Validating prior-informed Fisher-matrix analyses against GWTC data
验证基于先验的 Fisher 矩阵分析:与 GWTC 数据的对比
Fisher-matrix methods are widely used to predict how accurately parameters can be estimated. Being computationally efficient, this approach is prompted by the large number of signals simulated in forecast studies for future gravitational-wave detectors, for which adequate analysis tools and computational resources are still unavailable to the scientific community. However, approximating the full likelihood function with a Gaussian may lead to inaccuracies, which we investigate in this work. To assess the accuracy of the Fisher approximation, we compare the results of the Fisher code GWFish against real data from the Gravitational Wave Transient Catalogs provided by the Virgo/LIGO Bayesian analyses. Additionally, we present a sampling algorithm to include priors in GWFish, not only to ensure a fair comparison between GWFish results and the Virgo/LIGO posteriors but also to investigate the role of prior information and to assess the need to include it in standard Fisher analyses. We find that the impact of priors depends mostly on the level of signal-dependent degeneracy of the waveform parametrization, and priors are generally more important when the level of degeneracy is high. Our findings imply that Fisher-matrix methods are a valid tool for Einstein Telescope science-case studies.
展开 ▾首次针对全部公开 GWTC 事件,系统验证了先验信息对 Fisher 估计的作用;发现当参数简并严重时先验能显著改善误差,但无法消除多模态带来的角度参数偏差,为三探测器网络下的可靠性提供了直接证据。
Fisher 矩阵近似因计算高效被广泛用于下一代引力波探测器参数估计的预报 Cutler & Flanagan 1994、Vallisneri 2008,但其高斯似然近似可能导致的偏差一直受到讨论 Rodriguez+ 2013。多个 Fisher 代码如 GWFish Dupletsa+ 2023、GWFast Iacovelli+ 2022 相继开发并通过交叉验证 Branchesi+ 2023,部分工作已尝试加入先验信息约束物理边界 Vallisneri 2008、Iacovelli+ 2022,但缺乏基于真实数据的系统性检验。本文提出一种新方法将先验融入 GWFish,并首次与全部 78 个公开双黑洞事件的 LVK 贝叶斯后验进行系统对比,发现当参数简并较强时先验能显著改善质量、距离和自旋的误差估计,但对存在多模态的角度参数(如天空定位)改善有限。这些结果为未来爱因斯坦望远镜等多探测器网络的科学预研提供了重要校验,支持 Fisher+先验方法在多数科学用例中的可信度,也明确其在定位研究中的适用边界。
预印本 2024-04-24 · 刊出 2025-01-13 · 收录 2026-10-12