用于具有非平稳噪声的引力波旋近信号的紧凑时域降阶求积
A compact time-domain reduced order quadrature for gravitational wave inspirals with non-stationary noise
浏览论文内容
中文总结 AI 辅助
针对非平稳噪声下引力波旋近信号，提出首个时域降阶求积，通过频率分区减半存储权重，保持高精度，使时域搜索可行。
中文摘要 AI 辅助
我们开发了首个用于早期旋近阶段引力波的时域降阶求积方法，其动机是在存在数据间隙、不均匀采样和稠密协方差矩阵的情况下，对数据集进行压缩的需求。一个包含1290个元素的单一降阶基可以表示来自超大质量黑洞双星源、啁啾质量在$10^7-10^{10}$ $M_\odot$之间的旋近信号，但在存在非平稳噪声的情况下，必须存储约$1290^2$个降阶求积权重。我们通过按引力波频率划分参数空间，能够将存储的权重减少一半，划分后波形/似然度的相对精度仍保持在$10^{-5}$/$10^{-3}$以下。最后，我们指出外差法可以与降阶方法结合使用，且不损失精度。我们的降阶求积方法使得时域旋近搜索对于脉冲星计时阵列和相对恒星天体测量均可行。
英文摘要
We develop the first time-domain reduced order quadrature for gravitational waves in the early inspiral period, motivated by the need for dataset compression in the presence of gaps, uneven sampling, and dense covariance matrices. A single reduced basis with 1290 elements can represent inspirals from supermassive black hole binary sources with chirp masses between $10^7-10^{10}$ $M_\odot$, but in the presence of non-stationary noise $\sim 1290^2$ reduced order quadrature weights must be stored. We are able to reduce the stored weights by a factor of two by partitioning our parameter space by GW frequency, with relative waveform/likelihood accuracies remaining below $10^{-5}$/$10^{-3}$ after partition. Finally, we point out that heterodyning can be combined with reduced order methods with no loss in accuracy. Our reduced order quadrature makes inspiral searches in the time domain feasible for both pulsar timing arrays and relative stellar astrometry.
发表机构
- University of Southern California（南加州大学）
- Institute for Gravitational Wave Astronomy and School of Physics and Astronomy, University of Birmingham（伯明翰大学引力波天文研究所与物理天文学院）
机构由 AI 辅助整理，请以论文原文为准。
