Advertisement

Electroweak Higgs production with HiggsPO at NLO QCD

  • Admir Greljo
  • Gino Isidori
  • Jonas M. Lindert
  • David Marzocca
  • Hantian Zhang
Open Access
Special Article - Tools for Experiment and Theory

Abstract

We present the HiggsPO UFO model for Monte Carlo event generation of electroweak VH and VBF Higgs production processes at NLO in QCD in the formalism of Higgs pseudo-observables (PO). We illustrate the use of this tool by studying the QCD corrections, matched to a parton shower, for several benchmark points in the Higgs PO parameter space. We find that, while being sizable and thus important to be considered in realistic experimental analyses, the QCD higher-order corrections largely factorize. As an additional finding, based on the NLO results, we advocate to consider 2D distributions of the two-jet azimuthal-angle difference and the leading jet \(p_T\) for new physics searches in VBF Higgs production. The HiggsPO UFO model is publicly available.

1 Introduction

The framework of Higgs pseudo-observables (PO) allows to describe in great generality Higgs production [1] and decay [2] processes in terms of few parameters defined from the properties of the relevant on-shell physical amplitude.1 In particular, \(h \rightarrow 4 f\) decays and electroweak Higgs production processes, associated \(p p \rightarrow Z h\) and \(p p \rightarrow Wh\) production (VH) as well as vector boson fusion (VBF) production \(p p \rightarrow h j j\), can be completely characterized by a set of on-shell correlation functions \(\langle 0 | \mathscr {T} \{ J_f^\mu (x), J_{f^\prime }^\nu (y), h(0) \} | 0 \rangle \), where \(J_{f, f^\prime }^\mu \) are the corresponding fermionic currents for each process. Higgs PO are defined directly from the residues of physical poles in these amplitudes.

The Higgs PO decomposition describes the short-distance contribution to the process and, in particular, encompasses any possible deviation due to some heavy new physics. In order to compare with experimental results it is also important to include long-distance contributions due to soft and collinear radiation. In the case of Higgs decays to four leptons the main effect is due to QED radiation. As it has been shown in Ref. [4], to a very good approximation this effect can be accommodated for in terms of a universal (i.e. independent of the short-distance dynamics) radiator function or, equivalently, by QED shower algorithms. Instead, in the case of electroweak Higgs production at the LHC, where quark currents are involved, QCD corrections are expected to yield the dominant effect. The issue of higher order radiative corrections has been extensively studied in the SM [3, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23], and in the context of effective field theory approaches [24, 25, 26]. The purpose of the present paper is to illustrate how this effect can be accounted for within the PO formalism.

In order to utilise the Higgs PO formalism in experimental analyses, it is important to embed it into an appropriate Monte Carlo event generator framework. The final goal is the possibility to simulate the electroweak Higgs production processes in a realistic environment, for arbitrary values of the Higgs PO, and with a theoretical precision exceeding the experimental one. Such an implementation was initiated in [1] with the realisation of the universal FeynRules (UFO) model, called HiggsPO, which can be used for example within the MadGraph5_aMC@NLO  [27] or Sherpa  [28] event generators at leading order (LO) in QCD. In this paper we present the upgrade able to generate electroweak Higgs production events at next-to-leading order (NLO) accuracy in QCD matched to parton showers [29]. We present numerical results obtained within MadGraph5_aMC@NLO. A corresponding interface within Sherpa+OpenLoops is under development. The HiggsPO NLO UFO model is publicly available at http://www.physik.uzh.ch/data/HiggsPO.

The paper is organized as follows. In Sect. 2 we briefly review the Higgs PO framework for electroweak Higgs production, its dressing with NLO QCD corrections and the corresponding implementation within a UFO model. In Sect. 3 we illustrate the effect of the NLO QCD corrections in a few relevant kinematical distributions for several benchmark points for both VH and VBF production. In Sect. 4 we use this tool to obtain the expression for the simplified template cross-section bins for Zh production as a function of the Higgs PO at NLO in QCD. We finally conclude in Sect. 5.

