
62: ranges 0.8 AU $\leq a < 1.0$ AU and 1.0 AU $ < a < 1.30$ AU
63: (except 5:2 MMR and several unstable cases) with relatively low
64: eccentricities. The Trojan planets with low eccentricities and
65: inclinations can secularly last at the triangular equilibrium
66: points of the two massive planets. Hence, the 47 UMa planetary
67: system may be a close analog to our solar system, bearing a
68: similar dynamical structure.
69: \end{abstract}
70:
71: \keywords{celestial mechanics-methods:n-body simulations-planetary
72: systems-stars:individual (47 UMa)}
73:
74: \section{Introduction}
75: The main sequence star 47 UMa is of spectral type G0 V with a mass
76: of 1.03$M_{\odot}$. Butler \& Marcy (1996) reported the discovery
77: of the first planet in the 47 UMa system which has become one of
78: the most eye-catching systems particularly after the subsequent
79: release of an additional companion (Fischer et al. 2002). It is
80: sometimes thought to be a close analog of our own solar system:
81: for example, the mass ratio of the two giant companions in 47 UMa
82: is $\sim$ 2.62 (see Table 1), as compared to that of
83: Jupiter-Saturn (JS) of 3.34; and the ratios of two orbital periods
84: are close to each other. Following the analogy, one may wonder
85: whether there exists additional members in 47 UMa system.
86: Multi-planet systems seem to be common in nature (Fischer et al.
87: 2003; hereafter Paper I) more than ten of such systems have been
88: detected in Doppler surveys to date. Moreover, if the solar system
89: is a typical one, how are we to understand the analogous
90: structures in other multi-planet systems, e.g., the presence of
91: the inner low-mass terrestrial planets, the asteroidal belts or
92: Kuiper belts (e.g. \textit{Spitzer} Mission) around central stars?
93: The pioneering numerical experiments (Adams \& Laughlin 2003)
94: showed that planet-planet scattering can yield planetary orbits
95: with $a\sim 1$ AU for a packed system consisting of 10 planets
96: with widely spaced orbits. However, from an observational
97: viewpoint, because of the variability of stellar atmospheres, the
98: current ground-based observational precision $\sim$ 2-3 m s$^{-1}$
99: (Butler et al. 2003) indeed places a fundamental limit on the
100: discovery of Earth-mass planets\footnote{however, the recent
101: discovered Neptune-mass planets with close-in orbits ($a<0.1$ AU)
102: (Butler et al. 2004; McArthur et al. 2004; Santos et al. 2004) are
103: grouped as a new category linking giant gaseous planet and
104: Earth-mass planet.} around 1 AU (even smaller), and the preference
105: detection of planets with masses of several $M_{\oplus}$ should
106: benefit from future space projects that can carry out high
107: resolution astrometric measurements (e.g., SIM and GAIA).
108: Nevertheless, in the current stage, this leaves room for
109: researchers to study the dynamical structure of such planets or
110: their formation that may possibly reside in the systems, in
111: advance of the observing missions.
112:
113: \section{Formation scenario}
114: The planetary formation scenarios may be replayed as follows: in
115: the standard model of core-accretion scenario (Safronov 1969;
116: Lissauer 1993), solid cores (or planetary embryos) rapidly grow
117: larger by the accretion of kilometer-sized planetesimals via
118: runaway growth (Kokubo \& Ida 1998; Ida \& Lin 2004), leading to
119: the formation of the gaseous giant planets through the accretion
120: of disk gas onto Earth-mass solid cores and more planetesimals
121: into an envelope. When they formed outside the ice boundary $\sim
122: 3$ AU (Ida \& Lin 2004) in the system, two giant planets may
123: migrate inward through disk-planet interactions (Ward 1997; Kley
124: 2000; Nelson et al. 2000; Papaloizou 2003), and such migration
125: process may not halt until the two planets are eventually captured
126: into $\sim$ 8:3 mean motion resonance (Laughlin, Chambers \&
127: Fischer 2002; hereafter Paper II), see also the 2:1 resonant
128: capture scenario for two massive planets in GJ 876 (Lee \& Peale
129: 2002; Kley, Peitz, \& Bryden 2004). In the migration scenario, the
130: giant protoplanets may clear up wide gaps about their orbits
131: (Goldreich, Lithwick, \& Sari 2004) and cease accreting small
132: bodies. But the giant planets may make the residual unaccreted
133: planetesimals in the debris disk move inward or outward through
134: gravitational scattering mechanism, even with the eccentricities
135: and inclinations of these planetesimals excited by sweeping
136: secular resonance (Nagasawa \& Ida 2000), thus the orbits of the
137: small bodies can undergo mutual crossings and then they are either
138: directly cleared up or collided into fragments in the
139: post-formation stage. Another possible fate for these isolated
140: members is that they may be captured into mean motion resonances
141: with the giant planets under migration (e.g., the capture of the
142: KBOs in the Neptune migration [Malhotra, Duncan, \& Levison
143: 2000]), and such resonant swarm may outline the structure of
144: asteroidal belts in the system. In consequence, a dynamically
145: quasi-stable configuration with spaced orbits may be achieved to
146: support further secular evolution, which may determine the
147: resulting destinies of those members, leading to the final
148: assemblage of the planetary system (Lin 2004).
149:
150: Paper II and Gozdziewski (2002) studied the long-term stability of
151: 47 UMa and pointed out that the secular apsidal resonance can help
152: stabilize the two giant planets in an aligned configuration with
153: the libration of their relative periapse longitudes (Ji et al.
