The couplings of the Higgs boson and its CP properties from fits of the signal strengths and their ratios at the 7+8 TeV LHC

Abstract

Using the full set of the LHC Higgs data from the runs at 7 and 8 TeV center of mass energies that have been released by the ATLAS and CMS collaborations, we determine the couplings of the Higgs particle to fermions and gauge bosons as well as its parity or CP composition. We consider ratios of production cross sections times decay branching fractions in which the theoretical (and some experimental) uncertainties as well as as some ambiguities from new physics cancel out. A fit of both the signal strengths in the various search channels that have been conducted, H -> Z Z, W W, gamma gamma, tau tau and b b, and their ratios shows that the observed ~126 GeV particle has couplings to fermions and gauge bosons that are Standard Model-like already at the 68% confidence level (CL). From the signal strengths in which the theoretical uncertainty is taken to be a bias, the particle is shown to be at most 68% CP-odd at the 99%CL and the possibility that it is a pure pseudoscalar state is excluded at the 4 sigma level when including both the experimental and theoretical uncertainties. The signal strengths also measure the invisible Higgs decay width which, with the same type of uncertainty analysis, is shown to be Gamma_H^inv / Gamma_H^SM < 0.52 at the 68%CL.

Figures (7)

• Figure 1: Left plot Best-fit regions at $68.27\%{\rm CL}$ (green), $95.45\%{\rm CL}$ (yellow) and $99.73\%{\rm CL}$ (grey) in the plane $c_f$ versus $c_V$, based on the $\chi^2$ function of eq. (\ref{['eq:Chi2def']}); the best-fit location is indicated by a (black) cross. The 'concentric' best-fit domains at the same CLs obtained also from $\chi^2$ but for the two extreme theoretical predictions -- upper (plain contours) and lower (dashed contours) -- of the Higgs signal strengths, are presented. The SM (red) point at $c_f=c_V=1$ is also shown. Right plot Best-fit domains at the $68.27\%{\rm CL}$ (green), $95.45\%{\rm CL}$ (yellow) and $99.73\%{\rm CL}$ (light grey) based on the $\chi^2$ function; these domains were obtained by varying continuously the Higgs signal strengths from their lowest to highest theoretical predictions.
• Figure 2: Left: Best-fit regions at $68.27\%{\rm CL}$ (green), $95.45\%{\rm CL}$ (yellow) and $99.73\%{\rm CL}$ (grey) in the plane $c_f$ versus $c_V$, based on the $\chi_R^2$ function. The best-fit (dotted) contours obtained from the $\chi^2$ function in case of a theoretical error added in quadrature (as in Fig. \ref{['fig:mu']}) are superimposed (in red). The associated best-fit point (cross) and SM (red) point are also shown. Right: Same plot as the left one but the best-fit domains from the $\chi^2$ analysis are now derived for the two extreme theoretical predictions of the signal strengths (as in Fig. \ref{['fig:mu']}).
• Figure 3: Best-fit regions at $68.27\%{\rm CL}$ (green), $95.45\%{\rm CL}$ (yellow) and $99.73\%{\rm CL}$ (grey) in the plane $c_t$ versus $c_V$ as obtained from a three-dimensional fit (whose best-fit point is the black cross on the central plot) of the 3 free parameters, $c_f$, $c_t$, $c_V$, based on the $\chi^2_R$ function. The three two-dimensional plots correspond to the slices of these three-dimensional domains at $c_f=0.52$, $1.14$ and $1.55$. Superimposed are the best-fit domains at $68.27\%{\rm CL}$, $95.45\%{\rm CL}$, $99.73\%{\rm CL}$ obtained from $\chi^2$ for the two theoretical signal strength predictions -- upper (red plain) and lower (red dashed contours).
• Figure 4: Best-fit regions at $68.27\%{\rm CL}$ (green), $95.45\%{\rm CL}$ (yellow) and $99.73\%{\rm CL}$ (grey) in the plane $c_f$ versus $c_V$, based on the $\chi_R^2$ function and including hypothetical data from the $14$ TeV LHC with ${\cal L} =300$ fb$^{-1}$ [left plot] or $3000$ fb$^{-1}$ [right plot]. The best-fit $\Delta \chi^2$ contours at $95.45\%{\rm CL}$ and $99.73\%{\rm CL}$ obtained in the same conditions, for the two extreme theoretical predictions of signal strengths (red plain and dashed ellipses), are superimposed; the $68.27\%{\rm CL}$ domain presented (in red) was obtained by varying continuously the signal strengths from their lowest to highest theoretical predictions. So typically the length of this domain indicates the theoretical uncertainty and its width the experimental error. The exactly symmetric domains, obtained via $c_f \to -c_f$, $c_V \to -c_V$, are not shown.
• Figure 5: Best-fit regions at $68.27\%{\rm CL}$ (green), $95.45\%{\rm CL}$ (yellow) and $99.73\%{\rm CL}$ (grey) in the plane $c_f$ versus $c_V$, based on the $\chi_R^2$ function and including hypothetical data from the $14$ TeV LHC with ${\cal L} =3000$ fb$^{-1}$ [as in Fig. \ref{['fig:ratioFUT']}]. The best-fit $\Delta \chi^2$ contours at $68.27\%{\rm CL}$, $95.45\%{\rm CL}$, $99.73\%{\rm CL}$ obtained in the same conditions, and with the theoretical error added in quadrature, are superimposed as dotted (red) contours; the best central point is indicated as a (black) cross.