In a recent measurement, the CMS experiment confirms a slight but persistent disagreement between the simulated and the observed rates of events in which a top quark pair is produced accompanied by a W boson.
One of the goals of the CMS experiment is to find hints of new physics phenomena that are not included in the so-called standard model of particle physics, a comprehensive theoretical description of all elementary particles and their interactions known to date. The first step in finding these beyond-the-standard-model effects is figuring out where the standard model is not able to make correct predictions. Therefore, we measure the properties of the standard model, most notably the probabilities of various interesting processes to occur, and compare them to expectations from simulations based on state-of-the-art theoretical knowledge.
Among these interesting processes is ttW, the production of a top quark-antiquark pair in association with a W boson. The top quark (t) is the heaviest known elementary particle; the W boson is a heavy vector boson, and the only one that carries an electric charge. Their simultaneous production forms an intriguing combination. Most of the time, our measurements agree nicely with the prediction. But not so for ttW! In a number of studies conducted over the past years, a slight but persistent tension has been observed between the measured and the predicted probability for this process.
Above: Left: the particles of the standard model, with the top quark (t) and the W boson highlighted. Right: diagrammatic representation of ttW production in proton-proton collisions.
A sign of new physics, something beyond the standard model? Perhaps, though the currently observed tension is far too small to be conclusive. A somewhat more likely possibility is that the simulation, used to translate the underlying physics into concrete predictions for ttW, is not perfect. Compared to other processes that are similar at first sight (for example ttZ, a top quark pair and a Z boson), ttW is particularly difficult to simulate. The calculations proceed in steps of increasing complexity. For most processes, including only one or two of these steps is sufficient, but for ttW, the contributions of further steps are known to be non-negligible. Including these extra steps in the calculation is very challenging, and this is a very active frontier of research.
Above: The gap between the measured and predicted ttW production rate has decreased over time, as more accurate simulation techniques became available – from “leading order” (LO), the baseline, to approximate “next-to-next-to-leading order” (NNLO), the current state-of-the-art. Yet, even with the current most accurate techniques, a slight but notable tension remains.
To try to get to the bottom of the ttW tension, we now measure the so-called “differential cross-section”, meaning that we measure the ttW production rate not just as a single number, but as a function of several interesting characteristics of the particles produced in the collisions. This provides more information and feedback about potential features that our simulation is not accounting for. For example, it could turn out that the differential cross-section agrees well with the prediction for events where the leptons are produced with little energy, but starts to slightly deviate for events where leptons are produced with a lot of energy. This could give clues on how to improve the simulation. At the same time, we compare our measurements to the latest state-of-the-art prediction.
From the experimental side, one of the special aspects of our study is that we measure the ttW differential cross-section with two different approaches: one that uses a machine-learning technique at its core (with slightly better accuracy) and one that doesn’t (which is easier to interpret and re-use for later studies). Both methods yield the same results (within their experimental uncertainties), and hence also serve as a mutual cross-check that shows the robustness of both strategies.
Above: One of the major results of this new ttW study. The upper panel shows the differential ttW cross-section as a function of the number of jets in the event. The horizontal blue and red lines show two predictions, each with estimated uncertainties indicated by the semi-transparent patches. The markers show our measurements, with corresponding uncertainties indicated by the vertical lines. The “MVA-based” method uses a machine-learning algorithm to identify potential ttW events in data and reject background events, while the “Counting” method does not.
Our results confirm the previously observed tension between the observation and the theoretical predictions. Our results also indicate that the tension increases when ttW production occurs with a large number of additional jets. It is however not at a level of precision yet to completely resolve the mystery surrounding ttW. But not to despair! As the current measurement is limited by the size of the data recorded so far, we will be able to improve the precision with more CMS data. Given the astounding performance of the LHC, we are sure to further unravel the mysteries of ttW production in the future! At the same time, this remains a very active area of research among modelers, and more accurate predictions are likely to emerge in the coming years.
So, what is causing the gap between the predictions and the measurements for ttW? Is it really just a matter of including more complex steps in the simulation? Or is there something more to it? Further research on both the experimental and theoretical front will tell. Stay tuned!
Read more about these results:
-
CMS Physics Analysis Summary (TOP-24-003): " Measurement of the ttW differential cross section and charge asymmetry at 13 TeV "
-
Display of collision events: CERN CDS
-
@CMSExperiment on social media: LinkedIn - Facebook - Twitter - Instagram
- Do you like these briefings and want to get an email notification when there is a new one? Subscribe here