154: 2003), then the eccentricities are well maintained to free from
155: larger vibrations due to this mechanism, as a result, this system
156: can even survive for billion years (Barnes \& Quinn 2004). Several
157: pioneer works were concentrated on the structure of the system and
158: presented a preliminary understanding of this issue. Jones, Sleep
159: \& Chambers (2001) investigated the existence of Earth-mass
160: planets in the presence of one known giant planet, and
161: subsequently Paper II and Asghari et al. (2004) further studied
162: the stability of massless test particles about the so-called
163: Habitable Zones (HZ) according to some earlier solutions (Fischer
164: et al. 2002), where the dynamical model was treated as a
165: restricted multi-body problem. Nevertheless, as the terrestrial
166: planets possess significant masses, they can interact with the two
167: giant planets by mutual gravitation, which may result in secular
168: effects for the planetary system. Accordingly, we should take into
169: account the masses of terrestrial bodies in the model when
170: exploring the dynamical architecture. In this paper, we performed
171: extensive simulations to examine the dynamical architecture in
172: both the HZ and extended areas, for Earth-like planets (with
173: masses from 0.1 $M_{\oplus}$ to 10 $M_{\oplus}$) of 47 UMa with
174: stable coplanar planetary configuration, based on the best-fit
175: orbital parameters given by Paper I. These new reliable orbital
176: solutions are derived from additional follow-up observations,
177: hence they can represent the actual motions of the system under
178: study. On the other hand, as mentioned previously, the discovery
179: of three close-in Neptune-mass planets demonstrates that it may be
180: possible for less massive planets ($\sim M_{\oplus}$) to move
181: close to the star. Therefore, in the extended study, we also
182: explored low-mass planets in the region 0.05 AU $\leq a < 0.4$ AU
183: and we found that the secular resonance arising from the inner
184: giant planet can render the eccentricity excitations for the
185: Earth-like planets (see \S3.2). In addition, we also carried out
186: two other runs to predict potential bodies in 47 UMa : In \S3.1,
187: to compare with the inner solar system, we further investigated
188: the planetary configuration of 2 giant planets plus 4 terrestrial
189: planets, and in \S3.3 we examined the case of presence of Trojan
190: planets with respect to the two giant planets. The results suggest
191: that the 47 UMa system may have a similar architecture to the
192: solar system.
193:
194: In the simulations, we use an N-body codes (Ji, Li \& Liu 2002) of
195: direct numerical simulations with the RKF7(8) and symplectic
196: integrators (Wisdom \& Holman 1991). We always take the stellar
197: mass and the minimum planetary masses from Table 1. The adopted
198: time stepsize is usually $\sim$ 1\%-2.5\% of the orbital period of
199: the innermost planet, which is sufficiently small for the
200: integration. Additionally, the numerical errors were effectively
201: controlled over the integration timescale, and the total energy is
202: generally conserved to $10^{-6}$ - $10^{-8}$ for the
203: integrations. Our main results now follow.
204:
205: \section{Simulations Results}
206: \subsection{2 Giants plus 4 terrestrial planets}
207: The announcement of the fourth planet in the 55 Cancri system
208: (McArthur et al. 2004) suggests that multiple planetary systems
209: resembling our solar system could be quite common in the galaxy.
210: In the first runs, we examine the configuration consisting of 2
211: giant planets (2G) and 4 terrestrial planets (4T) to study the
212: coexistence of multiple bodies. Specifically, this means that we
213: directly place Mercury, Venus, Earth and Mars into the 47 UMa
214: system to simulate "the inner solar system", where the orbital
215: elements for above terrestrial planets are calculated from JPL
216: planetary ephemerides DE405 at Epoch JD 2448750.9 corresponding to
217: the outer companion (see Table 2), e.g., the semi-major axes are
218: respectively, 0.387, 0.723, 1.00 and 1.523 AU. The giant planets
219: are always assumed to be coplanar in the simulations, thus the
220: inclinations for 4T refer to the fundamental plane of the 2G's
221: orbits. In Figure 1a, the orbital evolution of 4T is shown, where
222: $Q=a(1+e)$, $q=a(1-e)$ are, respectively, the apoapsis and
223: periapsis distances. The simulation results indicate that the
224: orbits of these low-mass planets become chaotic after $2\times
225: 10^{5}$ yr: Mars leaves its initial orbit in less than $10^{4}$ yr
226: and its $Q$ can amount to $\sim 10^{3}$ AU at $t>1.5\times 10^{5}$
227: yr. Such dynamical instability stems from strong perturbation by
228: 47 UMa b (see also \S3.2.4), and the Hill radius is
229: $R_{H}={\left[{M_{1}}/(3M_{c})\right]}^{1/3}a_{1}$ ($M_{c}$, the
230: stellar mass; $M_{1}, a_{1}$, the planetary mass and semi-major
231: axis of Companion B), then $3R_{H}\simeq 0.6$ AU, showing that the
232: orbit of Mars is quite close to the outskirts of $3R_{H}$ sphere.
233: In addition, as of its eccentricity grows, Venus begins to cross
234: the orbits of Earth and Mercury at $t>2.0\times 10^{4}$ yr, which
235: leads to eventual destruction of the system.
236:
237: However, as the two giant planets in 47 UMa are much closer to the
238: central star than the JS-pair to the Sun, for perfect analogy with
239: the inner solar system, the semi-major axes of 4T should be
240: shifted to 0.15, 0.29, 0.40 and 0.61 AU. Bearing this in mind, we
241: restarted a new run for the 2G-4T system, where we adopted the
242: rescaled semi-major axes for 4T together with all the other
243: initial values in Table 2. In this numerical experiment, we found
244: that the 2G-4T system can be dynamically stable and can last at
245: least for 20 Myr (see Figure 1b), where the time behaviors of $Q$
246: and $q$ for 4T show regular motions that their semi-major axes and
247: eccentricities do not dramatically change in their secular orbital
248: evolution. More terrestrial planets can be possibly created in the
249: later stage of planetary formation, Chambers (2001) studied the
250: accretion of $\sim150$ planetary embryos with lunar-to-Mars masses
251: in giant planets systems and recovered 4T similar to those in the
252: solar system. However, Levison \& Agnor (2003) underlined that the
253: population and masses of the resulting terrestrial planets may be
254: affected by the giant planets, because the growth and evolution of
255: the planetary embryos are determined by the possibly experienced
256: dynamical mechanisms (e.g., secular resonances).
