(g-2)_mu and alpha(M_Z^2) re-evaluated using new precise data

TL;DR

This work revises Standard Model predictions for the muon g-2 and for hadronic contributions to the running of the QED coupling by incorporating new precise e+e- hadronic cross-section data, including radiative-return results from KLOE and BaBar, and a QCD sum-rule analysis to adjudicate the 1.43–2 GeV region. The authors obtain a muon anomaly of about 11 659 182.8 × 10^-10 with an uncertainty near 4.9 × 10^-10, confirming a ~3.3-sigma deficit relative to the experimental value, while updating Delta alpha hadronic at MZ^2 to 276.26 × 10^-4 with a corresponding alpha(MZ^2)^-1 of 128.944. A key methodological result is the preference for summing exclusive channels below 2 GeV over using inclusive cross sections, aided by new BaBar data and BES pQCD consistency at higher energies. The findings have important implications for electroweak fits and Higgs-mass inferences, and they underscore the need for improved light-by-light estimates and future precision measurements at upcoming facilities.

Abstract

We update our Standard Model predictions for g-2 of the muon and for the hadronic contributions to the running of the QED coupling, Delta alpha_had^(5)(M_Z^2). Particular emphasis is put on recent changes in the hadronic contributions from new data in the 2pi channel and from the energy region just below 2 GeV. In particular, for the e+e- -> pi+pi- contribution we include the recent `radiative return' data from KLOE and BaBar. We also include the recent BaBar data on other exclusive channels. We make a detailed study of the effect of replacing the measurements of the inclusive cross section, sigma(e+e- -> hadrons), by the sum of the exclusive channels in the energy interval 1.43 < sqrt{s} < 2 GeV, which includes a QCD sum-rule analysis of this energy region. Our favoured prediction for the muon anomalous magnetic moment is (g-2)/2 = (11659182.8 +- 4.9)*10^(-10) which is 3.3 sigma below the present world-average measurement. We compare our g-2 value with other recent calculations. Our prediction for the QED coupling, obtained via Delta alpha_had^(5)(M_Z^2) = (276.26 +- 1.38)*10^(-4), is alpha(M_Z^2)^(-1) = 128.944 +- 0.019.

Figures (18)

• Figure 1: Fit with all data in the $2\pi$ channel: light (yellow) band. Radiative return data from BaBar BaBar2pi are shown by the darker (green) band, whereas the KLOE KLOE08KLOE10 data are displayed by the markers as indicated in the plot.
• Figure 2: Same as Fig. \ref{['fig:2pidataall']}, but 'zoomed' in for the $\rho-\omega$ interference region. In addition to the radiative return data, also the important data from CMD-2 CMD2oldCMD2new and SND SNDre are displayed.
• Figure 3: Low energy tail of the $\rho \to 2\pi$ channel: the light (yellow) band shows the result of our fit using all data, whereas the markers display the actual data points as indicated in the plot. The dashed line represents the prediction of chiral perturbation theory (labelled ChPT) used for the lowest energies from threshold to the first BaBar point at $0.305$ GeV.
• Figure 4: Normalised difference between the data sets based on radiative return from KLOE KLOE08KLOE10 and BaBar BaBar2pi and the fit of all data in the $2\pi$ channel, as indicated on the plot. The (lilac) band symmetric around zero represents the error band of the fit given by the diagonal elements of the fit's covariance matrix, with local error inflation as explained in the text, whereas the light (yellow) band indicates the error band of the fit without inflation.
• Figure 5: Dependence of the global $\chi^2_{\rm min}/{\rm d.o.f.}$ (solid red line, left scale), the globally inflated error $\Delta a_{\mu}^{2\pi}\cdot 10^{10}$ (dashed red line, left scale) and the mean value $a_{\mu}^{2\pi}\cdot 10^{10}$ (dotted blue, right scale) on the choice of the cluster size parameter $\delta$. The dash-dotted green line indicates $\Delta a_{\mu}^{2\pi}\cdot 10^{10}$ with local error inflation. Recall $a_{\mu}^{2\pi}$ is the $2\pi$ contribution in the range $0.305 < \sqrt{s} < 2$ GeV.