PaperlayerView exact revision on arXiv

Source-linked AI summary

HiggsBounds 2.0.0: Confronting Neutral and Charged Higgs Sector Predictions with Exclusion Bounds from LEP and the Tevatron

Philip Bechtle, Oliver Brein, Sven Heinemeyer, Georg Weiglein, Karina E. Williams

arXiv:1102.1898v1hep-phhep-ex

TL;DR

Higgs searches produce fragmented, heterogeneous limits whose applicability depends on a model’s production and decay properties. HiggsBounds 2.0.0 provides a unified test using model predictions, expected-limit sensitivity selection, and observed 95% C.L. limits. It expands coverage to charged Higgs sectors and newer LEP and Tevatron analyses, and applies the results to invisible Higgs decays and the Randall–Sundrum scalar sector.

  • Problem

    Higgs-search constraints are distributed across analyses with differing normalisations, updates, statistical overlaps, and model assumptions, complicating their application to arbitrary Higgs sectors.

  • Method

    HiggsBounds converts supplied Higgs-sector predictions into analysis-specific quantities, selects the most sensitive application using expected limits, and tests it against the corresponding observed limit.

  • Results

    Version 2.0.0 adds singly charged-Higgs tests and expanded LEP/Tevatron coverage; applications find exclusion near mH ≈2mW despite invisible branching ratios up to 32% and constrain the Randall–Sundrum scalar sector.

  • Takeaways & Limitations

    The release enables updated direct-search constraints to be applied across broader Higgs-sector models, including invisible-decay scenarios and the Randall–Sundrum model.

  • Takeaways & Limitations

    The demonstrated Randall–Sundrum scenario is extreme and may already conflict with other observations, while the method assumes narrow widths and sufficiently compatible signal behaviour.

Abstract

from arXiv · show

HiggsBounds 2.0.0 is a computer code which tests both neutral and charged Higgs sectors of arbitrary models against the current exclusion bounds from the Higgs searches at LEP and the Tevatron. As input, it requires a selection of model predictions, such as Higgs masses, branching ratios, effective couplings and total decay widths. HiggsBounds 2.0.0 then uses the expected and observed topological cross section limits from the Higgs searches to determine whether a given parameter scenario of a model is excluded at the 95% C.L. by those searches. Version 2.0.0 represents a significant extension of the code since its first release (1.0.0). It includes now 28/53 LEP/Tevatron Higgs search analyses, compared to the 11/22 in the first release, of which many of the ones from the Tevatron are replaced by updates. As a major extension, the code allows now the predictions for (singly) charged Higgs bosons to be confronted with LEP and Tevatron searches. Furthermore, the newly included analyses contain LEP searches for neutral Higgs bosons (H) decaying invisibly or into (non flavour tagged) hadrons as well as decay-mode independent searches for neutral Higgs bosons, LEP searches via the production modes tau+ tau- H and b b-bar H, and Tevatron searches via t t-bar H. Also, all Tevatron results presented at the ICHEP'10 are included in version 2.0.0. As physics applications of HiggsBounds 2.0.0 we study the allowed Higgs mass range for model scenarios with invisible Higgs decays and we obtain exclusion results for the scalar sector of the Randall-Sundrum model using up-to-date LEP and Tevatron direct search results.

1 Introduction

HiggsBounds 2.0.0 streamlines the comparison of arbitrary neutral and singly charged Higgs sectors with LEP and Tevatron exclusion limits. This release substantially expands the included searches and supports applications beyond standard benchmark models.

  • Motivation: Higgs searches constrain Higgs-sector parameter spaces, but their limits are distributed across publications, normalisations, updates, and model-dependent assumptions.A search performed under Standard Model assumptions may not apply when a model’s production or decay mechanisms differ substantially.
  • Contribution: HiggsBounds is a Fortran code that compares Higgs-sector predictions with existing exclusion limits for rapid testing of many models.The program is designed to make state-of-the-art Higgs-search constraints convenient to apply.
  • Contribution: Version 2.0.0 adds comparisons for singly charged Higgs bosons using current LEP and Tevatron limits.This is identified as a major extension relative to earlier releases.
  • Contribution: New neutral-Higgs coverage includes invisible, diphoton, non-flavour-tagged hadronic, decay-mode-independent, τ+τ−H, and b b-bar H searches.The release also adds several Tevatron analyses, including ICHEP 2010 updates, t t-bar H, H→γZ, and a W+W− combination.
  • Scope boundary: This release excludes charged-Higgs channels with H±→neutral Higgs plus W and searches for doubly charged Higgs bosons.Accordingly, “charged Higgs bosons” denotes singly charged states in the paper.
  • Paper scope: The paper documents the program’s workflow, required model inputs, implemented analyses, and two physics applications.The remaining sections describe the general approach, input variants, experimental analyses, and applications.

