Introduction

At the time of this workshop, there are now more than 150 detected extrasolar planets discovered and 13 confirmed multiple planet systems with more candidate planets and systems being evaluated (G. Marcy in Chapter 11 of this book). There is a growing number of ground-based observations of young circumstellar and protoplanetary disks as well as volumes of data from Spitzer and Cassini. There are two scenarios for gas-giant planet formation that are sufficiently sophisticated to provide results and predictions. Clearly, the ingredients are present for planetary scientists to develop a comprehensive or, at least, a cohesive model for the formation of the gas-giant planets in the Solar System and in extrasolar systems and, to initiate resolution of the age old question: how do planets form?

The subject of this chapter is the core accretion–gas capture model, the more generally accepted scenario for gas-giant planet formation (see also Chapter 8 by Thommes and Duncan for the core accretion itself). Please note that this model has had many names over the decades of its development: planetesimal hypothesis, nucleated instability model, and core instability model. However, it seems best to call it by the more descriptive label: the Core Accretion–Gas Capture model, the CAGC or, the short version, the core accretion model.

Introduction

The two mechanisms that have been advanced for explaining the formation of giant planets are core accretion (“bottom up”) and disk instability (“top down”). Core accretion, the conventional mechanism, relies on the collisional accumulation of planetesimals to assemble ∼ 10 M⊕ solid cores, which then accrete massive gaseous envelopes from the disk gas (Mizuno, 1980; Lissauer, 1987; Pollack et al., 1996; Kornet et al., 2002; Inaba et al., 2003). In this scenario, the ice-giant planets probably did not form in situ (Levison and Stewart, 2001), but rather formed between Jupiter and Saturn and then were scattered outward to their present orbits (Thommes et al., 1999, 2002).

The alternative to core accretion is disk instability, where gas-giant protoplanets form rapidly through a gravitational instability of the gaseous portion of the disk (Cameron, 1978; Boss, 1997, 1998b, 2000, 2002a,b, 2003, 2004; Gammie, 2001; Mayer et al., 2002, 2004; Nelson, 2000; Nelson et al., 1998, 2000; Pickett et al., 1998, 2000a,b, 2003; Rice and Armitage, 2003, Rice et al., 2003a) and then more slowly contract to planetary densities. Solid cores form simultaneously with protoplanet formation by the coagulation and sedimentation of dust grains in the clumps of disk gas and dust.

Recent improvements in radial velocity precision

Since the first discovery of an extrasolar planet around a Solar-type star ten years ago (Mayor and Queloz, 1995), research in this field has been very productive and has led to the detection of more than 140 exoplanets. The vast majority of these discoveries has been made with the radial-velocity (RV) technique, i.e. the precise measurement of the RV wobble that a planet induces in its parent star due to its orbital movement. A major effort to improve the accuracy of the RV measurements has been undertaken by several groups, since this is absolutely necessary to detect the RV signatures of giant planets, in the range 1–100 ms−1. Two main techniques were developed: one using a ThAr calibration simultaneously with each observation (Baranne et al., 1996) to track instrumental drifts, and one using an iodine absorption cell, superimposing a reference spectrum on the stellar spectrum (Butler et al., 1996). Both techniques have been able to deliver RV precision at the level of ∼3 ms−1, opening the way to the discovery of many planetary systems.

Over the past decade, the exoplanet group at Geneva Observatory has been operating two high-resolution spectrographs able to achieve high RV precision, namely the ELODIE instrument mounted on the 1.93 m telescope at Observatoire de Haute-Provence (France), and the CORALIE instrument installed on the Swiss 1.2 m telescope at La Silla Observatory (Chile). Both ELODIE and CORALIE are high-resolution (R = 50 000), fiber-fed echelle spectrographs.

Introduction

Planets form within circumstellar disks composed of a mixture of gas and dust grains. These disks result from the gravitational collapse of rotating molecular cloud cores. They are initially rather massive and consist of about 0.3 M*, where M* is the mass of the central star (e.g. Yorke et al., 1995). In contrast, the minimum mass required to build the planets of our Solar System is only about 0.01 Solar masses (M⊙). Evidently, there are processes that redistribute the mass, transform the dust to larger particles, and disperse much of the gas and dust.