257:
258: \subsection{Terrestrial planets in Habitable zones}
259: The Habitable Zones are generally conceived as places where the
260: biological evolution of life is able to develop on planetary
261: surfaces in environment of liquid-water, subtle temperature and
262: atmosphere components of CO$_{2}$, H$_{2}$O and N$_{2}$ (Kasting
263: et al. 1993); at the same time, the planetary habitability is also
264: related to the stellar luminosity and the age of the star-planet
265: system (Cuntz et al. 2003). The HZ could be considered to be
266: centered at $\sim$ 1 AU $(M_{c}/M_{\odot})^{2}$. For 47 UMa, the
267: inner and outer boundaries of HZ range from 0.7 AU to 1.3 AU
268: (Menou \& Tabachnik 2003), however, in our practical simulations,
269: we extended the HZ to other areas for the purpose of a more
270: comprehensive study.
271:
272: In the second series of runs, we extensively investigated the case
273: of two giant companions with one terrestrial planet in the HZ. The
274: mass of the assumed terrestrial planet ranges from 0.1
275: $M_{\oplus}$ to 10 $M_{\oplus}$. And the adopted initial orbital
276: parameters are as follows: numerical scanning was carried out for
277: [$a, e$] space by direct integrations, where the low-mass bodies
278: were placed at equal intervals of 0.01 AU for 0.05 AU $\leq a \leq
279: 2.0$ AU, the eccentricities were uniformly spaced every 0.01 for
280: $0.0 \leq e \leq 0.2$ (for 0.05 AU $\leq a < 0.4$ AU, where $0.0
281: \leq e \leq 0.1$), the inclinations are $0^{\circ} < I <
282: 5^{\circ}$, and the other angles were randomly distributed between
283: $0^{\circ}$ and $360^{\circ}$. Thus, over 3000 simulations were
284: exhaustively performed for typical integration spans from 1 Myr
285: to 10 Myr, totalling several $10^{10}$ yr.
286:
287: \subsubsection{0.05 AU $\leq a < 0.4$ AU}
288: In these runs, we explored the secular evolution of 385 "hot
289: Earths" or "hot Neptunes" for a time span of 1 Myr. All the
290: simulations are dynamically stable for $10^{6}$ yr, and 96\% of
291: the orbits bear $e_{final}<0.20$. However, Figure 2a shows that
292: the eccentricities for the bodies at $\sim 0.30$ AU are excited to
293: $\sim 0.40$, where the secular resonance $\nu_{1}$
294: ($41^{"}.11$/yr) of the inner companion (similar to $\nu_{5}$ for
295: Jupiter) is responsible for the excitation of eccentricity. The
296: debris disk at $\sim 0.30$ AU is also shown by Malhotra (2004),
297: who presented similar results of the eccentricity excitation of
298: massless bodies by nonlinear analytic theory for secular
299: resonance. In addition, even more terrestrial planets with spaced
300: orbits can simultaneously survive for longer lifetime (e.g., the
301: second simulation of 2G-4T system in \S3.1). Nevertheless, one can
302: easily see that this region is not a good location for
303: habitability, due to extra high temperature. Recently, Narayan,
304: Cumming, \& Lin (2005) discussed the detectability of low-mass
305: objects with orbital periods $P \leq 10$ days, and they estimated
306: the present threshold for detecting planets with close-in orbits
307: is $\geq 10-20 M_{\oplus}$, depending on the number of the
308: observations and the precision of Doppler measurements. Indeed,
309: the improvement of precision of the ground-based observations will
310: lead to the discovery of additional low-mass planets ($\sim 10
311: M_{\oplus}$) in other known planetary systems (G. W. Marcy 2004,
312: private communication).
313:
314: However, it brings about great difficulty in catching the
315: planetary formation for hot planets that move so close to their
316: host stars. Although the gas giant planets may undergo inward
317: orbital migration from several AU to the vicinity of host stars,
318: Ida \& Lin (2004) found that a large fraction (90\%-95\%) of the
319: planets that have migrated to $a <0.05$ AU must perish, because
320: the tidal heating can make the planetary radius inflate (Gu, Lin,
321: \& Bodenheimer 2003) and the body may be directly engulfed by the
322: star (Israelian et al. 2001) or survive only as rocky "cores" with
323: the gas envelope removed due to the mass loss of the planet
324: through the material exchange at the Roche radius. Nevertheless,
325: it is hopeful to detect such planets with $a \ge 0.05$ AU in
326: future space missions (e.g., COROT, KEPLER, TPF), while the
327: survival of the short-period terrestrial planets (Mardling \& Lin
328: 2004) is also related to the relativistic potential of the star.
329:
330: \subsubsection{0.4 AU $\leq a < 1.0$ AU}
331: We carried out 1260 integrations in this region for 5 Myr and we
332: found that none of the orbits escaped during this time span and
333: 94\% of them were in the resulting $e < 0.25$, see the final
334: status shown in Figure 2b. We find that the eccentricities of the
335: orbits with 0.70 AU $< a < $ 0.78 AU can be pumped up and in the
336: 2:9 MMR at $\sim 0.76$ AU, $e$ can reach $\sim 0.90$, indicating
337: that there may exist a gap near this resonance. Most of the
338: Earth-like planets about 1:4 MMR at $\sim 0.82$ AU move stably in
339: bounded motions with low-eccentricity trajectories, except for two
340: cases where the eccentricities eventually grow to high values.