2 Theoretical considerations and UFO implementation

2.1 Effective PO Lagrangian

The PO decomposition for electroweak Higgs production is based on a momentum expansion of the relevant on-shell amplitudes around the physical poles corresponding to the propagation of SM gauge bosons. In particular, the Higgs PO are defined from the residues of such poles [1, 2]. The widely-used UFO formalism to interface model information with Monte Carlo event generators, instead, relies on amplitude generation starting from a set of interaction Feynman rules, derived from a Lagrangian. While the PO are strictly defined from gauge-invariant analytic properties of on-shell scattering amplitudes, the resulting amplitude decomposition can effectively be reproduced by the tree-level matrix elements derived from the following effective Lagrangian [1, 2],
$$\begin{aligned} \mathscr {L}^\mathrm{eff}_\mathrm{HPO}= & {} {\kappa _{ZZ} }\frac{m_Z^2}{v} Z_\mu Z^\mu h + {\kappa _{WW} }\frac{2 m_W^2}{v} W_\mu ^+ W^{- \, \mu } h \nonumber \\&- {\varepsilon _{\gamma \gamma }} \frac{h}{ 2v} A_{\mu \nu } A^{\mu \nu } - {\varepsilon _{Z\gamma } }\frac{h}{v} Z_{\mu \nu } A^{\mu \nu } - {\varepsilon _{ZZ}} \frac{h}{2v} Z_{\mu \nu } Z^{\mu \nu } \nonumber \\&- {\varepsilon _{\gamma \gamma }^\mathrm{CP}} \frac{h}{2v} A_{\mu \nu } \widetilde{A}^{\mu \nu } - {\varepsilon _{Z\gamma }^\mathrm{CP}} \frac{h}{v} Z_{\mu \nu } \widetilde{A}^{\mu \nu } - {\varepsilon _{ZZ}^\mathrm{CP}} \frac{h}{2v} Z_{\mu \nu } \widetilde{Z}^{\mu \nu }\nonumber \\&- {\varepsilon _{WW}} \frac{h}{v} W^+_{\mu \nu } W^{- \, \mu \nu } - {\varepsilon _{WW}^\mathrm{CP}} \frac{h}{v} W^+_{\mu \nu } \widetilde{W}^{- \, \mu \nu } \nonumber \\&+ \sum _{f} \sum _i {\varepsilon _{Z f^i } }~ \frac{2 h}{v} Z_\mu \bar{f}^i \gamma ^\mu f^i \nonumber \\&+ \sum _{i, j} \frac{2 h}{v} \left[ {\varepsilon _{W e^{i}\nu ^{j}} } ~ W_\mu ^+ \bar{\nu }_{e^i L} \gamma ^\mu e^j_L\right. \nonumber \\&\left. +{\varepsilon _{W u^i_L d^j_L}} ~ W_\mu ^+ \bar{u}^i_{L} \gamma ^\mu d^j_L + { \varepsilon _{W u^i_R d^j_R} }~ W_\mu ^+ \bar{u}^i_{R} \gamma ^\mu d^j_R + \mathrm{h.c.} \right] ,\nonumber \\ \end{aligned}$$
(1)
where \(f = \{e_L, e_R, \nu _L, u_L, u_R, d_L, d_R\}\) and \(i,j= \{1, 2, 3\}\) are flavour indices. Moreover \(V_{\mu \nu } = \partial _\mu V_\nu - \partial _\nu V_\mu \) and \(\widetilde{V}^{\mu \nu } \!=\! \frac{1}{2} \varepsilon ^{\mu \nu \rho \sigma } V_{\rho \sigma }\). The coefficients \(\kappa _{VV}, \varepsilon _{VV'}, \varepsilon ^\mathrm{CP}_{VV'}, \varepsilon _{Z,f^i}, \varepsilon _{W,f^if^j}\) are considered as the proper Higgs pseudo-observables. The flavour-universal PO (\(\kappa _{VV}\), \(\varepsilon _{VV'}\), and \(\varepsilon ^\mathrm{CP}_{VV'}\)) and the leptonic contact terms (\(\varepsilon _{Z,e^i}\), \(\varepsilon _{Z,\nu ^i}\), and \(\varepsilon _{W,e^i\nu ^j}\)), will most likely be strongly constrained, or measured, from \(H\rightarrow 4\ell , 2\ell 2\nu \) decays. The remaining contact terms for light quarks can instead be probed in electroweak Higgs production processes [1]. The charged-current PO \(\varepsilon _{W,f_if_j}\) are complex, while all the others are real (in the limit where we neglect re-scattering effects due to light-quark loops). We introduce the complex phase as:
$$\begin{aligned} \varepsilon _{W,f_if_j} \rightarrow \varepsilon _{W,f_if_j} e^{i \phi _{W,f_i f_j}}~, \end{aligned}$$
(2)
where now \(\varepsilon _{W,f_if_j}\) is taken real. Assuming the Higgs to be a parity-even state and CP to be conserved, then all couplings are real and all the \(\varepsilon ^\mathrm{CP}_{VV'}\) vanish. This Lagrangian, together with the following parametrization of \(Vf\bar{f}\) interactions of the W and Z bosons with fermions,
$$\begin{aligned} \begin{aligned} \mathscr {L}^\mathrm{eff}_{Vff}&= \sum _{f} \sum _i {g_Z^{f_i}} ~ Z_\mu \bar{f}^i \gamma ^\mu f^i \\&\quad + \sum _{i,j} \left[ {(g_{W}^{e_L})_{ij}} ~W_\mu ^+ \bar{\nu }_{e^i L} \gamma ^\mu e^j_L + {(g_{W}^{q_L})_{ij}} ~ W_\mu ^+ \bar{u}^i_{L} \gamma ^\mu d^j_L\right. \\&\quad \left. + {(g_{W}^{q_R})_{ij}} ~ W_\mu ^+ \bar{u}^i_{R} \gamma ^\mu d^j_R + \mathrm{h.c.} \right] ~, \end{aligned} \end{aligned}$$
(3)
and combined with the corresponding gauge boson kinetic terms (based on an unbroken \(SU(3)_\mathrm{QCD} \times U(1)_\mathrm{QED}\)) and ghost terms for QCD, is implemented in a FeynRules (version 2.3.24) [30] model. All SM particle fields are defined in the mass (unitary) eigenbasis with SM QED and QCD gauge interactions. The masses for W, Z, h, and third generation charged fermions are put by hand, while the first two generations and neutrinos are kept massless. The couplings in Eq. (1) are defined with interaction-order label HPO. The corresponding leading-order UFO model generated with FeynRules is publicly available [29] since the publication of Ref. [1].
Under the assumption of an \(U(2)^3\) flavour symmetry acting on the first two generations of quarks, the number of independent light-quark PO reduces to six [1], namely
$$\begin{aligned} \varepsilon _{Z u_L},~\varepsilon _{Z u_R},~ \varepsilon _{Z d_L},~ \varepsilon _{Z d_R},~ \varepsilon _{W u_L}~, \phi _{W u_L}~, \end{aligned}$$
(4)
or five if we further neglect CP-violating contributions (in which case the phase \(\phi _{W u_L}\) vanishes). These six PO are directly accessible in the UFO model (and thus for example in the MadGraph5_aMC@NLO  param_card.dat) as:
$$\begin{aligned} \mathtt{eZuL},~\mathtt{eZuR},~\mathtt{eZdL},~\mathtt{eZdR},~\mathtt{eWuL},~\mathtt{phiWuL}\,. \end{aligned}$$
(5)
This \(U(2)^3\) symmetry assumption can be justified by flavour constraints in the light quark sector, as well as by the experimental difficulties of differentiating the light quark flavour. As demonstrated in Ref. [1], it will be instead possible to separately constrain the PO appearing in Eq. (4) by studying in detail VH and VBF Higgs production.

