[Paper Review] Extra doublets - Global analyses of Standard Model extensions in the fermionic or scalar sector

Otto Eberhardt|arXiv (Cornell University)|Sep 5, 2013

Particle physics theoretical and experimental studies1 references3 citations

TL;DR

This thesis performs global fits of two Standard Model extensions—fourth-generation fermions (SM4) and the Two-Higgs-Doublet Model (2HDM)—using electroweak precision observables, Higgs signal strengths, and flavor data. It demonstrates that the SM4 is excluded at high significance due to Higgs data, while the 2HDM remains viable in its decoupling limit.

ABSTRACT

I present comprehensive analyses of the viable parameter regions of the Standard Model of particle physics and two of its extensions: the Standard Model with four fermion generations and the Two-Higgs-Doublets model of type II. Applying frequentist methods, I have performed global fits which show that the Standard Model is compatible with electroweak precision data and the measured Higgs signal strengths. The Standard Model with four generations is excluded by the experimental results with a p-value corresponding to more than five standard deviations. In the Two-Higgs-Doublets model of type II, the fit prefers the decoupling limit where it mimics the Standard Model; however, neutral heavy Higgs boson masses below 200 GeV are still allowed.

Motivation & Objective

• To assess the viability of the Standard Model with a fourth fermion generation (SM4) using global experimental constraints.
• To evaluate the Two-Higgs-Doublet Model (2HDM) against electroweak precision observables, Higgs signal strengths, and flavor measurements.
• To determine whether SM4 or 2HDM can consistently explain pre- and post-Higgs-boson discovery data.
• To develop and apply parallelized likelihood fitting tools for high-dimensional parameter scans in non-nested models.
• To quantify model exclusion using likelihood ratio tests, particularly for non-decoupling models like SM4.

Proposed method

• Performs global likelihood fits combining electroweak precision observables (e.g., $m_W$, $R_b^0$), Higgs signal strengths, and flavor data.
• Uses the CKMfitter software framework with custom parallelization scripts to enable efficient 2D and 1D parameter scans across high-dimensional parameter spaces.
• Applies a novel unification strategy ('unify' script) to merge results from parallel jobs, recalculating $\chi^2$ and $p$-values using the global minimum to avoid bias.
• Employs likelihood ratio tests to compare non-nested models (SM vs. SM4, SM vs. 2HDM), accounting for systematic uncertainties and model complexity.
• Introduces analysis cards to configure scan parameters and automate job submission across computational clusters.
• Validates results using dedicated scripts ('redo') to re-run failed 1D scans and merge improved results.

Experimental results

Research questions

• RQ1 Can the Standard Model with a fourth fermion generation (SM4) be excluded by global fits of electroweak and Higgs data?
• RQ2 How do Higgs signal strengths and electroweak precision observables constrain the parameter space of the Two-Higgs-Doublet Model (2HDM)?
• RQ3 To what extent do flavor observables such as $B_s^0 \to \bar{B}_s^0$ mixing and $b \to s\gamma$ decay rates constrain the 2HDM parameter space?
• RQ4 How do the LHC Higgs boson discovery measurements affect the viability of SM4 and 2HDM extensions?
• RQ5 Can the parallelized likelihood fitting framework in CKMfitter reliably explore high-dimensional, non-nested model parameter spaces?

Key findings

• The SM4 is excluded at high significance (p-value < 0.01) based on combined Higgs signal strength and electroweak precision data, particularly due to the observed Higgs boson properties.
• The SM4 model cannot decouple from the SM, leading to strong constraints from Higgs signal strengths and electroweak observables, which are inconsistent with data.
• The 2HDM remains viable in the decoupling limit, where the heavy Higgs states are heavy enough to suppress tree-level FCNCs and satisfy flavor constraints.
• Flavor observables such as $\Delta m_{B_s}$ and $\mathcal{B}(b \to s\gamma)$ significantly constrain the $\tan\beta$–$m_{H^+}$ plane in the 2HDM.
• The Higgs signal strengths measured at the LHC are consistent with the SM but disfavor the SM4, with the $\gamma\gamma$ and $ZZ$ decay channels providing the strongest exclusion power.
• The developed parallelization and unification tools for CKMfitter enable efficient, scalable likelihood scans across complex, high-dimensional models, reducing computation time by orders of magnitude.