Global analysis of $b\to s\ell\ell$ anomalies

TL;DR

This paper analyzes the $b o s\,\ell\ell$ anomalies, confirming previous New Physics hints with updated data, and explores various scenarios with high significance, emphasizing the need for more data to distinguish among them.

Contribution

It provides a comprehensive analysis of recent $b\to s\ell\ell$ anomalies, confirming the significance of a specific New Physics contribution and exploring a wider range of scenarios with detailed fits.

Findings

Confirmed the $\mathcal{C}_9^{\rm NP} \simeq -1$ solution

Identified multiple high-significance New Physics scenarios

Highlighted the importance of $b\to s e^+ e^-$ data

Abstract

We present a detailed discussion of the current theoretical and experimental situation of the anomaly in the angular distribution of $B \to K^*(\to K\pi)\mu^+\mu^-$, observed at LHCb in the $1~\text{fb}^{-1}$ dataset and recently confirmed by the $3~\text{fb}^{-1}$ dataset. The impact of this data and other recent measurements on $b\to s\ell^+\ell^-$ transitions ($\ell=e,\mu$) is considered. We review the observables of interest, focusing on their theoretical uncertainties and their sensitivity to New Physics, based on an analysis employing the QCD factorisation approach including several sources of hadronic uncertainties (form factors, power corrections, charm-loop effects). We perform fits to New Physics contributions including experimental and theoretical correlations. The solution that we proposed in 2013 to solve the $B\to K^*\mu^+\mu^-$ anomaly, with a contribution ${\mathcal C}_9^{\rm NP}\simeq -1$, is confirmed and reinforced. A wider range of New-Physics scenarios with high significances (between 4 and 5 $\sigma$) emerges from the fit, some of them being particularly relevant for model building. More data is needed to discriminate among them conclusively. The inclusion of $b\to s e^+ e^-$ observables increases the significance of the favoured scenarios under the hypothesis of New Physics breaking lepton flavour universality. Several tests illustrate the robustness of our conclusions.