Top Quark Data Analysis

Since top quarks are produced so rarely--perhaps once every few billion proton-anti-proton collisions--studying the top is a bit like examining a haystack to find the properties of a needle. In all of these publications, the plan is similar. Start with a mountain of collision events containing quark jets, leptons, and their energies and momenta; make some gross cuts on the data--such as chosing events with high transverse momentum and a particular number of jets; then analyze the remaining top "candidate" events to come up with the top mass that could have produced them. All of the work depends critically on computer simulations that accurately model the particle interactions and the performance of the detectors, because every top candidate event has a certain probability of coming from "background"--non-top processes that would create the same set of jets and leptons. Simulations must be combined with the data to determine the the number of non-top events that would be expected with a given signature, and a claim of top measurement is only justified if the number of events sufficiently exceeds background expectations.