Документ взят из кэша поисковой машины. Адрес оригинального документа : http://comet.sai.msu.ru/~gmr/p208.pdf Дата изменения: Mon Feb 13 12:20:57 2006 Дата индексирования: Sat Apr 9 22:43:03 2016 Кодировка: Поисковые слова: южная атлантическая аномалия

2ULJLQDQG3URSDJDWLRQRI6KRFNV LQWKH$WPRVSKHUHVRI0LUD7\SH6WDUV
*HRUJLM05XGQLWVNLM

3

6WHUQEHUJ$VWURQRPLFDO,QVWLWXWH0RVFRZ6WDWH8QLYHUVLW\ 8QLYHUVLWHWVNLMSURVSHNW0RVFRZ5XVVLDHPDLOJPU#VDLPVXUX
5HVXOWVRIORQJ WHUPVWXGLHV RI0LUDW\SHYDULDEOHVDUHDQDO\VHG'DWDRQWKHPDVHUHPLVVLRQRI+ 2PROHFXOHV FP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
GLVSOD\VEULHIHSLVRGHVODVWLQJQRWPRUHWKDQDFRXSOHRIPRQWKVRI+ HPLVVLRQ 6XFKHSLVRGHVDSSHDURQO\LQRQHRIDIHZYLVXDOOLJKWF\FOHVLQRWKHUF\FOHVWKH%DOPHUHPLVVLRQVDUHXQGHWHFWDEOH7KHVHVWDUV PD\EHLQWULQVLFDOO\ZHDNHUSXOVDWRUVZLWKORZHUPDVV ORVVUDWHVDFFRUGLQJO\WKLQQHUFLUFXPVWHOODUHQYHORSHVDQG IDLQWHU XQVWDEOHPDVHUV2QFHSHUVHYHUDOSHULRGVCVXSHUSHULRG
DVWURQJHUVKRFNGHSDUWVIURPWKHVWHOODUVXUIDFH7KHYLVXDOOLJKW FXUYHVVXJJHVWWKDWWKLVLVSUREDEO\FRQQHFWHGZLWKEULJKWHUPD[LPDRIWKHVWDUV$QRWKHUSRVVLELOLW\IRUWKHCWUDQVLHQ WV
LV FRQQHFWHGZLWKUHOLFSODQHWDU\V\VWHPVDURXQGUHGJLDQWV0RYLQJDWDVXSHUVRQLFYHORFLW\ZLWKLQWKHDWPRVSKHUHDSODQHW SURGXFHVDFRQLFVKRFNZDYHYLVLEOHDKRWVSRW7KLVVSRWLVDVRXUFHRI%DOPHUHPLVVLRQOLQHVLWPD\DOVRDFFRXQWIRUFKDQJL QJ DV\PPHWULHVLQWKHLQIUDUHGEULJKWQHVVGLVWULEXWLRQVREVHUYHGLQWHUIHURPHWULFDOO\LQVRPH0LUDV,IWKHFRPSDQLRQRUELWKDVD VXIILFLHQWHFFHQWULFLW\DQGDSHULRGRIVHYHUDO\HDUVWKHQWKHHSLVRGHRIVWURQJ+ DQGIROORZLQJ+ 2PDVHUIODUHPD\EHGX HWR WKHSHULDVWURQSDVVDJHRIWKHFRPSDQLRQ
:RUNVKRSRQ0DVV/RVLQJ3XOVDWLQJ6WDUVDQG7KHLU&LUFXPVWHOODU0DWWHU6HQGDL-DSDQ0D\±