2 General Approach

HiggsBounds converts model predictions into analysis-specific quantities, selects the statistically most sensitive application using expected limits, and tests that application against the observed limit. Its interpretation assumes narrow Higgs widths and sufficiently SM-like signal kinematics and relevant production behaviour.

  • Inputs: The program accepts Higgs masses, total widths, branching ratios, production cross sections, effective couplings, and CP information for neutral states.Simpler input variants are available when specified approximations are valid, and missing quantities can be set to zero.
  • Input limitations: Incomplete model information can cause some analyses to be treated as inapplicable, even though fuller inputs might make them usable.This limits the coverage of exclusions derived from partial predictions.
  • Analysis applications: For each analysis application X, HiggsBounds calculates the normalised model quantity Qmodel(X) from the relevant predicted cross section and branching ratio.The application refers to a search analysis applied to one or two model Higgs bosons at specified mass values.
  • Normalisation: The code normalises limits using Standard Model production cross sections and branching ratios because Tevatron analyses generally used those predictions.Users should account for this normalisation choice when interpreting output.
  • Statistical procedure: The program chooses X0 as the application with the largest Qmodel/Qexpec ratio, using expected limits to identify the highest statistical sensitivity.Only the observed limit for this selected application is then used for the exclusion test.
  • Assumptions: The method requires the narrow-width approximation and assumes models do not substantially alter background signatures, signal kinematics, or the LEP VBF-to-Higgsstrahlung relation.The implemented exclusion bounds were obtained under the narrow-width assumption.

3 Theoretical predictions required as input

HiggsBounds accepts model predictions through three input formats, ranging from general hadronic cross sections to reduced effective-coupling inputs. These inputs cover Higgs masses, widths, branching ratios, couplings, production cross sections, and relevant top-quark branching ratios.

  • Input formats: Users can choose among hadronic, partonic, and effective-coupling input formats, with the required inputs depending on the Higgs sector and searches being tested.Neutral-only or LEP-only applications require fewer inputs than combined neutral, charged, LEP, and Tevatron tests.
  • whichinput=hadr: The general hadronic format requires masses, total widths, Higgs branching ratios, top-quark branching ratios, and normalised LEP and Tevatron production cross sections.The charged-Higgs and top-decay inputs are needed when applying relevant charged-Higgs searches.
  • Normalisation and couplings: Normalised production cross sections use a reference cross section, generally the Standard Model equivalent, while processes without one use specified fictitious or model-based references.The invisible branching ratio counts decays whose products appear only as missing energy.
  • whichinput=part: The partonic format specifies normalised Tevatron partonic production ratios, which HiggsBounds converts internally into hadronic cross-section ratios.Partonic ratios are easier to calculate in many models, and an approximation can reduce the required ratios from twelve to four.
  • whichinput=effC: The effective-coupling format reduces the input to masses, widths, effective fermion and boson couplings, and branching ratios without Standard Model equivalents.HiggsBounds then derives the corresponding partonic inputs; this option is intended for cases where width- and cross-section-based gluon couplings agree closely.

4 Analyses in HiggsBounds 2.0.0

HiggsBounds 2.0.0 implements a broad set of LEP and Tevatron analyses and selects applicable limits according to model assumptions and predicted signal topologies. It includes updated Tevatron results, additional neutral-Higgs channels, and charged-Higgs analyses.

  • Implemented analyses: The database contains expected and observed 95% C.L. cross-section limits with varied normalisations, including model-independent and selected dedicated analyses.The program interpolates the tabulated Q-values linearly between neighbouring Higgs-mass points.
  • Implemented analyses: 81 Higgs search analyses are implemented: 28 from LEP and 53 from the Tevatron.Several Tevatron analyses from version 1.2.0 were replaced with updated analyses based on more data.
  • LEP analyses: New LEP coverage includes invisible, γγ, non-flavour-tagged hadronic, decay-mode-independent, b b-bar H, and τ+τ−H searches.The listed neutral-Higgs final states include bb, ττ, cascade decays, and multi-fermion final states.
  • Tevatron analyses: Tevatron coverage includes single-topology neutral-Higgs analyses, combined Standard Model Higgs analyses, and charged-Higgs analyses.The included results encompass analyses released publicly at the ICHEP 2010 conference and combinations from CDF and D0.
  • Applicability conditions: HiggsBounds determines which analyses apply by checking assumptions such as Standard Model likeness, top-decay branching relations, and Higgs CP properties.The default SM-likeness tolerance is 2%, but the method can occasionally be overly restrictive.
  • Mass-degenerate Higgs bosons: Cross sections of neutral Higgs bosons with sufficiently similar masses may be added, although this feature is disabled by default and requires small interference effects.For analyses requiring an SM-likeness test, full cross sections are not added; the recommended mass difference threshold is at most 10 GeV.