Besides the six quark contact-terms PO listed above, all flavour-universal PO, as well as the leptonic contact terms, the Zff and \(Wff'\) vertices, are also implemented. This allows the HiggsPO model to generate any electroweak Higgs production process, as well as \(h \rightarrow 4f, ~ 2f\gamma ,~Z\gamma ,\) and \(\gamma \gamma \) decays in the Higgs PO framework at NLO in QCD.

2.2 Dressing the Higgs PO with NLO QCD corrections

The Higgs PO Lagrangian in Eq. (1) can be considered as an extension of the EW sector of the SM without new additional states. Therefore, the dressing with NLO QCD effects and the matching with QCD patron showers is straightforward based on modern Monte Carlo frameworks, that include the automated generation of the required born, one-loop and real matrix elements, and subtraction of infrared singularities. In fact, the HPO effective couplings in Eq. (1) do not require UV renormalization at NLO QCD. Still, the inclusion of the quark contact-term PO at NLO QCD requires dedicated so-called \(R_2\) rational terms.2 The contact-term \(R_2\) can easily be derived from the \(R_2\) contributions of the related \(Vf_i\bar{f}_j\) interactions in the SM [34]:
$$\begin{aligned} \begin{aligned} R_2^{\bar{f}_i f_i Z h}&= -\frac{i g_s^2}{3 \pi ^2 v} \varepsilon _{Z,f^i}~, \\ R_2^{\bar{u}_i d_j W^+ h}&= -\frac{i g_s^2}{3 \pi ^2 v} \varepsilon _{W,u_L^i d_L^j} e^{-i \phi _{W u}}~. \end{aligned} \end{aligned}$$
(6)
Technically, we employ the NLOCT package (version 1.02)  [35] together with FeynArts (version 3.9) [36] to generate all UV and \(R_2\) QCD counterterms for the SM interactions. The resulting model is exported in the UFO format. Since the present public version of the NLOCT package is restricted to renormalizable interactions, we supplement the \(R_2\) rational terms related to the contact-term PO shown in Eq. (6) by hand. As already mentioned, UV counterterms for any of the PO interactions are not needed.

The resulting NLO UFO model can directly be imported into MadGraph5_aMC@NLO, where all required one-loop amplitudes are automatically generated with MadLoop [37] and Ninja  [38, 39]. Infrared subtraction of the real contributions is automatically performed á la FKS [40] in MadFKS [41], where the corresponding real amplitudes are generated from the underlying UFO model. For the VBF Higgs production process the colour suppressed and thus numerically very small pentabox contribution in the virtual amplitude can either explicitly be included or excluded (default in MadGraph5_aMC@NLO). The letter case results in a formal mismatch of the IR poles of the virtual and real contributions, which can be dealt with as described in Sect. 3.2. At the same order of perturbation theory as VBF Higgs production also Higgsstrahlung with hadronic decays, i.e. \(pp\rightarrow (V\rightarrow q\bar{q})H\), contributes to the same \(H+2\) jet signature. Once VBF selection cuts (large invariant masses of the leading jets and/or large rapidity separation of the leading jets) are applied, these contributions (and their interference with VBF topologies) are very small. For simplicity in the simulations presented in Sect. 3.2 we disabled those contributions. In case the Higgsstrahlung subprocess is not considered as a dedicated background in a VBF analysis, it should be included in the VBF process, resulting in additional PO contributions that are automatically generated. We employ the described implementation in MadGraph5_aMC@NLO for the numerical predictions of electroweak Higgs production processes (VBF and VH) at NLO QCD matched to Pythia 6 [42] as presented in the following section. Even though we do not use it in this work, we expect the re-weighting technique at NLO accuracy described in Ref. [43] to be completely compatible with the HiggsPO model.