1

Observations

Since 1980, our team has been monitoring a sample of about 20 Miras in the H2 O maser line at = 1.35 cm (Berulis et al. 1983). Beginning from 1994 we have also been doing optical spectroscopy of these stars (Esipov et al. 1999 and references therein). In this poster I give special emphasis for the Balmer-line emission and H2O circumstellar masers, for which we have traced a long-term variability. Two Groups of Masers · `Stable' masers, which have been displaying H2 O maser emission (however, with varying intensity) throughout our observational interval, never falling below our detection threshold ( 10 Jy). Examples: R Aql, U Her, RS Vir. For these stars the H2 O maser variations correlate with the visual light curves, following them with a certain delay of about 0.3­0.4P (P is the brightness variation period). For U Ori this delay, analysed on a time interval of 12P , probably varies itself with a `supeperiod' of 9P (Rudnitskij et al. 2000). A cross-correlation analysis between the visual light curve and H2O maser variations for the Mira RS Vir (Lekht et al. 2001) shows that actually the correlation is maximum for a delay between the visual and H2O light curves of 4­5P . These stars also frequently show Balmer emission lines in their spectra, nearly in every light cycle. · `Transient' masers, which sometimes disappeared from our view, falling below 10 Jy for certain intervals. Some of them (R Leo, R Cas, U Aur) remained `H2O-silent' for more than 15 years, but then episodically reappeared. The H emission was seldom observed in these stars, usually in the form of short epsiodes, lasting a couple of weeks, near selected light minima (Esipov et al. 1999). The H2 O masers from both groups flare from time to time. In particular, U Ori, which to some extent may be classified as a `transient', because in early 1980 (and prior to this, during several years, as can be found in the literature) it was `H2O-silent'), but between June and October 1980 bursted to more than 1000 Jy (Rudnitskij et al. 2000). A similar event happened in the H2 O maser associated with the miralike semiregular variable W Hya in 1981 (Berulis et al. 1983) and in 2001.

1


2

Sho cks as a Common Cause of Variations of the Optical Emission Lines and H2O Masers

The correlation between the visual light curve and the H2 O maser variations can be explained by the model of shock impact on the masering region (Rudnitskij & Chuprikov 1990, Fig. 1). Shock waves, departing from the stellar surface under the action of pulsations, reach some time after the layers of the circumstellar envelope containing the H 2 O molecules. At this, collisional maser pumping is enhanced, and maser line flux increases. The visual­H2O correlation function for RS Vir suggests that this delay may be as long as a few stellar perio ds (Lekht et al. 2001), corresponding to a shock velocity of 10 km s-1. In the star R Leo (Esipov et al. 1999, see Fig. 2) -- and, to a smaller scale, in R Cas and U Aur -- we have observed short events of an H line flare, followed a year and a half later by a flare in the H2O radio line. We have assumed a shock as a common cause of these two events. If such an event is rare and takes place once per 10­15 stellar light cycles, there can be various explanations. The main problem may be that shocks in Miras are not so strong as it has been believed. Basing on the H emission line profiles in o Cet, Gillet et al. (1983) inferred sho ck velocities vs of up to 90 km s-1. However, other data of optical spectroscopy (Fox, Wood, & Dopita 1984) and weakness or lack of microwave continuum, which should accompany the H emission (Reid & Menten 1997, Chapman & Rudnitskij 2002) suggest vs values not higher than 20-25 km s-1. This is difficult to reconcile with the intense Balmer emission. · Merging shocks. Wood (1979) made calculations of consecutive shocks leaving the mira's surface and showed that shocks that leave the stellar surface later propagate in the gas `prepared' by previous shocks. The `later' shocks can overtake the `earlier' ones, merge with them and produce a stringer effect, both on Balmer emission and H2O maser. · Quasi-Stationary Layer. Rudnitskij & Chuprikov (1990) have explained the `superperiod' by building and disruption of a quasi-stationary circumstellar layer hosting the masering H2O molecules. This layer may appear in a stronger mass-loss episode once per several stellar cycles. Another cause may be that pulsation-driven shocks, consecutively departing from the stellar surface, can overtake their predecessors and merge with them, thus producing, again once per several cycles, a stronger-than-average shock (Wood 1979), also resulting in a maser flare.
2