5 New HiggsBounds Operating Instructions

HiggsBounds 2.0.0 offers library, command-line, and online interfaces for supplying model predictions, selecting analyses, and interpreting exclusion results.

  • New HiggsBounds Operating Instructions: Three usage formats are provided: a subroutine library, a command-line version, and an online version.The command-line format is broadly applicable, while the online version avoids installation for checking a few points.
  • New HiggsBounds Operating Instructions: The command-line version reads model data from text files, whereas the library integrates HiggsBounds into other programs.The online version provides access to the same functionality through a web form.
  • 5.1 Common features: Input: Input settings specify the numbers of neutral and singly charged Higgs bosons, the considered analysis subset, and the neutral-sector prediction format.The corresponding variables are nHzero, nHplus, whichanalyses, and whichinput.
  • 5.1 Common features: Input: Model inputs include effective normalized squared couplings, CP properties, branching ratios, and cross-section ratios, with array sizes and ordering defined in the input tables.The tables document the recognized arrays and their required data formats.
  • 5.2 Common features: Output: Outputs report whether a point is excluded at 95% C.L., the most sensitive analysis application, the contributing Higgs-boson count, and the observed-rate ratio.Key.dat associates analysis reference numbers with their applications.
  • Installation: The code can be compiled as a library and linked to another Fortran program with -L<HBpath> -lHB.The HiggsBounds subroutines reserve file handles 10, 11, 44, 45, and 87.

Subroutine initialize HiggsBounds (⋆)

Initialization configures the HiggsBounds calculation by declaring the neutral and charged Higgs content, selecting analyses, and supplying the corresponding model-prediction arrays.

  • Initialization: initialize_HiggsBounds is called once before any other HiggsBounds subroutine and sets nHzero, nHplus, and whichanalyses.Setting nHzero or nHplus to zero disables testing of the corresponding sector.
  • Neutral-sector input: Neutral-sector predictions are passed through one of three input routines corresponding to effC, part, or hadr formats.The selected routine is required only when nHzero is non-zero.
  • Neutral-sector input: The neutral effective-coupling input includes Higgs masses, total widths, normalized couplings, and invisible or cascade branching ratios.The partonic and hadronic options instead accept cross-section ratios and branching-ratio arrays.
  • Input conventions: Unneeded branching-ratio, effective-coupling, or cross-section arrays may be filled with zeros, making the corresponding theoretical rates zero.A zero quantity used in an SM-likeness test can nevertheless cause that test to fail.
  • Input conventions: Some arrays are ignored depending on whichanalyses, so setting an irrelevant array to zero does not affect the result.For example, the cascade branching ratio is irrelevant when whichanalyses='onlyT'.
  • Charged-sector input: Charged-sector input supplies charged-Higgs masses and widths, production ratios, and branching ratios for top and charged-Higgs decays.This routine is required only when nHplus is non-zero.

Subroutine run HiggsBounds ⋆

run_HiggsBounds performs the main exclusion calculation after all model inputs are supplied, returning the exclusion result and diagnostic variables.

  • Execution: run_HiggsBounds is called once at the end of the program after all other HiggsBounds subroutines.The Fortran 90 finish routine deallocates HiggsBounds allocatable arrays.
  • Standard Model data: The library provides Standard Model branching ratios, total decay widths, and hadronic Tevatron cross sections used internally by HiggsBounds.The SM branching ratios and total width come from HDECAY 3.4, while hadronic cross sections primarily use TEV4LHC inputs.
  • Standard Model data: The auxiliary SM top-quark decay-width function can help calculate BR(t → H+b) but is not used internally by HiggsBounds.It depends on the top-quark pole mass and is given at next-to-leading order.
  • Standard Model data: Built-in functions are valid only within specified mass ranges and return −1 when called outside those ranges.This boundary applies to the provided Standard Model functions.
  • Examples: Example programs demonstrate use with Fourth Generation Model predictions and with FeynHiggs and CPsuperH.The Fourth Generation example compares SM and model input through the effC interface.

Command line and input file format (⋆)

The command-line interface maps model predictions stored in structured files to the selected HiggsBounds analyses, with optional online and example workflows.

