The ``Gravitational Interaction of Compact Objects'' conference took place at the Kavli Institute of Theoretical Physics from May 12th to May 14th, 2003. Aside from bringing together researchers in the field to discuss the status of activities in the area, the conference served to ``kick-off'' a two-month workshop under the same name at KITP. Each day of the conference was divided into sections that focused on specific areas, which ranged from broad themes to particular issues.
The first day consisted of a series of talks to survey the status of different aspects of gravitational wave data astronomy. Fred Raab described the status of the LIGO detectors, and Joan Centrella gave an overview of the possible sources for both space based and ground based detectors. Lee Lindblom reviewed aspects of accretion induced collapse of white dwarfs and the marginal likelihood of detection of gravitational waves from these systems with advanced interferometers. Following these there were four review talks on the dynamics of binary systems. Oliver Poujade reviewed post-Newtonian computations of the gravitational waveform, based on matching a post-Newtonian expansion of the field equations in the near zone to a post-Minkowskian expansion in the far zone. Currently there are regularization ambiguities which arise at post-3-Newtonian order, and Poujade described ongoing attempts to resolve these ambiguities. John Baker described the status of the Lazarus project to simulate binary black holes, and the present efforts and results to obtain complete information from initial data sets describing equal mass spinning black holes. Luis Lehner gave an overview of different approaches to numerical relativity and to the binary black hole problem, and Miguel Alcubierre reviewed the status of binary black hole simulations within the Cauchy approach.
The second day's talks were mostly devoted to the current understanding of specific sources, together with one talk on data analysis. Phil Arras reviewed the generation of gravitational radiation by neutron star Rossby modes and the role gravitational waves can play in determining the observed/inferred properties of neutron stars. He argued that r-modes in low mass X-ray binaries (LMXBs) are a possible candidate for detection with LIGO II. This depends somewhat on unknown properties of the star's viscosity that govern the stability of a steady state solution in which the mode is excited. Detectable gravitational waves from LMXBs may also be produced via inhomogeneities in the crust. Next, Masaru Shibata reviewed the status of his simulations of binary neutron star mergers, presenting simulations of both equal and unequal mass cases which merge after a couple of orbits. Shibata argued that useful physical information should be obtainable from the current codes given larger computers and more efficient computer use. He also discussed incorporation of more realistic physical components like improved equations of state, neutrino cooling etc. Christian Cardall then surveyed the present status of supernovae simulations. He discussed shortcomings in present codes (lower dimensionality, Newtonian dynamics, incomplete physical processes incorporated), and reviewed the five-years Terascale Supernova Initiative which aims to address all these shortcomings.
The subject then switched to gravitational wave data analysis. Patrick Brady discussed data analysis methods for four different types of gravitational wave signals: known waveforms, unknown burst waveforms, periodic signals, and stochastic signals. He presented preliminary results from the analysis by the LIGO Science Collaboration (LSC) of LIGO's first science run, giving upper limits on the event rates of various types of sources.
Peter Meszaros then reviewed current understanding of gamma ray bursts (GRBs), their phenomenology, models for the central engine, and the likelihood of detecting gravitational waves from them. He also described planned detectors which will increase the number of GRBs detected and the accuracy of their sky locations.
Subsequently, Leor Barack highlighted extreme mass ratio binaries, in which neutron stars or solar mass black holes inspiral into supermassive black holes, as important sources for spaced based detectors like LISA. He discussed theoretical problems related to obtaining accurate templates for these inspirals, and recent results which should provide a way to compute such templates for generic orbits around spinning black holes. The day ended with Greg Cook's review talk on methods for solving the constraint equations to obtain initial data for interacting compact binaries. He showed that new methods based on the conformal thin sandwich approach agreed better with post-Newtonian results than did older methods.
On the third day, talks concentrated on pressing issues related to simulations of compact objects and highlighted several promising new ingredients. Two talks discussed ways to improve the availability computational power and the efficiency of its use. Ed Seidel gave an overview of how so-called Grid computing can provide considerable enhanced resources for realistic simulations. Frans Pretorius surveyed the application and promises of adaptive mesh refinement, showing explicit examples both in Cauchy and characteristic implementations.
The next two talks turned to general relativistic hydrodynamics simulations, and treated in depth what will be needed for accurate and realistic results. Mark Miller reviewed the effect on evolutions of the choice of initial data sets used, and discussed how to estimate the associated errors in physical observables that one extracts from the simulation. He illustrated these issues by showing explicit evolutions of multiple orbits of neutron star binary systems. Charles Gammie discussed the incorporation of magnetic fields in simulations, describing the implementation of the magneto-hydrodynamics equations and presenting explicit examples of accretion tori on spinning black hole backgrounds.
The conference then turned to new techniques that address issues which had previously been poorly understood. Jeffrey Winicour described the problems associated to dealing with boundaries, and in particular the issue of obtaining boundary conditions that make the initial/boundary value problem well posed. He described different efforts to achieve a solution of this problem in special cases. Manuel Tiglio described a set of novel techniques which, for linear problems, guarantee numerical stability of the implementation. He also described how the parameter freedom in families of formulations of Einstein's equations could be exploited to minimize the growth of constraint violating modes. He presented explicit examples in test problems where the use of this technique made a considerable improvement.
The conference served to highlight several recent advances in the simulation of compact objects, to review current outstanding problems associated with these simulations, and to improve our understanding of possible methods for overcoming the problems. In addition, it reviewed the context for these simulation and modeling efforts to detect and analyze gravitational wave signals, and the possible payoffs for scientific knowledge. All of the conference talks are available online at