On detection of black hole quasinormal ringdowns: Detection efficiency and waveform parameter determination in matched filtering
On detection of black hole quasinormal ringdowns: Detection efficiency and waveform parameter determination in matched filtering
复制标题
DOI：
10.1103/physrevd.71.103005
复制
发表时间：
2004-10
影响因子：
5
通讯作者：
Y. Tsunesada;N. Kanda;H. Nakano;D. Tatsumi;M. Ando;M. Sasaki;H. Tagoshi;H. Takahashi
中科院分区：
文献类型：
--
作者：
Y. Tsunesada;N. Kanda;H. Nakano;D. Tatsumi;M. Ando;M. Sasaki;H. Tagoshi;H. Takahashi
广义相对论可以很好地理解从一个带有衰荡波形的轻微扭曲的黑洞发出的引力辐射。它为直接观测黑洞和确定其物理参数、质量和角动量（克尔参数）提供了一个探针。对于利用引力波探测器的数据进行环域搜索，匹配滤波技术是有用的。本文介绍了利用观测数据进行引力波搜索中匹配滤波分析问题的研究。最重要的是，我们专注于模板构建，匹配或信噪比（SNRs），银河系事件的检测概率，以及波形参数或黑洞毛发评估的准确性。在匹配滤波的模板设计中，搜索参数范围和模板分离由可接受的最大信噪比损失、检测效率和计算成本等要求决定。然而，在使用观测数据的实际搜索中，非平稳噪声的影响会导致信噪比降低，并增加波形参数确定的误差。这些问题将潜在地出现在任何匹配滤波搜索的任何类型的波形。为此，我们对利用蒙特卡罗技术产生的人工衰铃信号进行了匹配滤波分析，并将其注入TAMA300观测数据中。我们采用了一种有效的方法来构建Nakano等人最近提出的一组衰铃滤波器，并使用从标准生成的模板组，使任何信号的信噪比损失不超过2%。通过检查模拟信号的信噪比分布，我们发现该标准在使用TAMA300数据的铃响搜索中得到满足。在TAMA300的灵敏度下，对于质量大于$20{M}_{\ensuremath{\bigodot}}$、$\ mathm {SNR}g10$的黑洞，银河系环灭事件的探测概率约为$50%$。波形参数估计的精度与模板间距一致，黑洞质量和Kerr参数的分辨率分别为$%$和$\ensuremath{\sim}40%$。通过在分层搜索策略中使用细网格模板库，可以将$\ mathm {SNR}\ensuremath{\ge}10$的事件分别提高到$l0.9%$和$l24%$。
Gravitational radiation from a slightly distorted black hole with ringdown waveform is well understood in general relativity. It provides a probe for direct observation of black holes and determination of their physical parameters, masses and angular momenta (Kerr parameters). For ringdown searches using data of gravitational wave detectors, matched filtering technique is useful. In this paper, we describe studies on problems in matched filtering analysis in realistic gravitational wave searches using observational data. Above all, we focus on template constructions, matches or signal-to-noise ratios (SNRs), detection probabilities for Galactic events, and accuracies in evaluation of waveform parameters or black hole hairs. In template design for matched filtering, search parameter ranges and template separations are determined by requirements from acceptable maximum loss of SNRs, detection efficiencies, and computational costs. In realistic searches using observational data, however, effects of nonstationary noises cause decreases of SNRs, and increases of errors in waveform parameter determinations. These problems will potentially arise in any matched filtering searches for any kind of waveforms. To investigate them, we have performed matched filtering analysis for artificial ringdown signals which are generated with Monte-Carlo technique and injected into the TAMA300 observational data. We employed an efficient method to construct a bank of ringdown filters recently proposed by Nakano et al., and use a template bank generated from a criterion such that losses of SNRs of any signals do not exceed $2%$. We found that this criterion is fulfilled in ringdown searches using TAMA300 data, by examining distribution of SNRs of simulated signals. It is also shown that with TAMA300 sensitivity, the detection probability for Galactic ringdown events is about $50%$ for black holes of masses greater than $20{M}_{\ensuremath{\bigodot}}$ with $\mathrm{SNR}g10$. The accuracies in waveform parameter estimations are found to be consistent with the template spacings, and resolutions for black hole masses and the Kerr parameters are evaluated as a few $%$ and $\ensuremath{\sim}40%$, respectively. They can be improved up to $l0.9%$ and $l24%$ for events of $\mathrm{SNR}\ensuremath{\ge}10$ by using fine-meshed template bank in the hierarchical search strategy.