Taming a leading theoretical uncertainty in HH measurements via accurate simulations for bbH production

TL;DR

This paper introduces a new NLO simulation for $b\bar{b}H$ production, reducing theoretical uncertainties in Higgs pair measurements at the LHC by accurately modeling the process and its backgrounds.

Contribution

It provides the first full NLO simulation of $b\bar{b}H$ with massive bottom quarks, improving the accuracy of Higgs pair production background estimates.

Findings

NLO corrections reduce scale dependence from 100% to 50%.

Simulation improves limits on Higgs pair production cross section.

Comparison reveals issues with massless bottom quark schemes.

Abstract

We present a new simulation for Higgs boson production in association with bottom quarks ($b\bar{b}H$) at next-to-leading order (NLO) accuracy matched to parton showers in hadronic collisions. Both contributions, the standard one proportional to the bottom-quark Yukawa coupling and the loop-induced one proportional to the top-quark Yukawa coupling from the gluon-fusion process, are taken into account in a scheme with massive bottom quarks. Therefore, we provide the full simulation of the $b\bar{b}H$ final state in the Standard Model, which constitutes also a crucial background to measurements for Higgs-boson pair ($HH$) production at the Large Hadron Collider when at least one of the Higgs bosons decays to bottom quarks. So far, the modeling of the $b\bar{b}H$ final state induced one of the dominant theoretical uncertainties to $HH$ measurements, as the gluon-fusion component was described only at the leading order (LO) with uncertainties of $\mathcal{O}(100\%)$. Including NLO corrections in its simulation allows us to reduce the scale dependence to $\mathcal{O}(50\%)$ so that it becomes subdominant with respect to other systematic uncertainties. As a case study, we provide an in-depth analysis of the $b\bar{b}H$ background to $HH$ measurements with realistic selection cuts in the $2b2\gamma$ channel. We also compare our novel simulation with the currently-employed ones, discussing possible issues and shortcomings of a scheme with massless bottom quarks. Finally, we propagate the effect of the new $b\bar{b}H$ simulation to $HH$ searches in the $2b2\gamma$ and $2b2\tau$ final states, and we find an improvement of up to 10% (20%) on the current (HL-LHC) limits on the $HH$ cross section.