利用CMS開放數據分析輕子型衰變Z玻色子附近的多事例對撞堆積效應與對撞潛在事件之強子分佈

Abstract

為了測量大型強子對撞機（LHC）中的對撞潛在事件（underlying event）和多事例對撞堆積效應（pileup），本論文使用歐洲原子核研究中心開放數據平台（CERN open data portal）釋出的能量8 TeV之質子質子對撞緊湊緲子線圈偵測器（CMS）開放數據進行研究。輕子型衰變Z玻色子事件中幾乎所有重子都來自於多事例對撞堆積效應，因此被選為適當的分析途徑，而這之中，此論文主要聚焦於輕子型衰變Z玻色子週遭的重子特性。此外，多事例對撞堆積效應削除工具——SoftKiller及PUPPI——的表現，也將被檢驗，並藉由開放數據平台釋出的模擬樣本進行蒙地卡羅匹配分析策略（Monte Carlo matching），將其結果進行比較。PUPPI為兩者中在此輕子型途徑下運作較為良好者，因此使用他進行多事例對撞效應提取（extracting pileup）。除了測量觀測量粒子重複度（multiplicity）、橫動量（transverse momentum）及不變質量（invariant mass）以外，一個新的紅外與共線安全（infrared and collinear safety）觀測量——衰退質量（dropped mass）也因為其對於多事例對撞堆積效應的敏感性與被強子化現象（hadronization）影響較輕微，而被使用在此論文的主要分析中。最後，多事例對撞堆積效應在輕子型衰變Z玻色子周遭空間的均勻分布被確認，而對撞潛在事件則被確認對於空間沒有相關性。另一方面，多事例對撞堆積效應與多事例碰撞數目（number of vertex）之間的線性關係被成功檢驗，同時對撞潛在事件也被驗證與多事例碰撞數目無關。

To probe underlying events and pileup contributions at the LHC, 8 TeV pp collision CMS open data released on the CERN open data portal is used. Leptonic Z bosons are selected as an optimal channel, that most of the hadrons in the events are from pileup. Distributions of hadrons around leptonic Z bosons are focused on. The performance of pileup mitigation tools, SoftKiller and PUPPI, are examined with the Monte Carlo matching method by using simulation samples released on the Open Data portal. PUPPI works better and is used to extract pileup contributions. Several observables such as particle multiplicity, transverse momentum, invariant mass, as well as a new infrared and collinear safety observable, dropped mass are utilized in the main analysis. The dropped mass is sensitive to pileup but less affected by hadronization than multiplicity. Uniform distribution of pileup in the space around leptonic Z bosons is checked as well as underlying events is confirmed to have no dependence on the space. The linearity between pileup and number of vertices is presented, and the independence between underlying events and number of vertices is also confirmed.