341: Paper II pointed out that the secular resonance $\nu_{2}$ arising
342: from the outer companion (similar to $\nu_{6}$ for Saturn) can
343: remove the test bodies. Would the $\nu_{2}$ also influence the
344: Earth-like planets in this system? Nevertheless, we did not find
345: this mechanism at work at about $\sim 0.85$ AU (see Paper II) when
346: we examined the results, because the terrestrial planets under
347: study that all bear finite masses that may change the strength of
348: this resonance; on the other hand, the location of the secular
349: resonance is changed due to the orbital variation of the outer
350: companion. For a terrestrial planet with a mass of $10
351: M_{\oplus}$, the region for $\nu_{2}$ secular resonance is now
352: shifted to $\sim 0.70$ AU, where two eigenfrequencies for the
353: terrestrial body and outer giant planet given by the
354: Laplace-Lagrange secular theory are, respectively, $211^{"}.37$/yr
355: and $225^{"}.48$/yr. This indicates that both planets almost have
356: the same secular apsidal precession rates in their motion. At the
357: new location, the $\nu_{2}$ resonance, together with the mean
358: motion resonance, can work at clearing up the planetesimals in the
359: disk (see Fig. 2b) by the excitation of the eccentricity;
360: qualitatively, our results are in accord with those of Paper II.
361:
362: While the inner edge of HZ marks out a narrow unstable area, it is
363: very possible to discover Earth-like planets in this wider area,
364: 0.4 AU $< a < 1.0$ AU, in future surveys, which are to be the best
365: candidate of habitable places for biological evolution of
366: intelligent beings.
367:
368: \subsubsection{1.0 AU $\leq a < 1.3$ AU}
369: There were 630 simulations in this region for 10 Myr and we found
370: that 88\% survived the integration, confirming the results given
371: in Paper II that most of the test particles with $a < 1.3$ AU can
372: eventually remain in the system. Here, for the 3:1 resonance at
373: $\sim 1.0$ AU, Figure 3a shows that there are stable orbits in the
374: zones about 1 AU with $e\leq0.1$, which agrees with the work by
375: Rivera \& Haghighipour (2004) who showed that a test particle can
376: last 100 Myr at 1 AU in 47 UMa; while for $0.1<e\leq 0.2$, the
377: orbits tend to be in unstable state owing to the excitation of the
378: eccentricities. The simulations may imply that the Earth-mass
379: planets near 3:1 resonance are possibly on the edge of stability.
380: However, the previous studies on this system showed that the 3:1
381: resonance is a gap with no survivors. Let us mention that such
382: differences may arise from the adopted initial planetary
383: configurations, and here we adopt the reliable best-fit orbital
384: solutions given in Paper I that can describe the exact motions for
385: the two giant planets. Hence, a comparative run was carried out to
386: examine this, again we ran 630 simulations for 10 Myr but with the
387: earlier orbital elements for the two massive planets to reproduce
388: the previous results at the 3:1 resonance. Our results with the
389: earlier data show that most of the Earth-mass planets about 3:1
390: resonance are unstable for the investigated time, and their
391: eccentricities can be pumped up to $\sim 1$ through resonance;
392: besides, the inclinations are excited to high values ranging
393: $90^{\circ}$ to $180^{\circ}$, indicating that the orbits of the
394: Earth-mass planets become retrograde in the dynamical evolution
395: and cross those of the prograde giant planets before they
396: terminate their dynamical lifetimes. Thus, we may safely conclude
397: that the stability of the terrestrial planets is dependent on the
398: initial planetary configuration.
399:
400: In Fig.3a, a narrow unstable stripe appears at the 5:2 MMR at
401: $\sim 1.13$ AU, although several of them can be luckily left
402: behind, most of the Earth-size planets are removed at $\sim$ 1 Myr
403: due to the perturbation of 47 UMa b, and in this sense it is
404: analogous to the situation in the solar system. However, a wider
405: area between 3:1 and 5:2 MMR is assumed to be a qualified
406: candidate habitable environment where the Earth-mass planet will
407: not encounter the problem of dynamical stability, and this is
408: almost true for the region (1.13 AU, 1.30 AU) with $e\leq0.1$,
409: except several unstable islands near 7:3 MMR at 1.18 AU. The
410: smaller eccentricity (near-circular orbits) may not cause dramatic
411: variations of temperature on the planet's surface, so favoring
412: habitability. Therefore, in a dynamical sense, if the 47 UMa
413: system can be adopted as a candidate target for SIM, it is also
414: possible to detect other Earths with stable orbits about 1 AU.
415:
416: \subsubsection{1.3 AU $\leq a \leq 1.6$ AU}
417: We performed 651 simulations for 10 Myr, and found the dynamical
418: structure in this regime to be quite complicated: 14\% of them can
419: finally survive for this time span, and 86\% are lost by ejection
420: into hyperbolic trajectories, indicating the chaotic nature for
421: these bodies. In Figure 3b, we can notice that the 2:1 MMR region
422: is at $\sim 1.31$ AU and also close to the outer edge of HZ, and
423: the orbits with $0.0 < e< 0.10$ are unstable, while for $0.10 \leq
424: e\leq 0.20$, there are stable islands where fictitious planets can
425: remain in bounded motions in the final system. We observe that the
426: 2:1 resonance marks out a remarkable boundary between chaotic and
427: regular orbits, indicating that orbits with $a < 1.31$ AU can have
428: much larger surviving rates than those of $a > 1.31$ AU. However,
429: there are wider stable region about 9:5 MMR at $\sim 1.40$ AU for
430: low eccentricities $0.0<e\leq0.05$. Most of the unstable orbits
431: are in the region $1.43$ AU $<a <1.56$ AU, using resonance
432: overlapping criterion (Murray \& Dermott 1999), the separation in
433: semi-major axis
434: $\Delta{a}\approx1.3{\left({M_{1}}/M_{c}\right)}^{2/7}a_{1}\doteq0.496$
435: AU, thus, the inner boundary $R_{O}$ for 47 UMa b is at $\sim
436: 1.58$ AU, and the orbits in this zone become chaotic during the
437: orbital evolution because the planets are both within $3R_{H}$ and
438: also close to $R_{O}$. And the characterized ejecting time $\tau
439: \sim 1$ Myr, which means the apparent gap (e.g., 5:3 MMR at $\sim
440: 1.48$ AU) in the inner belt, except for several stable islands.