As a first validation of the code we compare results for the total decay widths \(h \rightarrow V \bar{f} f'\) obtained with the HiggsPO UFO model in MadGraph5_aMC@NLO at LO and NLO against a LO analytical computation of the decay width in the PO formalism. Moreover, the ratio K of the NLO result to the LO one is expected to be a simple universal factor \(K = 1 + \alpha _s / \pi \simeq 1.038\). For all final states and combinations of contact terms we find perfect agreement between the analytical and Monte Carlo results within the numerical uncertainties.

As a further cross check and for future applications within other Monte Carlo frameworks, we implemented the HPO Lagrangian, as given in Eq. (1), together with the \(R_2\) contributions of Eq. (6) for the contact-terms in OpenLoops  [44, 45]. We compared the amplitudes for the Higgsstrahlung and VBF Higgs processes at the amplitude level and found perfect agreement. Here, for the latter we included the \(pp\rightarrow (V\rightarrow q\bar{q})h\) subprocess. Furthermore, we compared NLO fixed-order differential cross sections obtained with MadGraph5_aMC@NLO against Sherpa+OpenLoops and found agreement within the statistical uncertainty of the Monte Carlo integrations.

3 Results and examples

In order to illustrate the usage of HiggsPO and the effect of NLO QCD corrections, we present the simulation of EW Higgs production processes (Zh, Wh, and VBF) at the LHC with \({\sqrt{S}}=13~\)TeV for several benchmark points close to the SM, shown in Table 1.3 In each benchmark point a different contact term is switched on, together with the SM contribution (\(k_{ZZ} = k_{WW} = 1\)). As parton distribution functions (PDFs) we employ NNPDF23NLO and use the value of \(\alpha _S\) they provide. SM input parameters are chosen in accordance with the defaults of the HiggsPO UFO model. Renormalisation and factorisation scales for both processes are chosen as \(\mu =\mu _\mathrm{R}=\mu _\mathrm{F}=H_\mathrm{T}/2\), where \(H_\mathrm{T}\) is the scalar sum of the \(p_\mathrm{T}\) of all final state particles.

We shower Les Houches events with PYTHIA6 [42] and reconstruct jets using the FastJet [47] implementation of the anti-\(k_T\) jet algorithm [48] with \(R=0.4\) and a minimum \(p_\mathrm{T, jet}\) of 10 GeV.
Table 1

Benchmark points in the PO parameter space used to generate events in \(p p \rightarrow Z H\) (I, II, III, IV and V), \(p p \rightarrow W H\) (I, VI and VII), and VBF (I, ..., VII) processes. In all benchmarks we set kZZ = kWW = 1

BP

eZuL

eZuR

eZdL

eZdR

eWuL

phiWuL

I

0

0

0

0

0

0

II

0.0195

0

0

0

0

0

III

0

0.0195

0

0

0

0

IV

0

0

0.0244

0

0

0

V

0

0

0

0.0244

0

0

VI

0

0

0

0

0.018

0

VII

0

0

0

0

0.018

\(\frac{\pi }{2}\)

Fig. 1

Invariant mass distribution of the ZH system in \(p p \rightarrow Z H\) production at 13 TeV. Solid (dotted) lines show NLO (LO) + PS predictions for five benchmark points in the PO parameter space listed in Table 1

Fig. 2

Invariant mass distribution of the WH system in \(p p \rightarrow W H\) production at 13 TeV. Solid (dotted) lines show NLO (LO) + PS predictions for three benchmark points in the PO parameter space listed in Table 1

3.1 Associated VH production

Events for on-shell Zh and Wh production in the HiggsPO UFO model can be generated at NLO in QCD with MadGraph5_aMC@NLO with the following commands4:

In Figs. 1 and 2 we show the ZH and WH invariant mass distributions respectively obtained at NLO (solid lines) and LO (dashed lines) for the benchmark points listed in Table 1. The lower panel of each figure shows the ratio between the NLO and LO results. We observe that, with the choice of \(H_\mathrm{T}\) as dynamical renormalisation and factorisation scales, the NLO K-factor is fairly flat and universal in most of the considered kinematic regime for all considered benchmark points.

3.2 VBF Higgs production

Events for VBF Higgs production, without \(V(\rightarrow q\bar{q})H\) contributions, are generated via5:
Formally this processes comprises IR-divergent pentagon one-loop diagrams with two massive vector boson propagators. In the available automated frameworks such contributions can typically only be evaluated in Feynman gauge. Thus, in line with the default of MadGraph5_aMC@NLO such diagrams should be discarded in conjunction with the HiggsPO model. This is motivated by the tiny numerical impact of these diagrams in the VBF phase-space. However, this introduces a formal mismatch of the IR poles in the computation. This is a well known issue for VFB Higgs production in MadGraph5_aMC@NLO and a pragmatic solution requires to turn off the check for IR pole cancellation. This can be achieved by modifying the #IRPoleCheckThreshold parameter in the ‘FKS_params.dat’ file from 1.0d-5 to -1.0d0.
In our numerical analysis we employ the same VBF cuts as in Ref. [1], i.e. we require
$$\begin{aligned} p_{\mathrm{T, j}_{1,2}}> 30 \text { GeV}, \quad |\eta _{\mathrm{j}_{1,2}}| < 4.5, \quad m_{\mathrm{j}_1 \mathrm{j}_2} > 500 \text { GeV}, \end{aligned}$$
(7)
where jets are ordered according to their \(p_T\). In Fig. 3 we show the NLO (solid lines) and LO (dashed lines) leading-jet \(p_T\) distribution for the benchmark points listed in Table 1. Again, in the lower panel the ratio of the NLO over LO predictions is presented, indicating a rather flat and universal K-factor across a large kinematic regime.
Fig. 3

