Research School of Astronomy & Astrophysics

RESEARCH PROGRAMS

* not a member of the University

non-RSAA member of the University

# formerly a faculty member

STARS

The Southern Nearby Star Program

After a period of relative neglect, there is renewed interest in nearby stars in the wake of the discoveries of extrasolar Jovian planets, planetary debris disks, and water in the interstellar medium. In anticipation of future space missions having unprecedented sensitivity to the detection of planets around nearby stars, it is highly desirable to fill gaps in our knowledge of these objects and to identify as completely as possible those stars within 25 pc or so of the sun. Evidence suggests we may only know half the stars in this sample.

With this in mind, Ianna* and Begam*, in collaboration with Bessell, are continuing the CCD parallax program with the 1 m reflector at Siding Spring Observatory. Emphasis is placed on brighter stars from the Catalog of Nearby Stars without previous parallaxes and newly identified nearby star candidates. These objects are easily observed, and we can reach precisions of 1 - 2 milliarcseconds in the parallaxes at this telescope. We have finished parallax determinations for 24 stars; 17 are within 25 pc, and three are within 10 pc. The result of these fundamental observations will be the establishment of a comprehensive database for southern nearby stars.

MACHO Observations of LMC Red Giants: definite pulsators and possible binary stars

The MACHO experiment has monitored stars in the Large Magellanic Cloud (LMC) almost daily for more than 6 years. This has resulted in light curve data of unprecedented quality and quantity for many tens of thousands of red giant stars. Wood and The MACHO Collaboration have analysed a small fraction of these light curves. They have solved some long-standing problems related to the pulsation of red giants, and found some intriguing new results.

The average luminosity (energy output rate) of variable red giant stars in the LMC is plotted against the logarithm of the period of variability in Figure 5. Five distinct period-luminosity sequences can be seen. Comparison of observed periods, luminosities and period ratios with theoretical models allowed Wood and collaborators to identify Mira variables (the largest amplitude variables) unambiguously as radial fundamental mode pulsators, thus solving a long-standing controversy regarding the pulsation mode of these stars. The smaller amplitude, semi-regular variables were found to be pulsating in the 1st, 2nd or 3rd overtone, or even the fundamental mode. All these variables lie on just 3 of the 5 distinct sequences in Figure 5 (A, B and C).

Stars on the fourth sequence (D) in Figure 5 show semi-regular, eclipse-like light curves. The light curves and periods of these stars can be interpreted as being due to a pair of orbiting stars, with the red giant star (the one that can be seen) transferring matter onto a close, invisible companion (such a system is called a semi-detached binary). The transferred matter forms a dusty, comet-like cloud around the orbiting companion star: each time the orbiting comet and its tail passes between us and the red giant, it causes an apparent fading of the red giant. These stars make up about 25% of the bright red giants.

Annual Report 1998

Figure 5. The luminosity in visual magnitudes of variable red giants in the LMC plotted against the period in days. Note that the stars tend to lie on five sequences A...E.

Stars on the fifth sequence in Figure 5 (E) are generally less luminous than the other variables: in fact, most stars at these luminosities are non-variable. The light curves strongly suggest that we are looking at a pair of orbiting stars so close to each other that they are in contact (called a contact binary system). The light variations arise as the two stars orbit each other so that we see the system at different orientations. These stars make up about 0.5% of the red giants at their observed luminosity.

The existence of so many close binary stars among the red giants is very hard to explain. In fact, the numbers of objects seen is more consistent with the number of red giants expected to have planets orbiting the tops of their atmospheres than with the number of red giants expected to have close stellar companions. Work is continuing to try to confirm or disprove the binary nature of the stars on sequences D and E. If the existence of these red giants in contact and semi-detached binaries is confirmed, then existing theories of binary star evolution will require substantial modification.

Soft X-ray Transients

A large fraction of stars in our Galaxy are found in binary systems. In some systems, one of the two components is a compact object (black hole, neutron star or white dwarf) accreting mass from the companion star. The mass lost by the companion star can be transferred onto the compact object via a stellar wind or through an accretion disk. In both cases a large amount of gravitational energy of the infalling gas is released in the form of thermal energy (UV or X-rays, depending on the mass of the accreting object).

