Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userTemplate-based searches for gravitational waves: efficient lattice covering of flat parameter spaces
37
0
0.0
(
0
)
Authors
Reinhard Prix
Ask ChatGPT about the research
No Arabic abstract
The construction of optimal template banks for matched-filtering searches is an example of the sphere covering problem. For parameter spaces with constant-coefficient metrics a (near-) optimal template bank is achieved by the A_n* lattice, which is the best lattice-covering in dimensions n <= 5, and is close to the best covering known for dimensions n <= 16. Generally this provides a substantially more efficient covering than the simpler hyper-cubic lattice. We present an algorithm for generating lattice template banks for constant-coefficient metrics and we illustrate its implementation by generating A_n* template banks in n=2,3,4 dimensions.
rate research
Read More
Gravitational wave astronomy opened dramatically in September 2015 with the LIGO discovery of a distant and massive binary black hole coalescence. The more recent discovery of a binary neutron star merger, followed by a gamma ray burst and a kilonova, reinforces the excitement of this new era, in which we may soon see other sources of gravitational waves, including continuous, nearly monochromatic signals. Potential continuous wave (CW) sources include rapidly spinning galactic neutron stars and more exotic possibilities, such as emission from axion Bose Einstein clouds surrounding black holes. Recent searches in Advanced LIGO data are presented, and prospects for more sensitive future searches discussed.
Wide parameter space searches for long lived continuous gravitational wave signals are computationally limited. It is therefore critically important that available computational resources are used rationally. In this paper we consider directed searches, i.e. targets for which the sky position is known accurately but the frequency and spindown parameters are completely unknown. Given a list of such potential astrophysical targets, we therefore need to prioritize. On which target(s) should we spend scarce computing resources? What parameter space region in frequency and spindown should we search? Finally, what is the optimal search set-up that we should use? In this paper we present a general framework that allows to solve all three of these problems. This framework is based on maximizing the probability of making a detection subject to a constraint on the maximum available computational cost. We illustrate the method for a simplified problem.
We introduce a method based on the loudest event statistic to calculate an upper limit or interval on the astrophysical rate of binary coalescence. The calculation depends upon the sensitivity and noise background of the detectors, and a model for the astrophysical distribution of coalescing binaries. There are significant uncertainties in the calculation of the rate due to both astrophysical and instrumental uncertainties as well as errors introduced by using the post--Newtonian waveform to approximate the full signal. We catalog these uncertainties in detail and describe a method for marginalizing over them. Throughout, we provide an example based on the initial LIGO detectors.
We describe the application of the lattice covering problem to the placement of templates in a search for continuous gravitational waves from the low-mass X-Ray binary Scorpius X-1. Efficient placement of templates to cover the parameter space at a given maximum mismatch is an application of the sphere covering problem, for which an implementation is available in the LatticeTiling software library. In the case of Sco X-1, potential correlations, in both the prior uncertainty and the mismatch metric, between the orbital period and orbital phase, lead to complications in the efficient construction of the lattice. We define a shearing coordinate transformation which simultaneously minimizes both of these sources of correlation, and allows us to take advantage of the small prior orbital period uncertainty. The resulting lattices have a factor of about 3 fewer templates than the corresponding parameter space grids constructed by the prior straightforward method, allowing a more sensitive search at the same computing cost and maximum mismatch.
Gravitational wave searches to date have largely focused on non-precessing systems. Including precession effects greatly increases the number of templates to be searched over. This leads to a corresponding increase in the computational cost and can increase the false alarm rate of a realistic search. On the other hand, there might be astrophysical systems that are entirely missed by non-precessing searches. In this paper we consider the problem of constructing a template bank using stochastic methods for neutron-star--black-hole binaries allowing for precession, but with the restrictions that the orientation of the total angular momentum of the binary is pointing towards the detector and that the neutron-star spin is negligible relative to that of the black-hole. We quantify the number of templates required for the search, and we explicitly construct the template bank. We show that despite the large number of templates, stochastic methods can be adapted to solve the problem. We quantify the parameter space region over which the non-precessing search might miss signals.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