Distribution of the leading-jet \(p_T\) in VBF Higgs prodution events at 13 TeV. Solid (dotted) lines show NLO (LO) + PS predictions for seven benchmark points in the PO parameter space listed in Table 1

Fig. 4

NLO+PS predictions of cross section (pb) for \(\varDelta \phi (\mathrm{j}_1, \mathrm{j}_2)\)-\(p_\mathrm{T, j_1}\) distributions in VBF Higgs production at 13 TeV. 2D distributions are shown for the SM and PO extensions with \(\varepsilon _{WW}=1\), \(\varepsilon _{WW}^{CP}=1\)

At particle-level more than two jets can be reconstructed in VBF events. Thus, different jet definitions and VBF tagging techniques can have relevant impact on kinematic distributions in VBF Higgs production [49]. In particular, it might be interesting to optimize these choices such that the sensitivity to the different PO is maximised. We leave this to future studies.

Finally, in Fig. 4 we illustrate the NLO 2D distributions of the azimuthal angle difference of two leading jets \(\varDelta \phi (\mathrm{j}_1, \mathrm{j}_2)\) and the leading jet transverse momentum \(p_{T,j1}\) in VBF Higgs production. These 2D distributions are sensitive to the presence of transversal (\(\varepsilon _{ZZ,WW}\)) and CP-odd (\(\varepsilon _{ZZ,WW}^{CP}\)) POs, since these are characterized by a different Lorentz structure than the SM hVV coupling. In order to illustrate such effects we present the distributions for two rather extreme PO benchmark points (\(\varepsilon _{WW}=1\) and \(\varepsilon _{WW}^{CP}=1\)), alongside the SM prediction. We also checked that taking such 2D distributions into account in a global PO fit will substantially improve the sensitivity to the transversal and CP-odd PO, compared to using only the leading jet \(p_T\) distribution.

4 PO-dependence of the simplified template cross sections at NLO in QCD

As an example of a practical application of the HiggsPO tool, we compute the simplified template cross sections (STXS) for Zh production as a function of the Higgs PO at NLO in QCD. The STXS, introduced in Chapter III.2 of Ref. [3], are cross sections defined in some simple and idealized bins. The choice of such bins is motivated by optimizing the sensitivity to BSM effects while minimizing the necessary acceptance corrections (thus theory dependence) by choosing simple selection cuts to a phase-space close to the realistic fiducial region. Several stages with more and more granular bins are envisaged with increasing luminosity.

In the case of Zh production, with \(Z \rightarrow \ell ^+ \ell ^-\) or \(\bar{\nu }\nu \) decays, the kinematical variable chosen for the binning is the \(p_T\) of the Z boson (\(E_T^\mathrm{miss}\) in the case of neutrinos), with bins \([0 - 150 - 250 - 400 - \infty ]\) GeV. Furthermore, a possible split of the \([150 - 250]\) GeV bin is envisaged in 0-jet and \(\ge \)1-jet categories. For simplicity of this preliminary analysis we neglect this jet categorization. An overall selection on the Higgs rapidity \(|y_h| < 2.5\) is applied.

The idea behind the STXS is that, on the one hand, the experimental collaborations could provide a measurement of such cross sections in a very model-independent way by combining Higgs data in different channels. On the other hand, since the STXS are defined at the reconstruction level with simple kinematical cuts, they can easily be computed in a given BSM model in order to derive the limits. The Higgs PO-dependence of the STXS \(p_T\) bins in Zh production, for example, can be used to obtain global fits in the PO formalism.

The \(p p \rightarrow Z h\) events are generated in the same way as described in Sect. 3 and subsequently analyzed in order to apply the selection cut on \(y_h\) described above and separate them into different \(p_T^Z\) bins. In this way we obtain the cross section in each STXS bin as a quadratic function of the Higgs PO. The result, normalised to the SM value, is:
$$\begin{aligned} R_{0-150}&= k_{ZZ}^2 + k_{ZZ} (27 \varepsilon _{Z u_L} - 28 \varepsilon _{Z d_L} - 14 \varepsilon _{Z u_R} + 4.7 \varepsilon _{Z d_R}) \nonumber \\&\quad + 10^2 [6.0 (\varepsilon _{Z u_L}^2 + \varepsilon _{Z u_R}^2) + 4.7 (\varepsilon _{Z d_L}^2 + \varepsilon _{Z d_R}^2)] ~, \nonumber \\ R_{150-250}&= k_{ZZ}^2 + k_{ZZ} (80 \varepsilon _{Z u_L} - 71 \varepsilon _{Z d_L} - 37 \varepsilon _{Z u_R} + 14 \varepsilon _{Z d_R}) \nonumber \\&\quad + 10^3 [4.5 (\varepsilon _{Z u_L}^2 + \varepsilon _{Z u_R}^2) + 3.2 (\varepsilon _{Z d_L}^2 + \varepsilon _{Z d_R}^2)] ~, \nonumber \\ R_{250-400}&= k_{ZZ}^2 + k_{ZZ} (180 \varepsilon _{Z u_L} - 145 \varepsilon _{Z d_L} - 83 \varepsilon _{Z u_R} + 31 \varepsilon _{Z d_R}) \nonumber \\&\quad + 10^4 [2.2 (\varepsilon _{Z u_L}^2 + \varepsilon _{Z u_R}^2) + 1.5 (\varepsilon _{Z d_L}^2 + \varepsilon _{Z d_R}^2)] ~, \nonumber \\ R_{400-\infty }&= k_{ZZ}^2 + k_{ZZ} (610 \varepsilon _{Z u_L} - 340 \varepsilon _{Z d_L} - 235 \varepsilon _{Z u_R} + 145 \varepsilon _{Z d_R}) \nonumber \\&\quad + 10^5 [2.5 (\varepsilon _{Z u_L}^2 + \varepsilon _{Z u_R}^2) + 1.4 (\varepsilon _{Z d_L}^2 + \varepsilon _{Z d_R}^2)] ~, \end{aligned}$$
(8)
where \(R_x \equiv \sigma _i / \sigma ^\mathrm{SM}_i\) in each STXS bin. As shown in Fig. 1, the NLO corrections to the Zh invariant mass distribution are largely independent of the PO parameters. This implies that the effect of NLO corrections in the ratios in Eq. (8) are rather small.

