Probing for Invisible Higgs Decays with Global Fits

TL;DR

The work develops and applies global fit techniques to Higgs signal-strength data to probe invisible Higgs decays, framing ${\rm Br}_{inv}$ as a key portal to beyond-SM states. It introduces a global PDF-based approach and demonstrates that current data allow ${\rm Br}_{inv}$ up to about 0.64 at 95% CL for $m_h\approx124$ GeV, with a mild preference for a nonzero Br_inv that is sensitive to correlations with other new physics.Robustness tests show that allowing for non-SM scalar couplings, EWPD constraints, and higher-dimensional operators can remove or weaken the initial Br_inv hint, underscoring degeneracies with modified production/decay dynamics. The paper then outlines direct-search strategies (ZH, VBF, $t\bar t h$) to confirm Br_inv, highlighting that a combined indirect-direct approach will be essential to conclusively establish invisible Higgs decays and map their implications for the Higgs portal.

Abstract

We demonstrate by performing a global fit on Higgs signal strength data that large invisible branching ratios Br_{inv} for a Standard Model (SM) Higgs particle are currently consistent with the experimental hints of a scalar resonance at the mass scale m_h ~ 124 GeV. For this mass scale, we find Br_{inv} < 0.64 (95 % CL) from a global fit to individual channel signal strengths supplied by ATLAS, CMS and the Tevatron collaborations. Novel tests that can be used to improve the prospects of experimentally discovering the existence of a Br_{inv} with future data are proposed. These tests are based on the combination of all visible channel Higgs signal strengths, and allow us to examine the required reduction in experimental and theoretical errors in this data that would allow a more significantly bounded invisible branching ratio to be experimentally supported. We examine in some detail how our conclusions and method are affected when a scalar resonance at this mass scale has couplings deviating from the SM ones.

Figures (15)

• Figure 1: Global fit to the best-fit signal strength parameters in SM Higgs searches as supplied by the four experiments for individual channels (left) and for their combinations (right). These results are based on post Moriond 2012 data (see the Appendix and Table I) when an invisible width is added to the SM as a free parameter. Left: The red (upper) solid curve is for $m_h = 125 \, {\rm GeV}$; the blue (lower) solid curve is for $m_h = 124 \, {\rm GeV}$. The one sigma region defined with the CDF for a one parameter fit is given by the horizontal dashed lines in each case and the best fit points are given by $(m_h, {\rm Br}_{inv}) = (124,0.12),\, (125,0.15)$. Right: The red (lower) solid curve is for $m_h = 125 \, {\rm GeV}$; the blue (upper) solid curve is for $m_h = 124 \, {\rm GeV}$. Now the best fit points are given by $(m_h, {\rm Br}_{inv}) = (124,0.10),\, (125,0.06)$. Comparing these results gives a sense of the effect of neglected correlations amongst the individual signal channels in such fits.
• Figure 2: Illustration of the probability density functions for the background-only (blue curve on the left) and SM (red curve on the right) and corresponding $p$-values for a hypothetical $\hat{\mu}_{c}=0.6$.
• Figure 3: Left: $p$-values for SM and background-only hypotheses (negative and positive-slope lines respectively) vs ${\rm Br}_{inv}$ for several values of the $1 \sigma$ error $\sigma_c$ on $\hat{\mu}$: its current value $\sigma_c=0.3$ (solid); half of it, $\sigma_c=0.15$, expected to be reached at the end of this year (dashed); and a future value $\sigma_c=0.05$ (dotted). The red dashed horizontal lines show the $p$-values corresponding to significances from $1 \, \sigma$ to $5 \sigma$'s. Right: $\sigma_c$ (plotted as $1/\sigma_c^2$) required to pinpoint a non-zero ${\rm Br}_{inv}$ with a significance from 1 to 5 $\sigma$ (curves from lower gray to upper red). The precise condition imposed is that, for a given $\hat{\mu}_{c}=1- {\rm Br}_{inv}$, both $p_{back}$ and $p_{SM}$ (with ${\rm Br}_{inv}=0$) are smaller than those corresponding to fluctuations from $1 \sigma$ to $5 \sigma$. The horizontal dashed lines show again the values $\sigma_c\simeq 0.3, 0.15$ and $0.05$.
• Figure 4: 95% CL exclusion limits for ${\rm Br}_{inv}$ as a function of the observed $\hat{\mu}_c$ for several values of its error $\sigma_c$: the current one ($\sigma_c=0.3$); the combined error that is estimated to be reached by both ATLAS and CMS at the end of the year ($\sigma_c=0.15$), with the 95% CL excluded region shaded; and with a future $\sigma_c=0.05$. A red vertical solid line indicates the current value of ${\rm Br}_{inv}$ obtained for $m_h = 124 \, {\rm GeV}$. The $95\%$ CL limit for $m_h = 124 \, {\rm GeV}$ obtained directly from the $\chi^2$ fit ($< 0.64$) is consistent with the PDF test results shown.
• Figure 5: Left: Current status of the experimental situation concerning ${\rm Br}_{inv}$, extracted from combining the $\hat{\mu}_c$ values reported by ATLAS, CMS and the Tevatron, and interpreting deviations from $\hat{\mu}_c=1$ as coming from an invisible Higgs width. Above the lower dashed line $p_{SM}<p_{2\sigma}$; below the upper dashed line, $p_{back}<p_{2\sigma}$, so that no strong evidence for a nonzero value of ${\rm Br}_{inv}$ is possible at this time. Right: Same as left, in a hypothetical future situation (the solid curve is obtained from the left figure data series by shifting the data series to larger values of ${\rm Br}_{inv}$ and reducing the error) assuming a factor $5$ improvement in the precision with which the combined $\hat{\mu}_c$ could be measured compared to current data. The dashed lines correspond now to p-values equal to $p_{5\sigma}$, so that finding $5\sigma$ evidence for a nonzero ${\rm Br}_{inv}$ would be possible in the white region between both lines.