3

A Low-Mass Companion as a Source of Sho cks

As a basis of a discussion, I propose another explanation for the lack of radio continuum and for the rare H­H2O episodes in the `transient' H2O source, namely, impact of a shock provoked by a low-mass companion (a planet?) to the Mira star. If a star possessed a planetary system during its main-sequence life, at the redgiant stage the closer-by planets, revolving at R 1 - 3 A.U., will be embedded within the star's atmosphere. The more massive ones of them will probably survive the red-giant phase (Struck-Marcell 1988, Rybicki & Denis 2001). Evolution of a red giant having a compact stellar companion (a brown/black dwarf with a mass of 0.02M or possibly a neutron star) was analysed in a series of papers on the `double-core evolution' (Soker 1999 and references therein). The fate of a lower-mass companion [(0.001­0.01) M ­ a planet], embedded in the atmosphere of a star that has become a red giant, was also considered in a number of works. In particular, Soker (1999) proposed to search for Uranus­Neptune-like planets in planetary nebulae, formed in course of the post-AGB evolution. Until recently considerations about exosolar planetary systems might be purely speculative. However, systematic observations of the last decade have led to discoveries of several tens of planets orbiting solar-type stars in the solar vicinity. Most of them have been detected by the Doppler technique, involving precise measurements of stellar radial velocities on a time interval of several years, aiming to find small velocity shifts, caused by the orbital motion of a low-mass companion. Depending on the planet's mass and semima jor axis, amplitudes of such velo city variations are from several meters per second to several scores of m s-1. A complete list of these detections is maintained by Schneider (2002) at the Paris­ Meudon Observatory. The Doppler technique, used on a limited time interval, selects in the first turn more massive planets in orbits closer to the central stars. The planets detected so far can be divided into two groups: · `hot Jupiters', planets of one to a few Jupiter masses, revolving in circular very closely to their stars, at 0.05­0.5 A.U.; · planets in eccentric orbits (with e up to 0.6) with revolution periods up to a few hundred days. Certainly, further observations will perhaps planetary systems similar to A planet orbiting around a 1M velocity Vp 30 km s-1. If the star find the star is a
3

lower-mass and longer-period planets, Solar System. at a distance of 1 A.U. would move at red giant, then the planet is embedded


in the star's atmosphere, having a temperature T 2000 K and particle number density of 1012 - 1013 cm-3. The velocity of sound as there would be about 3.4 km s-1. Thus, the planet's motion is supersonic, the Mach number M = Vp /as being about 9. This motion is similar to a motion of a large meteoritic body in the Earth's atmosphere (Tsikulin 1969). A strong conical shock wave, ionising gas and heating it to 10,000­15,000 K, is formed. We consider a simple model, in which a perturbing bo dy with diameter d is moving along a rectilinear tra jectory at velocity Vp > as through a medium with mass density 0 . Quantity E is the energy released at a unit path of the body's motion; E is numerically equal to the drag force exerted on the body by the medium: d2 F= (1) 0 Vp2 . 4 Owing to the drag, the planet is gradually spiraling into the red giant's atmosphere. The rate of decrease of its semima jor axis a is (see, e.g., Taam, Bodenheimer & Ostriker 1978): a FVp a . =- a GM Mp (2)

Using (1) and substituting a = 1 A.U., M = 1M , Vp = 30 km s-1 and Jupiter's parameters Mp = 1MJ = 1.9 в 1030 g, d = dJ = 1.4 в 1010 cm, we have a/a -8 -1 -8 в 10 year . Thus, during the red giant stage, which lasts not longer than 106 years, the semima jor axis of the planet's orbit decreases by 8%. For a larger planet (13MJ, 2.35dJ), braking is still smaller, 3.2%. The motion of the planet through the stellar atmosphere is similar to the case of propagation of a shock wave from a detonating cylindrical charge (Tsikulin 1969). That is, in any plane perpendicular to the tra jectory of the body, the propagating shock can be considered as a cylindrical one (see Fig. 3). The sho ck front radius in this plane is E 1/4 1/2 rf = t 0 with t = z/Vp . The front equation in the (r, z ) coordinates (Fig. 3, left) is (3)

rf 1/4 z 1/2 (4) = d 4 d Shock velocity D in the direction perpendicular to the tra jectory is decreasing with time as
4


For at a will rf ma side

1 E 1/4 -1/2 D= t . (5) 2 0 a planet with d = 2.35dJ, velocity D will fall to the velocity of sound as distance zmax = d(/4)1/2(M2/4) 5.6 в 1011 cm behind the body. There be no emission at greater z 's. The corresponding maximum front radius 11 cm. The maximum pro jected area of the shocked `cone' (for a x = 1.3 в 10 view, as in Fig. 3, right) is Ssh = 1 (4 ) 3
1/4

d

1/2 3/2 zmax

4.8 в 10

22

cm2 1.6 в 10-5S ,

(6)