5 Conclusions

In this work we have presented an implementation of the Higgs Pseudo-Observables framework for electroweak Higgs production in the HiggsPO  UFO model for Monte Carlo event generation at NLO in QCD. The extension of the previous version of the tool to NLO accuracy will allow to perform a robust interpretation of high-statistics data on Vh and VBF Higgs production in this general BSM formalism.

The tool takes into account all PO relevant for EW Higgs production processes, under the hypothesis of an \(U(2)^3\) flavour symmetry. This symmetry reduces the number of independent \(Vhqq'\) quark contact terms to six, or to five if one further assumes CP conservation. On top of this, all the flavour-universal PO and leptonic contact terms (relevant for leptonic Higgs decays) are also included. All these parameters can be easily accessed and modified for example from the param_card.dat in MadGraph5_aMC@NLO.

In order to illustrate the use of the HiggsPO model we have presented simulations of Zh, Wh, and VBF Higgs production for a set of benchmark points. For each benchmark point we have studied the differential cross section in a relevant kinematical variable, and the effect of the NLO corrections. Our analysis shows that, for a suitable choice of factorization and renormalization scales such as \(H_T/2\), the NLO corrections are relevant, \(\mathscr {O}(20\%)\), but rather flat in the kinematical regime of interest. This suggests a fair convergence of the perturbative expansion. In the case of VBF production we have shown how the correlation between the leading jet \(p_T\) and the azimuthal angle difference of the two leading jets, \(\varDelta \phi (j_1, j_2)\), is a very sensitive probe of coupling structures different from the SM ones. Finally, we have presented the first computation of simplified template cross sections for Zh production as a function of Higgs pseudo-observables at NLO accuracy.

As is well known, higher-order EW corrections in EW Higgs production become very important at large energies due to the appearance of Sudakov logarithms. The dominant EW Sudakov logarithms are universal and can be factorized from the hard part of the scattering process [50]. Nonetheless, being of electroweak nature, these corrections are expected to generate a non-trivial mixing among different PO. This implies that their implementation in the PO framework, especially in the context of a Monte Carlo event generator, is a non-trivial task [51]. While waiting for the completion of this program, we stress that current data on EW Higgs production are dominated by events near the kinematical threshold, where the effect of EW corrections is not log-enhanced.6

To sum up, the HiggsPO– event simulation tool for Higgs pseudo-observables framework – is upgraded to NLO in QCD and is publicly available at http://www.physik.uzh.ch/data/HiggsPO.

Footnotes

  1. 1.

    See Chapter III.1 of Ref. [3] for a comprehensive and up to date review of the topic.

  2. 2.

    These \(R_2\) contributions can be understood as the missing \((4 - D)\)-dimensional contributions, by construction not necessarily automatically generated within numerical one-loop generators. These contributions are universal and can be restored from process-independent effective counterterms [31, 32, 33].

  3. 3.

    Since differential cross sections are a second order polynomial in the PO, running a simulation for a handful of points is enough to cover the entire parameter space (see for example [46]).

  4. 4.

    For generating LO events it is sufficient to remove the [QCD] flag.

  5. 5.

    In order to include also the \(V(\rightarrow q\bar{q})H\) contribution one should use instead > generate p p> h j j HPO=1 QED=2 [QCD].

  6. 6.

    At present no general-purpose Monte Carlo event generator includes such EW corrections for BSM models.

Notes

Acknowledgements

This research was supported in part by the Swiss National Science Foundation (SNF) under contract 200021-159720.