441: Another possible population for terrestrial planets is located at
442: 3:2 resonance $\sim 1.59$ AU for $0.04< e < 0.20$, and 18 small
443: bodies can last for 10 Myr and confirm the results of Paper II.
444: The 3:2 MMR zone in 47 UMa is reminiscent of the Hilda asteroids
445: in the solar system moving in a stable region.
446:
447: \subsubsection{1.6 AU $< a \leq 2.0$ AU}
448: 840 Earth-like planets are placed in this region for integration
449: for 10 Myr, and the simulation revealed that 98\% are removed from
450: the system within a typical ejection time $\tau<3\times 10^{4}$
451: yr, which is much shorter than for 1.3 AU $\leq a \leq $1.6 AU,
452: implying a thoroughly chaotic situation in this area. It is not
453: difficult to understand that these terrestrial planets are
454: entirely thrown out by the gravitational influence of the inner
455: giant planet, using the $3R_{H}$ stability criterion.
456:
457: \subsection{Trojan planets}
458: In the solar system, at present there are 1690 Jupiter
459: Trojans\footnote{http://cfa-www.harvard.edu/iau/lists/JupiterTrojans.html}
460: (1062 preceding and 628 trailing Jupiter) located at L4 and L5.
461: Hence, a fascinating issue is that whether there exist "extrasolar
462: Trojan planets" in other planetary systems, and if such planets
463: can occur in the system harboring only one gas giant with $a \sim
464: 1 $ AU (Dvorak et al. 2004), it may essentially help understand
465: their presence in the HZ. Laughlin \& Chambers (2002) found the
466: equilateral configuration consisting of a star and two equal-mass
467: planets can be linearly stable for mass ratios less than 0.03812
468: and showed that a pair of Saturn-mass planets can sustain the
469: stability in the 1:1 orbital resonance.
470:
471: In the third series of runs, we let the experimental planets with
472: masses of 0.1 $M_{\oplus}$ to 10 $M_{\oplus}$ initially orbit
473: about the triangular equilibrium points of the two giant
474: companions, for each giant planet there are 100 orbits distributed
475: near L4 (or L5) point. Then, we implemented 400 dynamical
476: simulations to explore their secular evolution on the timescale of
477: $10^{7}$ yr. The starting conditions for the Trojan bodies are:
478: $0.00 < e < 0.10$, $0^{\circ} < I < 1^{\circ}$ and $|\lambda -
479: \lambda_{1,2} - 60^{\circ}|<1^{\circ}$ (where $\lambda$, and
480: $\lambda_{1,2}$, respectively, the mean longitudes of the Trojan,
481: the inner and outer companions). For the inner planet of 47 UMa b,
482: the survival rates for the low-mass bodies about L4 and L5 points
483: are, respectively, 96\% and 98\%, which is quite similar to the
484: Trojans and Greeks for Jupiter; while for the outer planet (47 UMa
485: c), the survival proportion goes down to 80\% and 70\% near L4 and
486: L5 region, as a comparison Saturn has no co-orbital bodies.
487: Obviously, the star-planet mass ratios $\mu \sim 10^{-3} < 0.0385$
488: (Murray \& Dermott 1999) both satisfy the linear stability
489: condition for the triangular points. In addition, it is also
490: suggested that the stable triangular region for the inner planet
491: is wider than that of the outer planet, and this can be easily
492: understood as this width depends on the value $\sim \mu^{1/2}$ of
493: the tadpole region.
494:
495: Figure 4 displays the long-term orbital evolution for the low-mass
496: planets moving about L4 (\textit{black line}) and L5
497: (\textit{yellow line}) points for 47 UMa, where $a$ and $e$ both
498: perform small modulations for $10^{7}$ yr, and $\lambda -
499: \lambda_{1,2}$ also librate about $60^{\circ}$ and $300^{\circ}$,
500: respectively, with low amplitudes for the same time span.
501:
502:
503: \section{Summary and discussions}
504: In this work, we have systematically studied the existence of
505: Earth-like planets in the region for 0.05 AU $\leq a \leq 2.0$ AU
506: for 47 UMa by numerical simulations. In addition, we also
507: investigated the packed system and Trojan planets in this system.
508: We now summarize the main results as follows:
509:
510: (1) The "hot Earths" can dynamically survive for $10^{6}$ yr for
511: 0.05 AU $\leq a<$ 0.4 AU, and they are probably detected by
512: transit time variations (Agol et al. 2005; Holman \& Murray 2005)
513: via their interaction with the transiting planet\footnote{Laughlin
514: G. and his collaborators are now maintaining the website (see
515: http://www.transitsearch.org/) to search for transiting "hot
516: Jupiter" planets through the worldwide cooperative observational
517: campaign. Besides, the other active group from OGLE project have
518: detected several transiting planets
519: (http://sirius.astrouw.edu.pl/$\sim$ogle/) in the observations.}.
520: These techniques may become a vital and effective observational
521: strategy to discover more transiting objects around main sequence
522: stars.
