Matthias Maurer ; Christoph Schwanda ; Gianluca Inguglia

2023
HEPHY Vienna

Abstract: The flavour changing neutral current b → sl+l− is considered to be one of the most promising particle transitions to search for new physics at the current point of time. It cannot occur directly in the Standard Model, but only through higher order processes, and is therefore highly suppressed. This suppression of the Standard Model amplitude makes the influences on angular distribution and branching ratio predicted by various new physics models easier to measure. Previous experiments, most notably Belle I in 2016 and LHCb with its most recent update in 2020, have reported hints of lepton flavour universality violation at a level of 2.6 σ and 2.5 σ respectively for the angular observable P5′ in the decay B → K∗l+l−. The thesis at hand is part of Belle II’s undertaking of improving these results and thereby either verifying or excluding these tensions. ′ As the full angular analysis necessary for a P5 measurement is out of scope for achieving within a single Master thesis, the present work’s aim is to lay the necessary foundation towards doing so in future projects. Since the over- arching analysis is not complete, no real data has been used in order to avoid the introduction of biases, thereby relying entirely on simulated Monte Carlo data. Nevertheless, the analysis is entirely compatible with real data as well. After the selection and reconstruction of signal candidates from raw data, an extensive analysis of higher order correlations was conducted to avoid biases in later parts of the angular analysis. Background was suppressed using gradient boosted decision trees, and the branching ratios were calculated using signal yields extracted from 1D fits on the beam constrained mass. Last but not least, the stability of the fitter against statistical fluctuations was tested using 10 000 poisson distributed toys. Unfortunately, major delays in data taking have made it very difficult to improve the precision achieved by previous experiments within the foreseeable future, with only ∼ 350fb−1 being available for now in comparison to the 711 fb−1 available to Belle I. While the predicted signal yields and purity for the B+ decays are competitive with the Belle I results, this is nowhere near the case for the B0. As a result, it seems unlikely that an improvement in P5′ precision can be achieved with the current amount of data available. An interesting option to circumvent the current statistical limits might be to use both Belle I and Belle II data sets simultaneously for a combined measurement.

Note: Presented on 19 01 2023
Note: MSc