2 where S = R 3 в 1027 cm2 is the stellar disc area for R 3 в 1013 cm. Observations and model calculations of the Balmer emission lines in Miras (e.g., Fox & Wood 1985) show that, for the above-mentioned parameters, the shock front yields up to 1020 H photons cm-2 s-1 ; with source area Ssh , this can account for the total Balmer line fluxes observed from a star at a distance of about 300 pc, a few в10-12 erg cm-2 s-1 (Fox, Wood & Dopita 1984). The side-view observations of the cone shock also naturally explain doubling and large linewidth ( 60 km s-1) of the Balmer lines, observed in Miras (e.g., Gillet et al. 1983, Udry et al. 1998). The lack of radio continuum is explained by the small angular size of the planetary source, which is though hot enough to pro duce the observed Balmer emission. Another effect of the planetary shock is enhanced H2O maser pumping by the mechanism proposed by Rudnitskij & Chuprikov (1990); the delay may be similar to the case of a spherical shock. As mentioned above, some stars (R Leo, R Cas and U Aur) displayed isolated bursts of the H emission, followed (about a year and a half later) by a flare of the H2O maser radio emission. This may be due to a periastron shock-wave episode of a planet in a highly eccentric orbit with a period P 15 years. Some other stars (e.g., U Ori, Rudnitskij et al. 2000) have already shown some hints to H2 O maser `superperiodicity' of 9 years -- which could be associated with planetary revolution periods. One of the main ob jections to the model proposed is that the drag of the surrounding medium tends to circularise the planet's orbit. However, Soker (2000) has shown that in the case of a mass-losing central star the passage of the periastron by a planet enhances the mass-loss rate, and this effect can support, and even increase the orbit eccentricity.

5


4

Observational Tests
· Repeated mapping of the H2O masers throughout the light. This would allow us to trace the propagation of a shock across the maser layer, igniting sequentially maser spots more and more distant from the stellar surface. · Optical and infrared interferometry of miras, aiming to detect hot spots (eventual planets) migrating across the stellar discs -- like the one probably observed in R Cas by Lopez et al. (1997). · Further radio continuum observations of red giants at shorter wavelengths (short centimetre and millimetre), similar to those performed on ATCA at = 6 and 3 cm by Chapman & Rudnitskij (2002). For a thermal spectrum, F 2, and shorter-wave observations are more promising for detection. In the millimeter continuum flux densities of about a millijansky from red giants are expected, due to photospheric emission. Episodes connected with shocks can also happen, in this case parallel monitoring in optical spectroscopy is useful for shock diagnostics.

References
Berulis, I.I., Lekht, E.E., Pashchenko, M.I., & Rudnitskij, G.M. 1983, SvA, 27, 179 Chapman, J.M., & Rudnitskij, G.M. 2002, PASA, submitted Esipov, V.F., Pashchenko, M.I., Rudnitskij, G.M., & Fomin, S.V. 1999, Astron. Lett., Fox, M.W., Wo o d, P.R., & Dopita, M.A. 1984, ApJ, 286, 337 Gillet, D., Maurice, E., & Baade D. 1983, A&A, 128, 384 Lekht, E.E., Mendoza-Torres, J.E., Rudnitskij, G.M., & Tolmachev, A.M. 2001, A&A, Reid, M.J., & Menten, K.M. 1997, ApJ, 476, 327 Rudnitskij, G.M., & Chuprikov, A.A. 1990, SvA, 34, 147 Rudnitskij, G.M., Lekht, E.E., Mendoza-Torres, J.E., Pashchenko, M.I., & Berulis, A&AS, 146, 385 Rybicki, K.R., & Denis, C. 2001, Icarus, 151, 130 Schneider, E. 2002, http://www.obspm.fr/planets/ Soker, N. 1999, MNRAS, 306, 806 Soker, N. 2000, A&A, 357, 557 Struck-Marcell, C. 1988, ApJ, 330, 986 Taam, R.E., Bo denheimer, P., & Ostriker, J.P. 1978, ApJ, 222, 269 Tsikulin, M.A. 1969, Sho ck Waves Induced in the Atmosphere by Motion of Large Bo dies (in Russian) (Moscow: Nauka) Udry, S., Jorissen, A., Mayor, M., & Van Eck, S. 1998, A&AS, 131, 25 Wo o d, P.R. 1979, ApJ, 227, 220 25, 672

376, 928

I.I. 2000,

Meteoritic

6