The dark sirens method combines gravitational waves and catalogs of galaxies to constrain the cosmological expansion history, merger rates and mass distributions of compact objects, and the laws of gravity. However, the incompleteness of galaxy catalogs means faint potential host galaxies are unobserved, and must be modeled to avoid inducing a bias. The majority of dark sirens analyses to date assume that the missing galaxies are distributed uniformly across the sky, which is clearly unphysical. We introduce a new Bayesian approach to the reconstruction of galaxy catalogs, which makes full use of our knowledge of large-scale structure. Our method quantifies the uncertainties on the estimated true galaxy number count in each voxel, and is marginalized over cosmological parameters and bias parameters. Crucially, our method further assesses the (absolute) magnitude distribution of galaxies, which is not known from the galaxy catalog itself. We present the details of our method and validate our approach on a galaxy catalog associated to the Millennium Simulation. The tools developed here generate physically-informed and robust host galaxy reconstructions, enabling more informative dark sirens analyses. Stage IV galaxy surveys will display greater redshift overlap with GW observations, whilst remaining incomplete — emphasizing the importance of our work.

中文翻译：

---

宇宙制图学：由大尺度结构提供的星系密度的贝叶斯重建


暗警报器方法结合了引力波和星系目录，以限制宇宙膨胀历史、致密物体的合并速率和质量分布以及万有引力定律。然而，星系表的不完整意味着无法观察到微弱的潜在宿主星系，必须对其进行建模以避免引起偏差。迄今为止，大多数黑暗警报器分析都假设缺失的星系均匀分布在天空中，这显然是非物理的。我们引入了一种新的贝叶斯方法来重建星系表，该方法充分利用了我们对大尺度结构的了解。我们的方法量化了每个体素中估计的真实星系数的不确定性，并且在宇宙学参数和偏差参数上被边缘化。至关重要的是，我们的方法进一步评估了星系的（绝对）星等分布，这在星系目录本身是未知的。我们展示了我们的方法的细节，并在与千年模拟相关的星系表上验证了我们的方法。这里开发的工具可以生成基于物理信息且强大的宿主星系重建，从而能够进行信息量更大的黑暗警报器分析。第四阶段星系巡天将显示与 GW 观测的更大红移重叠，同时保持不完整——这凸显了我们工作的重要性。