523:
524: (2) The Trojan planets with low eccentricities and inclinations
525: can last out at the triangular equilibrium points of the two
526: massive planets of 47 UMa. In this sense, it is analogous to the
527: case in solar system. Nevertheless, if they really exist, the
528: formation of these bodies is still a mystery and needs further
529: study.
530:
531: (3) The Earth-like planets can eventually remain in the system for
532: 10 Myr in the areas associated with mean motion resonance (e.g.,
533: 3:2 MMR) with the inner companion. We also showed that the 2:1 and
534: 3:1 resonances could be on the fringe of stability, but the 5:2
535: MMR is unstable and the bodies can be ejected as "extrasolar
536: comets". And this may sketch out an asteroidal belt structure
537: similar to the solar system. Moreover, the 2:1 MMR (near the outer
538: boundary of HZ at 1.30 AU) marks out a significant barrier between
539: chaotic and regular motions, implying that a large fraction of the
540: orbits inside this resonance can be survive, while most of them
541: are lost in the simulations outside the 2:1 resonance. Again,
542: considering the inner boundary of HZ and dynamical stability, we
543: point out that the most likely candidate for habitable environment
544: is terrestrial planets with orbits in the ranges 0.8 AU $\leq a <
545: 1.0$ AU and 1.0 AU $ < a < 1.30$ AU (except 5:2 MMR, and several
546: unstable cases) with low eccentricities (e.g., $0.0 \leq e \leq
547: 0.1$). However, in our own solar system there are no terrestrial
548: planets from the 1:4 MMR out to Jupiter, although there are stable
549: orbits there. This may suggest that although some orbits are
550: stable, conditions are such that terrestrial planets cannot form
551: so close to giant planets. Perhaps this is because runaway growth
552: is suppressed due to the increased eccentricities from the
553: perturbations of the giant planet. In 47 UMa, the corresponding
554: region runs from 0.82 AU on out (see Figure 2), almost completely
555: covering the HZ. Hence, it would be reasonable to conclude that
556: the only proper place to find habitable planets in this system
557: would be at about 0.8 AU. But this should be carefully examined by
558: forthcoming space measurements (e.g., SIM) capable of detecting
559: low-mass planets.
560:
561: In a word, we can see that the 47 UMa planetary system may bear a
562: similar dynamical structure to our solar system, in that it can
563: also own several terrestrial members resembling the inner solar
564: system. A comparative study can be performed in other planetary
565: systems with two giant planets ({\' E}rdi et al. 2004; Raymond \&
566: Barnes 2005) to explore whether Earth-like planets can exist
567: there, and to locate less massive planets (Malhotra 2005, in
568: preparation) in general planetary systems. The dynamical
569: habitability is relevant to both mean motion resonance and secular
570: resonance for a given system. Moreover, the formation of the
571: terrestrial planets in the HZ is difficult in the planetary
572: systems that contain a giant planet with a moderate-size eccentric
573: orbit (Veras \& Armitage 2005), because most of the initial
574: material within HZ was depleted by orbital crossings; in addition,
575: the accretion simulations (Paper II) show that the massive
576: planetary embryos in the HZ should be created prior to the
577: formation of giant gaseous planets. Hence, more efforts should
578: also be directed to the understanding of the formation of
579: habitable terrestrial planets.
580:
581: %%After we submitted this manuscript, we know that there is an
582: %%extensive work by ...
583:
584: \acknowledgments{We thank the anonymous referee for useful
585: comments and suggestions that helped to improve the contents. We
586: thank Lin D.N.C., Marcy G.W. and Malhotra R. for insightful
587: discussions. We gratefully acknowledge the language improvement by
588: Kiang T. We are also grateful to Zhou Q.L. for the assistance of
589: workstation utilization. Most of the computations were carried out
590: on high performance workstations at Laboratory of Astronomical
591: Data Analysis and Computational Physics of Nanjing University.
592: This work is financially supported by the National Natural Science
593: Foundations of China (Grant 10203005, 10173006, 10233020) and the