Research School of Astronomy & Astrophysics

A class of X-ray binary systems known as Soft X-ray Transients (SXTs) shows transitions between different states, with periods of quiescence separated by energetic outbursts; an increase in the optical brightness and radio flares often accompany the X-ray outbursts. Spectroscopic measurements of the binary periods and the optical velocities of the binary components strongly suggest that the accreting object in most SXTs is a black-hole candidate (a compact object with a mass larger than the maximum mass of a neutron star).

Soria, Wu*, Hunstead* and Johnston* conducted spectroscopic and photometric studies of some X-ray transients in different states, in order to determine their physical parameters and in particular the mass of the compact object, and to investigate the geometry and the structure of the accretion disk in both states. From the X-ray and radio they found that optical Balmer line emission is correlated with the formation of a hard X-ray emitting corona. They also investigated the similarities in the optical spectrum between Galactic SXTs and active galaxies.

Hipparcos Data for Two Open Clusters containing Cepheids

The distance scale of the universe rests heavily on the knowledge about the luminosity scale of the cepheids, which is calibrated from those cepheids that are members of open clusters. The very fact that some cepheids can be shown to be likely members of clusters is thus of interest. Using proper motion data obtained from the Hipparcos satellite Lynga and Lindegren* have shown the membership of the cepheids S Nor and U Sgr of two open clusters. The obtained parallaxes support the result.

The Most Metal-poor Stars

Norris, with Ryan* and Beers*, surveyed several thousand metal weak candidates, searching for the most metal poor stars. These objects were the first to form in the Galaxy, and their study provides clues as to the manner in which the first heavy elements were produced, the way in which the Galaxy itself formed, and constrains the theoretical models of the first supernovae. This work is part of a larger worldwide consortium of investigators organized by Beers, and preliminary results of the project were reported by him at the Third Stromlo Symposium. There were two highlights. First, the survey has uncovered some 110 objects having heavy element content less than 1/1000 that found in the Sun: a decade ago fewer than 10 such stars were known. This provides an important database for 8m telescope investigations of chemical enrichment at the earliest times. Second, no stars have been discovered with 1/10000 the heavy element content of the sun. The simplest models of galactic chemical enrichment would have predicted that about 10 such stars should have been found. The simplest explanation of this result is that the gas cloud from which the Galaxy formed had already experienced pre-Galactic chemical enrichment.

The Abundance of Lithium in the Most Metal-poor Stars

Norris, with Ryan*, Beers*, and Deliyannis*, determined the lithium abundances of 23 very metal-poor field main sequence stars, in order to determine the dispersion of abundance at the earliest times, and hence constrain the amount of lithium which emerged from the Big Bang. Temperatures were restricted to the range 6200-6400 K, and metallicities were less that 1/300 that of the sun. (Two decades ago the French astronomers Monique and Francois Spite discovered that for metal-poor halo stars in the effective temperature range 5700-6400 K the abundance of lithium was constant as a function of the stars' temperatures, leading to the term Spite Plateau. They postulated that the plateau value is the abundance which emerged from the Big Bang.)

Annual Report 1998

Of the 23 stars observed, 21 have Li abundances which are very tightly clustered, with a dispersion of only 0.03 dex (abundances varying by 7 percent) as is shown in Figure 6. This may be compared with previous determinations of 0.06-0.08 dex (15-20 percent). The two outliers in the figure clearly deviate from the population, and are members of a minority class of Li-depleted stars which are not at all understood. The observed dispersion is comparable to the errors of measurement, suggesting that the Spite Plateau is paper-thin in the ranges observed, which is a surprising result given that several mechanisms of Li destruction and formation are known to operate both in stars and the interstellar medium. It is quite impressive that 12 billion years after the event the dispersion is so small. The essentially zero intrinsic spread leads to the conclusion that either these stars have all changed their surface Li abundances very uniformly, or else they exhibit close to the cosmological primordial abundance.

Figure 6: The abundance of lithium, A(Li), as a function of stellar effective temperature for 23 very metal-poor stars.

