Theory Uncertainties for Higgs and Other Searches Using Jet Bins

TL;DR

This paper addresses the underestimation of theoretical uncertainties in Higgs and other searches using jet bins by proposing a scale variation method that accounts for large logarithmic effects at jet boundaries, improving the reliability of cross section predictions.

Contribution

It introduces a new method for estimating theory uncertainties in exclusive jet cross sections by using inclusive cross section variations, ensuring more accurate uncertainty estimates and consistency with resummed calculations.

Findings

The proposed method effectively captures uncertainties due to large logarithms at jet boundaries.

It provides a correlation matrix for jet bins that cancels uncertainties when combining neighboring bins.

The approach aligns fixed-order uncertainties with those from resummed calculations in known cases.

Abstract

Bounds on the Higgs mass from the Tevatron and LHC are determined using exclusive jet bins to maximize sensitivity. Scale variation in exclusive fixed-order predictions underestimates the perturbative uncertainty for these cross sections, due to cancellations between the perturbative corrections leading to large K factors and those that induce logarithmic sensitivity to the jet-bin boundary. To account for this, we propose that scale variation in the fixed-order calculations should be used to determine theory uncertainties for inclusive jet cross sections, whose differences yield exclusive jet cross sections. This yields a theory correlation matrix for the jet bins such that the additional uncertainty from large logarithms due to the jet boundary cancels when neighboring bins are added. This procedure is tested for H + 0, 1 jets, WW + 0 jets, and W + 0, 1, 2 jets, and found to be generally applicable. For a case where the higher-order resummation of the jet boundary corrections is known, we show that this procedure yields fixed-order uncertainties which are theoretically consistent with those obtained in the resummed calculation.