594: Foundation of Minor Planets of Purple Mountain Observatory.}
595:
596:
597: \clearpage
598:
599: \begin{thebibliography}{}
600:
601: \bibitem[Adams (2003)]{Ada03}
602: Adams, F., \& Laughlin, G. 2003, Icarus, 163, 290
603:
604: \bibitem[Agol (2005)]{Ago05}
605: Agol, E., et al. 2005, \mnras, 359, 567
606:
607: \bibitem[Asghari & (2004)]{Asg04}
608: Asghari, N., et al. 2004, \aap, 426, 353
609:
610: \bibitem[Barnes (2004)]{Bar04}
611: Barnes, R., \& Quinn, T. 2004, \apj, 611, 494
612:
613: \bibitem[Butler et al. (1996)]{But96}
614: Butler, R. P., \& Marcy, G. W. 1996, \apj, 464, L153
615:
616: \bibitem[Butler et al. (2004)]{But04}
617: Butler, R. P., et al. 2004, \apj, 617, 580
618:
619: \bibitem[Chambers (2001)]{Cha01}
620: Chambers, J.E. 2001. Icarus, 152, 205
621:
622: \bibitem[Cuntz et al.(2003)]{Cun03}
623: Cuntz, M., von Bloh, W., Bounama, C., \& Franck, S.\ 2003, Icarus,
624: 162, 214
625:
626: \bibitem[Dvorak et al. (2004)]{Dvo04}
627: Dvorak, R., et al. 2004, \aap, 426, L37
628:
629: \bibitem[{\' E}rdi et al.(2004)]{Erdi2004}
630: {\' E}rdi, B., Dvorak, R., S{\' a}ndor, Z., Pilat-Lohinger, E., \&
631: Funk, B.\ 2004, \mnras, 351, 1043
632:
633:
634: \bibitem[Fischer et al.(2002)]{Fis02}
635: Fischer, D., et al. 2002, \apj, 564, 1028
636:
637: \bibitem[Fischer et al.(2003)]{Fis03a}
638: Fischer, D., et al. 2003, \apj, 586, 1394 (Paper I)
639:
640: \bibitem[Gozdziewski (2002)]{Goz02}
641: Gozdziewski, K. 2002, \aap, 393, 997
642:
643: \bibitem[Goldreich(2004)]{Gol04}
644: Goldreich, P., Lithwick, Y., \& Sari, R. 2004, \apj, 614, 497
645:
646: \bibitem[Gu(2004)]{Gu04}
647: Gu, P., Lin, D. N. C., \& Bodenheimer, P. H. 2003, \apj, 588, 509
648:
649: \bibitem[Holman(2005)]{Hol05}
650: Holman, M., \& Murray, N. 2005, Science, 307, 1288
651:
652: \bibitem[Ida et al.(2004)]{Ida04}
653: Ida, S., \& Lin, D.N.C. 2004, \apj, 604, 388
654:
655: \bibitem[Israelian et al.(2001)]{Isr01}
656: Israelian, G., et al. 2001, \nat, 411, 163
657:
658: \bibitem[Ji et al.(2002)]{Ji02}
659: Ji, J., Li, G., \& Liu, L. 2002, \apj, 572, 1041
660:
661: \bibitem[Ji et al.(2003)]{Ji03}
662: Ji, J., et al. 2003, \apj, 591, L57
663:
664: \bibitem[Jones et al.(2001)]{Jon01}
665: Jones, B. W., Sleep, P.N., \& Chambers, J.E. 2001, \aap, 366, 254
666:
667: %%\bibitem[Jones et al.(2005)]{Jon05}
668: %%Jones, B.~W., Underwood, D.~R., \& Sleep, P.~N.\ 2005, \apj, 622,
669: %%1091
670:
671: \bibitem[Kasting et~al.(1993)]{Kas93}
672: Kasting, J. F., et al. 1993, Icarus, 101, 108
673:
674: \bibitem[Kley (2000)]{Kle00}
675: Kley, W. 2000, \mnras, 313, L47
676:
677: \bibitem[Kley et~al.(2004)]{Kle2004}
678: Kley, W., Peitz, J., \& Bryden, G. 2004, \aap, 414, 735
679:
680: \bibitem[Kokubo (1998)]{Kok98}
681: Kokubo, E., \& Ida, S. 1998. Icarus, 131, 171
682:
683: \bibitem[Laughlin et al.(2002)]{Lau02}
684: Laughlin, G., Chambers, J. \& Fischer D. 2002, \apj, 579, 455
685: (Paper II)
686:
687: \bibitem[Laughlin (2002)]{Lau2002}
688: Laughlin, G., \& Chambers, J. 2002, \aj, 124, 592
689:
690: \bibitem[Lee et al.(2002)]{Lee02}
691: Lee, M.H., \& Peale, S.J. 2002, \apj, 567, 596
692:
693: \bibitem[Levison et~al.(2003)]{Lev03}
694: Levison, H.F., \& Agnor, C. 2003. \aj, 152, 2692
695:
696: \bibitem[Lin et~al.(2004)]{Lin04}
697: Lin, D.N.C. 2004, "Dynamical Shake-up of Planetary Systems", KIPT
698: conference on Planet Formation, University of California, Santa
699: Barbara, CA
700:
701: \bibitem[Lissauer (1993)]{Lis93}
702: Lissauer, J.J. 1993, \araa, 31, 129
703:
704: \bibitem[Malhotra et al.(2000)]{Mal00}
705: Malhotra, R., Duncan, M. J., \& Levison, H. 2000, in Protostars
706: and Planets IV, ed. V. Mannings, A. P. Boss \& S. S. Russell
707: (Tucson: Univ. Arizona Press), 1231
708:
709: \bibitem[Malhotra (2004)]{Mal04}
710: Malhotra, R. 2004, "Locating sub-jovian planets and debris in
711: exo-planetary systems", American Astronomical Society, DPS meeting
712: \#36, \#42.04
713:
714: \bibitem[Mardling et al.(2004)]{Mard04}
715: Mardling, R. \& Lin, D.N.C. 2004, \apj, 614, 955
716:
717: \bibitem[McArthur et al. (2004)]{McA04}
718: McArthur, B.E., et al. 2004, \apj, 614, L81
719:
720: \bibitem[Menou & Tabachnik(2003)]{Men03}
721: Menou, K., \& Tabachnik, S. 2003, ApJ, 583, 473
722:
723: \bibitem[Murray & Dermott(1999)]{Mur99}
724: Murray, C. D., \& Dermott, S. F. 1999, Solar System Dynamics (New
725: York: Cambridge Univ. Press).