Photoelectric Heating by Dust in Planetary Nebulae

As stars like our Sun die, they eject clouds of gas and dust which become ionised in the planetary nebula (PNe) phase of evolution. Although the photoionisation codes used to analyse these PNe fully account for the atomic processes, they have not treated the physics of the dust very well. Dopita & Sutherland have shown that these can be very important in the interpretation of PNe spectra. Their image on the back cover shows the planetary nebula K-1-3 in the light of singly-ionised nitrogen (red) hydrogen (green) and doubly-ionised oxygen (blue). The hot regions around the central star appear purple in this image. This is the region where photoelectric heating by grains is important.

In the inner regions of the PNe, the gas is highly ionised, and photoelectric heating from atoms is not very large. However, if the PNe contains a lot of small dust grains, then the heating of the gas by photoelectrons from grains becomes the dominant heating mechanism in the inner zones of the PNe. Dopita & Sutherland have found that this has a very strong effect on both the thermal structure and the emission line spectrum of highly ionised PNe.

Research School of Astronomy & Astrophysics

Eta Carinae Low Excitation Event

Eta Carinae is possibly the most massive star in the Milky way galaxy. It is highly variable, and is a strong candidate for a supernova explosion. Eta Carinae underwent its "Great Eruption" in 1843 during which time it was the second brightest star in the sky. It then faded rapidly, brightened again briefly around 1892 during the "Lesser Eruption", but remained faint until about 1940. Since then it has been steadily increasing in brightness to about V=6 mag where it is today. Several "shell ejections" have been detected spectroscopically over the last 50 yr, but it was only in 1996 that it was recognised that these, and possibly the Great Eruption and Lesser Eruption, occur at regular intervals with a period of 5.52 yr. This raised the intriguing possibility that the periodicity may be fundamentally linked to the unknown cause of the Great Eruption. The 5.52 yr period is difficult to explain with a single star model because it does not correspond to any known oscillation mode of the star. However, it can be interpreted as the orbital period of a binary star system. Then the connection with the Great Eruption would be less direct. More data were needed, and the 5.52 yr period made it possible for the first time to predict the occurrence of the next "shell ejection" in late 1997/early 1998.

McGregor with Rathborne, an MSSSO summer student, used the 74 inch telescope Coude echelle spectrograph to monitor Eta Carinae between November 1997 and March 1998. The low excitation event indeed occurred as predicted. While dramatic emission-line profile variations were recorded, the velocities of the emission-line bases clearly did not change significantly in several cases. Either the profile variations were due to structural changes in the stellar wind from Eta Carinae and unrelated to orbital motion, or systematic orbital motion of the putative binary system was masked by the stellar wind. Prominent blueshifted absorption features were present on lines of H I and He I. These confirmed that a significant change had occurred in the stellar wind structure, irrespective of the possible binary nature of the star. The X-ray light curve for Eta Carina showed a dramatic decline around this time that has been interpreted as an occultation of the X-ray emitting region by the hypothesized companion. This result remains at odds with the interpretation from optical spectroscopy, and indicates that detailed modelling of the system will be required to further reveal its nature.

THE GALAXY

The Massive Astronomical Compact Halo Object Survey

The MACHO Project is searching for galactic dark matter by looking for its gravitational lensing effects on the light of background stars. The changing geometry of the lensing system as the lens moves relative to the background star causes the brightness of the background star to change with time. These 'microlensing' events can be detected by monitoring the brightnesses of large numbers of stars. MACHO began gathering data in 1992 using the 50 inch telescope at Mt. Stromlo, and now monitors roughly forty million stars in the Large and Small Magellanic Clouds, and in the Galactic Bulge. A large number of microlensing events have now been detected, with well over 200 along lines of sight to the Bulge, over 16 toward the LMC, and two towards the SMC.

During 1998 our major emphasis was on completing the analysis of the first five years of LMC data. This represents a delay of about a year from our previous analysis plan, due to delays in calculating the definitive detection efficiencies for the experiment. The efficiencies are now well in hand, and we expect the five year LMC paper to be submitted by the middle of this year.