The processes which are responsible for the dispersal of the gas influence the formation of planets. For example, the timescale for gas dispersal as a function of the disk radius affects the composition of the resulting planetary system. As long as the dust particles are small enough to be tightly coupled to the gas, they follow the gas flow. If the gas is dispersed before the dust particles have had a chance to grow, all the dust will be lost and planetesimals and planets cannot form. Even if there is time for particles to coagulate and build sufficiently large rocky cores that can accrete gas (Pollack et al., 1996; Hubickyj et al., 2004), the formation of gas-giant planets like Jupiter and Saturn will be suppressed if the gas is dispersed before the accretion can occur.

Introduction: geoscience meets astronomy

Much of our knowledge about the formation of planets in the Solar System and in particular concepts and ideas about the origin of the Earth are derived from studies of extraterrestrial matter. Meteorites (Sears, 2004; Lauretta et al., 2006; Krot et al., 2006) were available for laboratory investigations long before space probes were sent out for in situ investigations of planetary surfaces, or Moon rocks were brought back to Earth. Meteorite studies provided such important parameters as the age of the Earth and the time of formation of the first solids in the Solar System (Chen and Wasserburg, 1981; Allegre et al., 1995; Amelin et al., 2002), as well as the average abundances of the elements in the Solar System (Anders and Grevesse, 1989; Palme and Jones, 2004). Traditionally, the study of rocky material requires techniques that fundamentally differ from astronomical techniques. While electromagnetic radiation from stars is analyzed by spectroscopy, the solid samples of aggregated cosmic dust and rocky matter from planetary surfaces require the use of laboratory instruments that allow the determination of their chemical and isotopic composition. Planetary surface materials are present in polymineralic assemblages. Formation conditions and thermal stability of individual minerals provide important boundary conditions for the genesis and history of the analyzed materials. Such studies require a thorough mineralogical background. The abundances and properties of the rock-forming elements, the focus of geo- and cosmochemical research, are, however, not necessarily of major concern to astrophysicists.

Introduction

Given that brown dwarfs are usually much more massive than planets (see Section 15.6), it is somewhat surprising that the first incontrovertible discovery of a brown dwarf (Nakajima et al., 1995) and the discovery of the first extrasolar planet (Mayor and Queloz, 1995) were announced simultaneously in 1995. Over the past decade, the rapid progress made in both fields has been extraordinary. There are now more than 150 extrasolar planets known, including more than a dozen multiple-planet systems. The first brown dwarf, Gliese 229B, was found in orbit around an M-dwarf, but in the same year other candidates, later confirmed to be free-floating brown dwarfs, were announced (e.g. Teide 1 by Rebolo et al., 1995), along with PPl 15 which was later discovered to be a binary brown dwarf (Basri and Martin, 1999). Observations now suggest that brown dwarfs are as common as stars, although stars dominate in terms of mass (e.g. Reid et al., 1999).

Since the rest of this book is devoted to the topic of planets, in this chapter I will review the properties and potential formation mechanisms of brown dwarfs, comparing and contrasting them with planets, but referring the reader to the other chapters of the book for detailed information on planets.

Masses

The most fundamental parameter of a brown dwarf is its mass.

Introduction

According to the core-accretion model for planet formation, the building blocks of planets are formed by the coagulation of dust grains, growing from the initial sub-micron sizes inherited from the interstellar medium to the 100 kilometer sizes of full-grown planetesimals. This is a growth process over 12 orders of magnitude in size and 36 orders of magnitude in mass. The physics of dust is of crucial importance for the study of planet formation. It also plays a major role in the structure and evolution of protoplanetary disks, since the dust carries most of the opacity of the dust–gas mixture in these disks and provides the surface for chemical reactions. Moreover, infrared and (sub)millimeter observations of dust continuum emission from these disks can be used as a powerful probe for the disk structure and mineralogical composition. A deep understanding of the physics of dust and the coagulation of grains is therefore of paramount importance for the study of the formation of planets and the circumstellar disks in which they are formed.

The study of grain coagulation and the formation of planetesimals has a long history. At the start of the twentieth century an equation for coagulation of colloidal particles was formulated by Smoluchowski (1916), though not related to astrophysical applications. The continuous form of that equation was later used to study the size distribution of fog particles in the Earth's atmosphere (Shumann, 1940).