726:
727: \bibitem[Nagasawa \& Ida(2000)]{Nag00}
728: Nagasawa, M., \& Ida, S. 2000, \aj, 120, 3311
729:
730:
731: \bibitem[Narayan(2005)]{Nar05}
732: Narayan, R., Cumming, A., \& Lin, D.N.C. 2005, \apj, 620, 1002
733:
734: \bibitem[Nelson et al.(2000)]{Nel00}
735: Nelson, R. P., et al. 2000, \mnras, 318, 18
736:
737: \bibitem[Papaloizou (2003)]{Pap2003}
738: Papaloizou, J.C.B. 2003, Cel. Mech. \& Dyn. Astron., 87, 53
739:
740: \bibitem[Raymond \& Barnes(2005)]{Ray05}
741: Raymond, S.~N., \& Barnes, R.\ 2005, \apj, 619, 549
742:
743: \bibitem[Rivera (2004)]{Riv2004}
744: Rivera, E.J., \& Haghighipour, N. 2004, submitted to ApJ
745: (astro-ph/0406429)
746:
747: \bibitem[Santos et al. (2004)]{San04}
748: Santos, N.C., et al. 2004, A\&A, 426, L19
749:
750: \bibitem[Safronov (1969)]{Saf69}
751: Safronov,V.S. 1969, Evolution of the Protoplanetary Cloud and
752: Formation of the Earth and the Planets (Moscow:Nauka)
753:
754: \bibitem[Veras \& Armitage(2005)]{Ver2005}
755: Veras, D., \& Armitage, P.~J.\ 2005, \apjl, 620, L111
756:
757: \bibitem[Ward (1997)]{War97}
758: Ward, W. R. 1997, Icarus, 126, 261
759:
760: \bibitem[Wisdom & Holman(1991)]{Wis91}
761: Wisdom, J., \& Holman, M. 1991, \aj, 102, 1528
762:
763:
764: \end{thebibliography}
765:
766: \clearpage
767: \begin{deluxetable}{lll}
768: \tablewidth{0pt} \tablecaption{The orbital parameters of 47 UMa
769: planetary system (adopted from Fischer et al. [2003]). The stellar
770: mass is $1.03M_{\odot}$.} \tablehead{ \colhead{Parameter} &
771: \colhead{Companion B} & \colhead{Companion C} } \startdata
772: $M$sin$i$($M_{Jup}$) & 2.86 & 1.09 \\
773: Orbital period $P$(days) & 1079.2 & 2845.0 \\
774: $a$(AU) & 2.077 & 3.968 \\
775: Eccentricity $e$ & 0.05 & 0.00 \\
776: $\omega$(deg) & 124.3 & 170.89 \\
777: Periastron Time (JD) & 2452374.1 & 2448750.9
778: \enddata
779: \end{deluxetable}
780:
781: %%\clearpage
782: \begin{deluxetable}{lllllll}
783: \tablewidth{0pt} \tablecaption{The orbital elements for 4
784: terrestrial planets at JD 2448750.9 (From DE405).}
785: \tablehead{\colhead{Planet} &\colhead{a(AU)} &\colhead{ecc}
786: &\colhead{$I$} &\colhead{$\Omega$} &\colhead{$\omega$}
787: &\colhead{$M$}} \startdata
788: Mercury &.387 &.2056 &7.00 &48.34 &29.10 &260.40 \\
789: Venus &.723 &.0067 &3.39 &76.70 &55.06 &253.70 \\
790: Earth &1.000 &.0163 &0.001 &255.71 &207.06 &123.84 \\
791: Mars &1.523 &.0935 &1.850 &49.57 &286.48 &355.18
792: \enddata
793: \end{deluxetable}
794: \clearpage
795:
796:
797: \epsscale{1.10}
798: \begin{figure}
799: \figurenum{1} \plottwo{f1a.eps}{f1b.eps} \caption{The orbital
800: evolution of 4 terrestrial planets. \textit{Left}: chaotic case
801: with $a$ from Table 2, Mars leaves its initial orbit at $t
802: <10^{4}$ yr, and Venus begins to cross the orbits of Earth and
803: Mercury at $t>2.0\times 10^{4}$ yr. \textit{Right}: with rescaled
804: semi-major axes, 4T remain regular motions that $a$ and $e$ both
805: perform slight vibrations for 20 Myr. \label{fig1} }
806: \end{figure}
807: \clearpage
808:
809: \epsscale{1.10}
810: \begin{figure}
811: \figurenum{2} \plottwo{f2a.eps}{f2b.eps} \caption{The contour of
812: status of the final eccentricities for Earth-like planets, the
813: vertical axis for the initial $e$. \textit{Left:} 0.05 AU $\leq a
814: < 0.4$ AU for 1 Myr. Notice $\nu_{1}$ secular resonance at $\sim
815: 0.30$ AU pumps up the eccentricities. \textit{Right:} 0.4 AU $\leq
816: a < 1.0$ AU for 5 Myr. The $e$ of the orbits with 0.70 AU $< a <
817: 0.78$ AU can be excited and in the 2:9 MMR at $\sim 0.76$ AU, $e$
818: can reach $\sim 0.90$. \label{fig2}}
819: \end{figure}
820:
821: \clearpage
822:
823: \epsscale{1.10}
824: \begin{figure}
825: \figurenum{3} \plottwo{f3a.eps}{f3b.eps} \caption{The surviving
826: time for Earth-like planets for the integration of 10 Myr, the
827: vertical axis is the same as Fig.2. \textit{Left}: for 1.0 AU
828: $\leq a < 1.3$ AU, see the gap for the 5:2 MMR at $\sim 1.13$ AU.
829: \textit{Right}: for 1.3 AU $\leq a \leq 1.6$ AU, a population of
830: the terrestrial planets is about 9:5 MMR at $\sim 1.40$ AU for
831: low eccentricities (see texts for details). \label{fig3} }
832: \end{figure}
833:
834: \clearpage
835:
836: \epsscale{1.10}
837: \begin{figure}
838: \figurenum{4} \plottwo{f4a.eps}{f4b.eps} \caption{The secular
839: evolution for the low-mass planets moving about L4 (\textit{black
840: line}) and L5 (\textit{yellow line}) points for 47 UMa, where $a$
841: and $e$ both perform small modulations for $10^{7}$ yr, and
842: $\lambda - \lambda_{1,2}$ also librate about $60^{\circ}$ and
843: $300^{\circ}$ with low amplitudes, respectively. \textit{Left}:
844: for inner planet. \textit{Right}:for outer planet.\label{fig4}}
845: \end{figure}
846:
847:
848:
849:
850: \end{document}
851: %%
852: %% End of file 'ms.tex'.
853: