The habitable zone (HZ) around a star is typically defined as the region where a rocky planet can maintain liquid water on its surface. That definition is appropriate, because this allows for the possibility that carbon-based, photosynthetic life exists on the planet in sufficient abundance to modify the planet's atmosphere in a way that might be remotely detected. Exactly what conditions are needed, however, to maintain liquid water remains a topic for debate. In the past, modelers have restricted themselves to waterrich planets with CO 2 and H 2 O as the only important greenhouse gases. More recently, some researchers have suggested broadening the definition to include arid, "Dune" planets on the inner edge and planets with captured H 2 atmospheres on the outer edge, thereby greatly increasing the HZ width. Such planets could exist, but we demonstrate that an inner edge limit of 0.59 AU or less is physically unrealistic. We further argue that conservative HZ definitions should be used for designing future space-based telescopes, but that optimistic definitions may be useful in interpreting the data from such missions. In terms of effective solar flux, S eff , the recently recalculated HZ boundaries are: recent Venus-1.78; runaway greenhouse-1.04; moist greenhouse-1.01; maximum greenhouse-0.35; and early Mars-0.32. Based on a combination of different HZ definitions, the frequency of potentially Earth-like planets around late K and M stars observed by Kepler is in the range of 0.4-0.5.A s the appearance of this special issue of PNAS confirms, the search for exoplanets is by now well underway; indeed, it is one of the hottest research areas in all of astronomy. At the time of this writing, 428 exoplanets have been identified by groundbased radial velocity (RV) measurements and 279 planets have been discovered by the transit technique (1) and confirmed by various other methods, including RV. In addition, more than 3,000 "planet candidates" have been identified by the Kepler mission (2). Most of these planets are either too large or too close to their parent star to have any chance of harboring life. However, both astronomers and the general public are interested in identifying potentially habitable planets and in searching their atmospheres spectroscopically for evidence of life. This leads immediately to the question of how life can be recognized remotely, along with the related question of what are the conditions needed to support it. Liquid water is often mentioned as a prerequisite for life, and we will argue below that this restriction is appropriate for the astronomical search for life. However, some researchers have questioned this assumption (3), and so we begin by explaining why the presence of liquid water is so important in the search for life on planets around other stars.

Chemical composition affects virtually all aspects of astrobiology, from stellar astrophysics to molecular biology. We present a synopsis of the research results presented at the ''Stellar Stoichiometry'' Workshop Without Walls hosted at Arizona State University April 11-12, 2013, under the auspices of the NASA Astrobiology Institute. The results focus on the measurement of chemical abundances and the effects of composition on processes from stellar to planetary scales. Of particular interest were the scientific connections between processes in these normally disparate fields. Measuring the abundances of elements in stars and giant and terrestrial planets poses substantial difficulties in technique and interpretation. One of the motivations for this conference was the fact that determinations of the abundance of a given element in a single star by different groups can differ by more than their quoted errors. The problems affecting the reliability of abundance estimations and their inherent limitations are discussed. When these problems are taken into consideration, self-consistent surveys of stellar abundances show that there is still substantial variation (factors of *2) in the ratios of common elements (e.g., C, O, Na, Al, Mg, Si, Ca) important in rock-forming minerals, atmospheres, and biology. We consider how abundance variations arise through injection of supernova nucleosynthesis products into star-forming material and through photoevaporation of protoplanetary disks. The effects of composition on stellar evolution are substantial, and coupled with planetary atmosphere models can result in predicted habitable zone extents that vary by many tens of percent. Variations in the bulk composition of planets can affect rates of radiogenic heating and substantially change the mineralogy of planetary interiors, affecting properties such as convection and energy transport.

If mutual gravitational scattering among exoplanets occurs, then it may produce unique orbital properties. For example, two-planet systems that lie near the boundary between circulation and libration of their periapses could result if planet-planet scattering ejected a former third planet quickly, leaving one planet on an eccentric orbit and the other on a circular orbit. We first improve upon previous work that examined the apsidal behavior of known multiplanet systems by doubling the sample size and including observational uncertainties. This analysis recovers previous results that demonstrated that many systems lay on the apsidal boundary between libration and circulation. We then performed over 12,000 three-dimensional N -body simulations of hypothetical three-body systems that are unstable, but stabilize to two-body systems after an ejection. Using these synthetic two-planet systems, we test the planet-planet scattering hypothesis by comparing their apsidal behavior, over a range of viewing angles, to that of the observed systems and find that they are statistically consistent regardless of the multiplicity of the observed systems. Finally, we combine our results with previous studies to show that, from the sampled cases, the most likely planetary mass function prior to planet-planet scattering follows a power law with index -1.1. We find that this pre-scattering mass function predicts a mutual inclination frequency distribution that follows an exponential function with an index between -0.06 and -0.1.