Euclid：提高弱引力透镜剪切标定效率——像素噪声消除与响应方法的检验
Euclid: Improving the efficiency of weak lensing shear bias calibration. Pixel noise cancellation and the response method on trial
AI总结：
本文比较了Euclid弱引力透镜剪切标定中不同噪声消除方法的效率，发现响应方法在随机定位星系的小剪切场景下可将模拟效率提升70倍，像素噪声消除可将运行时间降低7倍，并证明旋转整个场景能增强混叠情况下的形状噪声消除效率。
摘要：
为了从即将开展的弱引力透镜巡天（例如由Euclid进行的巡天）中获得准确的宇宙学推断，剪切测量需要使用星系图像模拟进行标定。我们研究了不同噪声消除方法的效率，这些方法旨在减少达到给定剪切测量精度所需的模拟体量。具体而言，我们比较了具有不同噪声消除的拟合方法以及一种基于响应的方法。我们使用GalSim在网格上以及在更大场景中的随机位置模拟星系。将星系放置在随机位置需要对其进行检测，我们使用SExtractor完成了这一检测。在网格上，我们忽略了检测步骤，因此也忽略了由此可能产生的检测偏差。模拟图像的剪切使用基于快速矩的方法KSB进行测量，我们注意到该方法存在偏离纯线性剪切测量偏差的情况。对于不确定度的估计，我们使用自举法作为一种经验方法。我们发现，在形状噪声消除的基础上研究的每种方法都能进一步提高标定模拟的效率。改进程度取决于所考虑的剪切幅度范围和模拟类型（基于网格或随机位置）。在网格上针对小剪切使用响应方法提供了最大的改进。在更现实的随机定位星系情况下，使用响应方法在小剪切下仍能获得70倍的改进因子。或者，在形状噪声消除的基础上使用像素噪声消除，运行时间已经可以降低7倍。此外，我们证明，在存在混叠的情况下，如果旋转整个场景而不是单个星系，可以提高形状噪声消除的效率。
英文摘要：
To obtain an accurate cosmological inference from upcoming weak lensing surveys such as the one conducted by Euclid, the shear measurement requires calibration using galaxy image simulations. We study the efficiency of different noise cancellation methods that aim at reducing the simulation volume required to reach a given precision in the shear measurement. Explicitly, we compared fit methods with different noise cancellations and a method based on responses. We used GalSim to simulate galaxies both on a grid and at random positions in larger scenes. Placing the galaxies at random positions requires their detection, which we performed with SExtractor. On the grid, we neglected the detection step and, therefore, the potential detection bias arising from it. The shear of the simulated images was measured with the fast moment-based method KSB, for which we note deviations from purely linear shear measurement biases. For the estimation of uncertainties, we used bootstrapping as an empirical method. We find that each method we studied on top of shape noise cancellation can further increase the efficiency of calibration simulations. The improvement depends on the considered shear amplitude range and the type of simulations (grid-based or random positions). The response method on a grid for small shears provides the biggest improvement. In the more realistic case of randomly positioned galaxies, we still find an improvement factor of 70 for small shears using the response method. Alternatively, the runtime can be lowered by a factor of 7 already using pixel noise cancellation on top of shape noise cancellation. Furthermore, we demonstrate that the efficiency of shape noise cancellation can be enhanced in the presence of blending if entire scenes are rotated instead of individual galaxies.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