References

  1. 1.
    A. Greljo, G. Isidori, J.M. Lindert, D. Marzocca, Eur. Phys. J. C 76(3), 158 (2016). arXiv:1512.06135 ADSCrossRefGoogle Scholar
  2. 2.
    M. Gonzalez-Alonso, A. Greljo, G. Isidori, D. Marzocca, Eur. Phys. J. C 75, 128 (2015). arXiv:1412.6038 ADSCrossRefGoogle Scholar
  3. 3.
    LHC Higgs Cross Section Working Group Collaboration, D. de Florian et al. arXiv:1610.07922
  4. 4.
    M. Bordone, A. Greljo, G. Isidori, D. Marzocca, A. Pattori, Eur. Phys. J. C 75(8), 385 (2015). arXiv:1507.02555 ADSCrossRefGoogle Scholar
  5. 5.
    T. Han, G. Valencia, S. Willenbrock, Phys. Rev. Lett. 69, 3274–3277 (1992). arXiv:hep-ph/9206246 ADSCrossRefGoogle Scholar
  6. 6.
    T. Figy, C. Oleari, D. Zeppenfeld, Phys. Rev. D 68, 073005 (2003). arXiv:hep-ph/0306109 ADSCrossRefGoogle Scholar
  7. 7.
    P. Bolzoni, F. Maltoni, S.-O. Moch, M. Zaro, Phys. Rev. Lett. 105, 011801 (2010). arXiv:1003.4451 ADSCrossRefGoogle Scholar
  8. 8.
    P. Bolzoni, F. Maltoni, S.-O. Moch, M. Zaro, Phys. Rev. D 85, 035002 (2012). arXiv:1109.3717 ADSCrossRefGoogle Scholar
  9. 9.
    M. Cacciari, F .A. Dreyer, A. Karlberg, G .P. Salam, G. Zanderighi, Phys. Rev. Lett. 115(8), 082002 (2015). arXiv:1506.02660 ADSCrossRefGoogle Scholar
  10. 10.
    F.A. Dreyer, A. Karlberg, Phys. Rev. Lett. 117(7), 072001 (2016). arXiv:1606.00840 ADSCrossRefGoogle Scholar
  11. 11.
    M. Ciccolini, A. Denner, S. Dittmaier, Phys. Rev. Lett. 99, 161803 (2007). arXiv:0707.0381 ADSCrossRefGoogle Scholar
  12. 12.
    M. Ciccolini, A. Denner, S. Dittmaier, Phys. Rev. D 77, 013002 (2008). arXiv:0710.4749 ADSCrossRefGoogle Scholar
  13. 13.
    A. Denner, S. Dittmaier, S. Kallweit, A. Mck, Comput. Phys. Commun. 195, 161–171 (2015). arXiv:1412.5390 ADSCrossRefGoogle Scholar
  14. 14.
    G. Ferrera, M. Grazzini, F. Tramontano, Phys. Rev. Lett. 107, 152003 (2011). arXiv:1107.1164 ADSCrossRefGoogle Scholar
  15. 15.
    G. Ferrera, M. Grazzini, F. Tramontano, JHEP 04, 039 (2014). arXiv:1312.1669 ADSCrossRefGoogle Scholar
  16. 16.
    G. Ferrera, M. Grazzini, F. Tramontano, Phys. Lett. B 740, 51–55 (2015). arXiv:1407.4747 ADSCrossRefGoogle Scholar
  17. 17.
    C. Englert, M. McCullough, M. Spannowsky, Phys. Rev. D 89(1), 013013 (2014). arXiv:1310.4828 ADSCrossRefGoogle Scholar
  18. 18.
    D. Goncalves, F. Krauss, S. Kuttimalai, P. Maierhfer, Phys. Rev. D 92(7), 073006 (2015). arXiv:1509.01597 ADSCrossRefGoogle Scholar
  19. 19.
    A. Denner, S. Dittmaier, S. Kallweit, A. Muck, JHEP 03, 075 (2012). arXiv:1112.5142 ADSCrossRefGoogle Scholar
  20. 20.
    F. Granata, J.M. Lindert, C. Oleari, S. Pozzorini, JHEP 09, 012 (2017). arXiv:1706.03522 ADSCrossRefGoogle Scholar
  21. 21.
    LHC Higgs Cross Section Working Group Collaboration, S. Dittmaier et al. arXiv:1101.0593
  22. 22.
    S. Dittmaier et al. arXiv:1201.3084
  23. 23.
    LHC Higgs Cross Section Working Group Collaboration, J. R. Andersen et al. arXiv:1307.1347
  24. 24.
    F. Maltoni, K. Mawatari, M. Zaro, Eur. Phys. J. C 74(1), 2710 (2014). arXiv:1311.1829
  25. 25.
    K. Mimasu, V. Sanz, C. Williams, JHEP 08, 039 (2016). arXiv:1512.02572 ADSCrossRefGoogle Scholar
  26. 26.
    C. Degrande, B. Fuks, K. Mawatari, K. Mimasu, V. Sanz, Eur. Phys. J. C 77(4), 262 (2017). arXiv:1609.04833 ADSCrossRefGoogle Scholar
  27. 27.
    J. Alwall, R. Frederix, S. Frixione, V. Hirschi, F. Maltoni, O. Mattelaer, H.S. Shao, T. Stelzer, P. Torrielli, M. Zaro, JHEP 07, 079 (2014). arXiv:1405.0301 ADSCrossRefGoogle Scholar
  28. 28.
    S. Höche, S. Kuttimalai, S. Schumann, F. Siegert, Eur. Phys. J. C 75(3), 135 (2015). arXiv:1412.6478 ADSCrossRefGoogle Scholar