  • Command line and input file format (⋆): The command-line invocation is ./HiggsBounds <whichanalyses> <whichinput> <nHzero> <nHplus> <prefix>.The prefix is prepended to input and output filenames and may include directories.
  • Command line and input file format (⋆): Input files begin with a line number that links predictions belonging to the same model parameter point across files.Files must not contain comments or blank lines.
  • Command line and input file format (⋆): Neutral branching ratios are separated into BR_OP.dat for ordinary-particle decays and BR_NP.dat for decays without a Standard Model equivalent.The new filenames reflect the inclusion of invisible branching ratios.
  • Command line and input file format (⋆): The input tables define recognized arrays for couplings, CP properties, branching ratios, and cross-section ratios, including required ordering conventions.Only the lower-left triangle is required for g2hjhiZ and lepCS, while diagonal cascade-decay entries are not physical quantities.
  • Command line and input file format (⋆): Optional additional.dat columns can store scan variables not required by HiggsBounds, such as tan β for MSSM scans.The first entry on each line remains the line number.
  • Command line and input file format (⋆): The required input-file subset depends on whichinput, whichanalyses, nHzero, and nHplus; irrelevant files are not read or required.The command-line version selects files appropriate to the requested analysis and input settings.
  • Examples: The online version generates a form for selected neutral and charged Higgs content and reports the resulting HiggsBounds output on screen.It also identifies the processes with the second- and third-highest statistical sensitivities and provides pre-filled examples.

6 Physics Applications

The physics applications illustrate HiggsBounds using invisible Higgs decays and Randall–Sundrum scalar-sector constraints. The resulting exclusions depend on Higgs masses, branching ratios, couplings, and the most sensitive LEP or Tevatron search channels.

  • Invisible Higgs decays: The toy model varies the invisible Higgs branching ratio and mass while fixing other cross sections and widths to Standard Model values under the narrow-width approximation.The example program supplies one neutral Higgs boson, no charged Higgs bosons, effective couplings, decay widths, and branching ratios to HiggsBounds.
  • Invisible Higgs decays: Almost all invisible branching ratios are excluded up to nearly the LEP kinematical limit, with the dominant channel changing as the invisible branching ratio varies.Lower invisible branching ratios are constrained through H → b b-bar, while higher values use Higgsstrahlung searches with missing momentum; additional OPAL and L3 analyses cover M_H < 90 GeV.
  • Invisible Higgs decays: The Tevatron excludes a wedge extending to BR(H → invisible)=0.32 for Higgs masses between 158 GeV and 175 GeV.The toy-model predictions pass the SM-likeness test because relevant cross sections and branching ratios differ from their Standard Model values by a common factor.
  • Interpreting the variable obsratio: A Higgs boson with M_H = 95 GeV and BR(H → invisible)=0.5 cannot be excluded by LEP Higgsstrahlung searches alone.In the intermediate mass range, 1/obsratio indicates how much the measured Standard Model cross-section limit would need to decrease for exclusion; a 10% improvement at 165 GeV would extend exclusion to BR(H → invisible)=0.38.
  • Constraints on the Randall–Sundrum scalar sector: For the Randall–Sundrum model with Λ_φ=1 TeV and ξ=1/6, LEP excludes Higgs-like and radion-like regions while Tevatron exclusions mainly use single-scalar WW searches.The radion-like state’s enhanced gg coupling and suppressed b b-bar coupling make φ → hadrons sensitive below about 50 GeV, while Tevatron WW searches dominate for scalar masses above roughly 120 GeV.
  • Constraints on the Randall–Sundrum scalar sector: With Λ_φ=1 TeV and m_h=120 GeV, falling ξ generally enlarges the excluded m_φ interval, while a funnel near ξ≈0 remains unexcluded.Above 120 GeV, exclusion is entirely due to the Tevatron single-radion WW analysis; below 120 GeV, the same analysis can instead constrain the Higgs-like state.

7 Summary

HiggsBounds 2.0.0 extends Higgs-sector exclusion testing to neutral and singly charged Higgs bosons using updated LEP and Tevatron searches. Applications demonstrate constraints on invisible-decay scenarios and the Randall–Sundrum scalar sector, while its design supports incorporating future LHC limits.

  • HiggsBounds 2.0.0 tests neutral and singly charged Higgs sectors of arbitrary models against LEP and Tevatron exclusion bounds.The program requires model predictions and provides operating instructions, examples, analysis references, and applicability conditions.
  • The update adds LEP searches for invisible, hadronic, decay-mode-independent, τ+τ−H, and b b-bar H channels, plus Tevatron t t-bar H searches and updated analyses.All Tevatron results presented at ICHEP’10 are included, with many updates replacing previously implemented analyses.
  • A SM-like Higgs remains excluded around mH ≈2mW even with invisible decays having up to 32% branching ratio.This result comes from current Tevatron searches for the SM-like Higgs boson.
  • Including Tevatron results alongside earlier LEP constraints excludes large portions of the studied Randall–Sundrum scalar-sector parameter space through the model-independent H →W+W− search.The authors describe this as the first study, to their knowledge, of Tevatron Higgs-search impacts on this model.
  • HiggsBounds 2.0.0 is designed so upcoming LHC Higgs-search limits can be incorporated readily.LEP, Tevatron, and LHC bounds are presented as information sources for testing model interpretations of possible Higgs-like signals.
Loading 1102.1898v1…