Introduction

Rather few facts can be considered as acceptable to all who are working in the field of planet and planetesimal formation. Starting there, we will explore the possible pathways as suggested by experiments. It is certainly undisputed that the regular mode of planet formation is connected to protoplanetary disks. These disks consist mostly of gas, which makes up about 99% of their mass. The remaining 1% resides in the form of dust and – depending on the temperature – in the form of ice. As terrestrial planets are mostly built from heavier elements it is natural to assume that they are somehow assembled from the dust component in the disk.

Whatever model is placed between the dust and the planets, collisions between the solid bodies are unavoidable. In fact a large part of the process of planet formation can be based on collisions which can and (at least partly) will lead to the formation of larger bodies.

In the following sections we will review experiments that have studied these collisions and eventually put these results in a rough sketch of planetesimal formation. It is sometimes argued that collisions of large bodies might be too energetic to lead to the formation of a still larger body (Youdin and Shu, 2002). As described in this chapter it is true that collisions can lead to erosion rather than growth. However, we will show that this is not necessarily so for all collisions.

Introduction

The observation of a transit in our own Solar System is a long-lasting experience. Historical events related to transits can be traced back to Ptolemy who mentioned in his “Almagest” that the lack of detections of transits was not in contradiction with Mercury and Venus being closer to the Earth than the Sun (in the geocentric system) simply because they could be either too small to be detected or their orbital plane could be slightly tilted to the Solar one (Gerbaldi, personal communication). In 1607 Johannes Kepler thought he had directly observed a predicted Mercury transit but in fact only followed sun spots. He did, however, predict the next transits of Venus and Mercury to take place in 1631 following the extremely accurate observations of the planets by Tycho Brahe. The first transit to be observed was the Mercury transit in 1631 with the best observations leading Pierre Gassendi to evaluate its diameter to be less than 20 arcsec, much smaller than ever thought before. All the following transit observations led to new ephemerides and estimates of the size of the Solar System, but not as accurate as expected because of the difficulty of locating in time the entrance and exit of the planetary disk over the Solar one. First pictures of the Venus transit were made as early as in 1874 (Fig. 9.1). The transit of Mercury was also observed with the Solar and Heliospheric Observatory (SOHO) spacecraft from the L5 Lagrange point of the Earth.

Introduction

The observed characteristics of molecular clouds from which stars form can be reproduced by simulations of magnetohydrodynamic (MHD) turbulence, indicating the vital role played by magnetic fields in the processes of star formation. The fields support dense cloud cores against collapse, but they cannot do so indefinitely, because only charged particles couple to the field lines while neutral atoms and molecules can freely slip through. Through this process, called ambipolar diffusion, the cores slowly contract. The recombination rate in denser gas increases, causing the ionisation degree of the core to decrease. According to available observational data, once the core has contracted to ∼0.03 pc it decouples from the magnetic field and enters the dynamic collapse phase. During the collapse the angular momentum is locked into the core and remains unchanged (Hogerheijde, 2004).

Protostellar collapse and formation of disks

The typical specific angular momentum of a core on the verge of dynamic collapse, jc, amounts to ∼1021 cm2 s−1, and is many orders of magnitude larger than the typical specific angular momentum of a star (Hogerheijde, 2004). The inevitable conclusion is that the protostellar object resulting from the collapse must be surrounded by a large, rotationally supported disk (hereafter, protoplanetary disk) in which the original angular momentum of the core is stored. The outer radius of the disk, rd, may be roughly estimated based on Kepler's law.

This conference truly reflects a microcosm of an explosive revolution in the quest to understand the origin of planet and star formation. The diverse nature of this wide-open field necessitates a multi-facet attack on all relevant issues. In this pursuit, it is particularly important to find the missing links between the many seemingly independent observations as circumstantial clues around a global picture. The development of a comprehensive coherent interpretation requires an integrated approach to identify the dominant physical processes which determine the physical characteristics of planets and the dynamic architecture of planetary systems.