Annual Report 1998

During 1998 a second event was detected toward the SMC. The lensing object was a binary system, the light curve exhibiting two caustics. Our ability to detect microlensing events shortly after they begin was put to extremely good use during this event. The community was alerted early to the fact that a binary event was in progress, and as a result a number of followup telescopes were employed around the world to get the best possible time sampling of the event. This allowed the duration of the caustic crossing to be measured, which in turn allows the location of the lens to be roughly determined. The result is unambiguous: the lens is located in, or near to, the SMC. This result has resulted in renewed suggestions that all of the lenses observed toward the Magellanic Clouds may be in the Clouds themselves, or in an intervening stellar populations. Some of these suggestions cannot yet be tested with the available observational data. At least two, however, were tested during the year and found to be contradicted by the data. These are the suggestion by Evans et al that we are seeing stars from our own galactic disk, warped and flared into our line of sight; and the suggestion by Zaritsky et al that there is evidence in the colour-magnitude diagram of the LMC for an intervening stellar population.

Bulge Image Subtraction

Axelrod has been supervising the PhD thesis work of Andrew Drake, who is performing a pixel lensing analysis of MACHO bulge data through image subtraction. This work is nearing completion, and two papers have been accepted for publication. As we expected from our initial calculations, difference imaging detected nearly twice the number of events found by standard MACHO photometry. Additionally, the photometry of these events is of much higher quality, making them more valuable for analysis. We are seriously considering applying the difference imaging technique to the entire MACHO dataset.

THE MAGELLANIC CLOUDS

The Nature of the Faint Star Cluster AM-3

In the late 1970s visual searches of Sky Survey films by Madore and Arp revealed the existence of three previously uncatalogued faint star clusters, which they designated AM-1, -2 and -3. AM-1 has since been shown to be one of the most distant of the Galaxy's halo globular clusters, while AM-2 is now recognized as a distant, reddened Galactic open cluster. AM-3 however, has been essentially ignored despite its potential association with the SMC; it lies approximately 4.5 deg from the centre of the SMC just within the outermost density contours of the SMC field population. Analysis of CCD images by Da Costa have now confirmed the association of AM-3 with the SMC. In particular, the cluster colour-magnitude diagram shows a sparse clump of red giants at V = 19.6 +/- 0.1 and the top of the cluster main sequence at V ~= 22. The magnitude of the AM-3 clump stars is comparable to those for old and intermediate-age SMC clusters while, at distance of the SMC, the luminosity of the AM-3 main sequence turnoff indicates an age of approximately 6 Gyrs for the cluster. Thus, despite its extreme location relative to the SMC centre, AM-3 is an intermediate-age object. Its projected distance from the centre of the SMC is ~4.5 kpc, larger by almost 1 kpc than the previous western-most SMC cluster identified (Lindsay 1) and larger also than the projected distance of the eastern-most cluster Lindsay 113. The absolute magnitude of AM-3 is estimated as Mv = -3.5 +/- 0.5, an intrinsically faint value. Consequently, it is quite probable that AM-3 would not have been noticed if lay closer in to the main body of the SMC.

Research School of Astronomy & Astrophysics

The HI Structure of the Large Magellanic Cloud

Using the Australia Telescope Compact Array, a complete HI survey of the Large Magellanic Cloud (LMC) has been completed by Kim et al. (1998a). This survey covers 11.1x12.4 degrees on the sky at a resolution of one arc minute. These maps show an amazing complex of expanding giant and supergiant HI shells. For the giant shells, the expansion velocity increases from about 15 km s-¹ for the smallest up to about 20-35 km s-¹ showing that these are being accelerated by the energy input from the young massive stars within them. On the other hand, the supergiant shells show no trend to increasing velocity of expansion with radius. These supergiant shells have become larger than the scale height of the matter in the disk of the LMC, and the pressurised gas that drives the expansion is venting into the halo.

As can be seen from Figure 7, in which the HI is shown as white tones, the ionised gas is shown as dark tones, and identified HI giant shells are also shown, a high proportion of the giant and supergiant HI shells are colliding with one another. Many of them are interlocking, especially near the 30 Dor complex, where very active star formation has been occurring simultaneously in many different centres. The fact that the HI velocity structure in both the LMC and the SMC is dominated by a series of interlocking and interacting bubbles shows graphically that these are providing a fundamental control of the phase structure of the interstellar medium in this galaxy. The gas at the points of collision is compressed, and forced into collapse to form new generations of young stars. These are visible as regions of i