标题： 基于GPU加速的宇宙学和星系的引力透镜与动力学（GLaD）建模 显示英文标题

标题： GPU-Accelerated Gravitational Lensing & Dynamical (GLaD) Modeling for Cosmology and Galaxies

摘要： 时滞距离测量来自强透镜类星体，提供了一种稳健的独立方法来确定哈勃常数（$H_0$）。这种方法交叉验证了晚期宇宙中距离阶梯的$H_0$估计值和早期宇宙中的宇宙微波背景。然而，透镜模型中的质量板退化引入了系统不确定性，限制了精度。动力学建模通过使用独立观测数据约束质量分布来补充强透镜。我们开发了一种方法和软件框架，用于联合建模恒星运动学和透镜数据。使用透镜类星体RXJ1131$-$1131的模拟数据，我们证明高质量的运动学数据可以实现$\sim$4%的精度对$H_0$。通过广泛的建模，我们研究了透镜星系中存在超大质量黑洞以及运动学数据中的潜在系统偏差对$H_0$测量的影响。我们的结果表明，对黑洞质量和轨道各向异性施加先验条件，或排除中心运动学区间，可以减轻$H_0$估计中的偏差。通过对具有系统偏差的模拟运动学数据进行测试，我们强调了对运动学系统误差进行亚百分比控制的必要性，这可以通过当前技术实现。此外，我们利用GPU并行化加速贝叶斯推断，将之前需要一个月的过程缩短了一个数量级。这个流程为利用大数据集推进宇宙学和星系演化研究提供了巨大潜力。 显示英文摘要

摘要： Time-delay distance measurements from strongly lensed quasars provide a robust, independent method for determining the Hubble constant ($H_0$). This approach cross-checks $H_0$ estimates from the distance ladder in the late universe and the cosmic microwave background in the early universe. However, the mass-sheet degeneracy in lensing models introduces systematic uncertainty, limiting precision. Dynamical modeling complements strong lensing by constraining the mass distribution with independent observational data. We develop a methodology and software framework for joint modeling of stellar kinematics and lensing data. Using simulated data for the lensed quasar RXJ1131$-$1131, we demonstrate that high-quality kinematic data can achieve $\sim$4% precision on $H_0$. Through extensive modeling, we examine the impact of the presence of a supermassive black hole in the lens galaxy and potential systematic biases in kinematic data on $H_0$ measurements. Our results show that imposing priors on black hole mass and orbital anisotropy, or excluding central kinematic bins, mitigates biases in $H_0$ estimates. By testing on mock kinematic data with systematic biases, we highlight the need for sub-percent control of kinematic systematics, which is achievable with current technology. Additionally, we leverage GPU parallelization to accelerate Bayesian inference, reducing a previously month-long process by an order of magnitude. This pipeline offers significant potential for advancing cosmological and galaxy evolution studies with large datasets.