On the basic concept of planetary origin, there is very little difference between modern theories and the original Laplacian hypothesis. The coplanar geometry of all the major planets' orbits hardly needs any extrapolation for theorists to postulate the scenario that the planets formed long ago in a rotational flattened disk which is commonly referred to as the Solar Nebula. Today, we have direct images and multi-wavelength spectra of protostellar disks within which planet formation is thought to be an ongoing process. Perhaps the biggest advancement in the past decade is the discovery of over 100 planets around nearby stars other than the Sun. For the first time in this scientific endeavor, the Solar System reduces its unique importance to a single entry in the rapidly growing database of planetary-system census.

Abstract

This polemic considers the reality and implications of broad double-peaked Balmer, and super-broad asymmetric FeK alpha, emission lines in quasars. Current evidence suggest that both are rare. The lack of physical consistency and/or correlation in a disk model parameter space suggests little support for the claims that these lines arise from an accretion disk.

Introduction

Everyone knows that the central engine of a quasar involves accretion onto a supermassive black hole (SMBH). What else can it be? Especially if the Doppler interpretation of quasar redshifts is accepted and is correct. What else could the redshift be? This ideological approach to science is both good and bad. Without a paradigm research in this area would lack any focus or direction. Such anarchy is clearly out of favor. The danger, however, is that a paradigm can be confused as fact and efforts to explore, or even hypothesize, alternatives is discouraged and even suppressed. This tendency can be re-enforced in more ideologically oriented cultures because scientists, after all, are not immune to the weaknesses of the societies in which they work.

There is a tendency, when a paradigm becomes too strong, for observations to be treated with a measure of skepticism and even contempt. This is true unless they support the prevailing beliefs. Part of the disbelief can stem from genuine skepticism given the difficulty of obtaining good data. ‘Good’ is obviously an ill-defined term but, in this context, it involves a clear understanding of what a given set of data can, and cannot, tell you. In other words the ability of that data to constrain models. Quasar spectroscopy is a good example of this tendency.

Abstract

There is now a large consensus on the preferred cosmological model, which is known as the concordance model. This model relies on the introduction of a cosmological constant that represents the dominant form of energy densities in the present-day Universe. I briefly discuss the fact that from an astrophysicist point of view the evidence for a cosmological constant, although compelling, is not of sufficient robustness to consider that its existence has been demonstrated beyond reasonable doubt. I present the preliminary results of the Ω project, a large XMM program devoted to observing distant SHARC clusters. For the first time a measurement of the L–T evolution with XMM has been obtained. We found clear evidence for a positive evolution of the L–T relation, in agreement with most previous analysis based on ASCA and Chandra observations. Its cosmological implication is also discussed based on a new analysis of different X-ray surveys: EMSS, RDCS, MACS, SHARC, 160 deg2. It is found that a high matter density model fits remarkably well all these surveys, in agreement with all existing previous analyses following the same strategy. Concordance models produce far more high redshift massive clusters than observed in all existing X-ray surveys. This failure could indicate a deviation from the expected scaling of the M–T relation with redshift. However, no signature of such a possibility is found in existing data. I conclude that the properties of distant X-ray clusters as evidenced by XMM provide reliable indication in favor of an Einsteinde Sitter universe.

Abstract

Claims that ordinary spiral galaxies and some classes of QSO show periodicity in their redshift distributions have been investigated using high-precision data and rigorous statistical procedures. The periodicities are broadly confirmed. They are easily seen by eye in the data sets. Observational, reduction, or statistical artefacts do not seem capable of accounting for them.

Introduction

“Anomalous redshift” claims have appeared in the literature for about 30 years now and are associated with a few astronomers such as H. Arp, the Burbidges, and W. G. Tifft. The claims are controversial and the author has been engaged in a long-term project to examine them objectively. Probably the easiest to test are the claims of redshift periodicity. The search for periodicity in noisy data has a large literature and is a well-understood process. Three such claims have so far been examined, namely the 72 km s−1 periodicity (in the Coma cluster), the 36 km s−1 galactocentric periodicity (in wide-profile field spirals), and the periodicity 0.089 in log10(1 + z) (in the redshifts of QSOs close to bright, nearby spirals).

The approach in all cases has been the same: To use high-quality redshift data, not previously used in formulating the hypothesis, and rigorous statistical methods. Modern computing power now allows one to generate large numbers of synthetic data sets with which the real data can be compared. Here I describe the overall approach and results rather than the technicalities. The latter can be found in papers in the reference list.