  29. 29.
    The Higgs PO UFO model. http://www.physik.uzh.ch/data/HiggsPO
  30. 30.
    A. Alloul, N.D. Christensen, C. Degrande, C. Duhr, B. Fuks, Comput. Phys. Commun. 185, 2250–2300 (2014). arXiv:1310.1921 ADSCrossRefGoogle Scholar
  31. 31.
    G. Ossola, C.G. Papadopoulos, R. Pittau, JHEP 05, 004 (2008). arXiv:0802.1876 ADSCrossRefGoogle Scholar
  32. 32.
    T. Binoth, J.P. Guillet, G. Heinrich, JHEP 02, 013 (2007). arXiv:hep-ph/0609054 ADSCrossRefGoogle Scholar
  33. 33.
    A. Bredenstein, A. Denner, S. Dittmaier, S. Pozzorini, JHEP 08, 108 (2008). arXiv:0807.1248 ADSCrossRefGoogle Scholar
  34. 34.
    P. Draggiotis, M.V. Garzelli, C.G. Papadopoulos, R. Pittau, JHEP 04, 072 (2009). arXiv:0903.0356 ADSCrossRefGoogle Scholar
  35. 35.
    C. Degrande, Comput. Phys. Commun. 197, 239–262 (2015). arXiv:1406.3030 ADSMathSciNetCrossRefGoogle Scholar
  36. 36.
    T. Hahn, Comput. Phys. Commun. 140, 418–431 (2001). arXiv:hep-ph/0012260 ADSCrossRefGoogle Scholar
  37. 37.
    V. Hirschi, R. Frederix, S. Frixione, M.V. Garzelli, F. Maltoni, R. Pittau, JHEP 05, 044 (2011). arXiv:1103.0621 ADSCrossRefGoogle Scholar
  38. 38.
    T. Peraro, Comput. Phys. Commun. 185, 2771–2797 (2014). arXiv:1403.1229 ADSCrossRefGoogle Scholar
  39. 39.
    V. Hirschi, T. Peraro, JHEP 06, 060 (2016). arXiv:1604.01363 ADSCrossRefGoogle Scholar
  40. 40.
    S. Frixione, Z. Kunszt, A. Signer, Nucl. Phys. B 467, 399–442 (1996). arXiv:hep-ph/9512328 ADSCrossRefGoogle Scholar
  41. 41.
    R. Frederix, S. Frixione, F. Maltoni, T. Stelzer, JHEP 10, 003 (2009). arXiv:0908.4272 ADSCrossRefGoogle Scholar
  42. 42.
    T. Sjostrand, S. Mrenna, P.Z. Skands, JHEP 05, 026 (2006). arXiv:hep-ph/0603175 ADSCrossRefGoogle Scholar
  43. 43.
    O. Mattelaer, Eur. Phys. J. C 76(12), 674 (2016). https://doi.org/10.1140/epjc/s10052-016-4533-7. arXiv:1607.00763 [hep-ph]
  44. 44.
    F. Cascioli, P. Maierhofer, S. Pozzorini, Phys. Rev. Lett. 108, 111601 (2012). arXiv:1111.5206 ADSCrossRefGoogle Scholar
  45. 45.
    The OpenLoops one-loop generator by F. Cascioli, J. Lindert, P. Maierhöfer, S. Pozzorini. http://openloops.hepforge.org
  46. 46.
    ATLAS Collaboration, ATL-PHYS-PUB-2015-047 (2015)Google Scholar
  47. 47.
    M. Cacciari, G.P. Salam, G. Soyez, Eur. Phys. J. C 72, 1896 (2012). arXiv:1111.6097 ADSCrossRefGoogle Scholar
  48. 48.
    M. Cacciari, G.P. Salam, G. Soyez, JHEP 0804, 063 (2008). https://doi.org/10.1088/1126-6708/2008/04/063. arXiv:0802.1189 [hep-ph]
  49. 49.
    M. Rauch, D. Zeppenfeld, Phys. Rev. D 95(11), 114015 (2017). arXiv:1703.05676 ADSCrossRefGoogle Scholar
  50. 50.
    A. Denner, S. Pozzorini, Eur. Phys. J. C 21, 63–79 (2001). arXiv:hep-ph/0104127 ADSCrossRefGoogle Scholar
  51. 51.
    G. Isidori, A. Pattori. (work in progress) Google Scholar

Copyright information

© The Author(s) 2017

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Funded by SCOAP3

Authors and Affiliations

  • Admir Greljo
    • 1
    • 2
    • 3
  • Gino Isidori
    • 1
  • Jonas M. Lindert
    • 4
  • David Marzocca
    • 1
    • 5
  • Hantian Zhang
    • 1
  1. 1.Physik-InstitutUniversität ZürichZurichSwitzerland
  2. 2.PRISMA Cluster of Excellence and Mainz Institute for Theoretical PhysicsJohannes Gutenberg-Universität MainzMainzGermany
  3. 3.Faculty of ScienceUniversity of SarajevoSarajevoBosnia and Herzegovina
  4. 4.Department of Physics, Institute for Particle Physics PhenomenologyDurham UniversityDurhamUK
  5. 5.INFN, Sezione di Trieste, SISSATriesteItaly

Personalised recommendations