How to handle calibration uncertainties in high-energy astrophysics
How to handle calibration uncertainties in high-energy astrophysics
复制标题
如何处理高能天体物理学中的校准不确定性
DOI：
--
复制
发表时间：
2008
期刊：
影响因子：
--
通讯作者：
Taeyoung Park
中科院分区：
文献类型：
--
作者：
V. Kashyap;Hyunsook Lee;A. Siemiginowska;J. McDowell;A. Rots;J. Drake;P. Ratzlaff;A. Zezas;R. Izem;A. Connors;D. V. van Dyk;Taeyoung Park
统计误差的重要性在天文应用中已得到充分证实，而与此不同的是，仪器校准中的不确定性通常被忽略。尽管人们广泛认识到校准中的不确定性可能会导致巨大的系统误差，但尚未开发出可靠且有原则的方法来解释这些误差，因此没有任何机制可以将它们纳入标准天文数据分析中。在这里，我们提出了一个框架，可以对它们进行编码，以便将它们纳入分析范围。我们描述了这个基于改进的 MCMC 算法的框架，并提出了一种格式标准，该格式标准源自 Chandra 上 ACIS-S 探测器的有效面积测量经验，可应用于任何编码系统误差的仪器或方法。然后可以将校准不确定性传播到模型参数估计中，以产生包括系统误差信息的误差线。
Unlike statistical errors, whose importance has been well established in astronomical applications, uncertainties in instrument calibration are generally ignored. Despite wide recognition that uncertainties in calibration can cause large systematic errors, robust and principled methods to account for them have not been developed, and consequently there is no mechanism by which they can be incorporated into standard astronomical data analysis. Here we present a framework where they can be encoded such that they can be brought within the scope of analysis. We describe this framework, which is based on a modified MCMC algorithm, and propose a format standard derived from experience with effective area measurements of the ACIS-S detector on Chandra that can be applied to any instrument or method of codifying systematic errors. Calibration uncertainties can then be propagated into model parameter estimates to produce error bars that include systematic error information.