betaFeedback

WorksPhysics and Astronomy

N3^3LO Power Corrections for 00-jettiness Subtractions With Fiducial Cuts

Abstract

We compute the leading logarithmic power corrections at next-to-next-to-next-to-leading order for 00-jettiness subtractions for color singlet production. We discuss how to disentangle these power corrections from those arising from the presence of fiducial and isolation cuts by using Projection-to-Born improved slicing. We present the results for Drell-Yan and Higgs production in gluon fusion differential in both the invariant mass and rapidity of the color singlet. Our results include all the channels contributing at leading logarithmic order for these processes, including the off-diagonal channels that receive contributions from soft quark emission. We study the numerical impact of the power corrections for Drell-Yan and Higgs production and find it to become negligible compared to the size of the N3^3LO corrections only below τcut∼10−5τ_\text{cut} \sim 10^{-5}. We estimate that in a fully differential calculation at N3^3LO combining the Projection-to-Born improved slicing method and our results for the leading logarithmic power corrections may allow for keeping the slicing uncertainties under control already with τcut≲10−3τ_\text{cut} \lesssim 10^{-3}, marking a significant improvement in efficiency for these methods. These results constitute a crucial ingredient for fully differential N3^3LO calculations based on the NN-jettiness subtraction scheme.

Introduction

As the volume of events collected at the Large Hadron Collider (LHC) and level of refinement in experimental techniques increases, the need for producing more and more precise predictions for LHC observables is becoming of primary importance. Our ability to perform calculations in QCD at higher orders is an essential step in this direction. While techniques to perform fully differential calculation for hadronic collisions at next-to-next-to-leading order (NNLO) are well established [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13], and analytic results are available for inclusive cross sections at next-to-next-to-next-to-leading (N3^{3}LO) for several processes [14, 15, 16, 17], in the last couple of years a number of results are pushing the state of the art for fully differential calculations of the production of a color singlet towards the N3^{3}LO benchmark [18, 19, 20, 21, 22, 23, 24, 25]. Event shape observables, such as thrust [26], or the NN-jettiness resolution variable, TN\mathcal{T}_{N}, [27, 28], can be used to perform subtractions for fixed-order calculations at the LHC [7, 8] as well as at e+e−e^{+}e^{-} colliders. For the particular case of color singlet production, there has been intense progress in extending exclusive calculations to N3^{3}LO. With the recent calculation of the three loop qTq_{T} beam functions [29, 30], all the required perturbative ingredients for qTq_{T} subtractions [1] have been computed (the hard function and qTq_{T} soft function have been calculated before, see [31, 32]). This has enabled the first predictions for fully differential color singlet production at N3^{3}LO accuracy employing the qTq_{T} subtraction method [19, 20, 21, 23, 24, 25]. In the last few years, important results have also come out for obtaining the ingredients to extend NN-jettiness subtraction to N3N^3LO. These include the calculation of the hard, jet, beam, and soft function, which are the perturbative ingredients of the leading power factorization theorem for this observable in soft-collinear effective theory (SCET) [33, 34, 35, 36, 37]. The N3N^3LO jet function has been calculated in [38] and the three-loop beam functions have been obtained in [39] (see also [40]), therefore completely determining the collinear ingredients at this perturbative order. For the case of 2 light-like directions, which is relevant for color singlet production in hadron collisions, significant advancement has happened for the determination of the three loop soft function [41, 42, 43, 44], while the N3N^3LO hard function is the same as for qTq_T subtraction and has been known for a long time [31]. From the scalings of the power corrections [45, 46, 47], it is clear that performing subtractions at this order will dramatically benefit from the knowledge of the power corrections. In the recent years, substantial progress has been achieved in understanding the behaviour of QCD amplitudes and cross sections in the soft and collinear limit beyond leading power for colliders, see for example [47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77]. Using the techniques of [65], in this paper we give results for the leading logarithmic (LL) power corrections for both Drell-Yan and gluon-fusion Higgs production for the 0-jet resolution variable beam thrust [28].

We begin with a brief review of NN-jettiness subtraction [7, 8] and fixing the notation for our calculation. NN-Jettiness subtraction can be used to calculate a cross section σ(O)\sigma(\mathcal{O}) for a given NN-jet process. Here O\mathcal{O} is the differential measurement on the final state of the process, and we will leave the dependence on it implicit for the reminder of this section. In this work we will focus on the case of color singlet production in hadron collisions, therefore we will work with 0-jettiness. The 0-jettiness variable [78, 79] is defined as

T0=∑imin⁡{2qa⋅kiQa,2qb⋅kiQb},(1)\mathcal{T}_0 = \sum_i \min\left\{ \frac{2q_a \cdot k_i}{Q_a}, \frac{2q_b \cdot k_i}{Q_b} \right\}, \tag*{(1)}

where kik_i are the momenta of the hadronic final states. Here, qa,bq_{a,b} are label momenta in SCET, and can be obtained from the mass QQ and rapidity YY of the color singlet final state as

qaμ=xaEcmnμ2=QeYnμ2,qbμ=xbEcmnˉμ2=Qe−Ynˉμ2,(2)q_a^\mu= x_a E_{\mathrm{cm}} \frac{n^\mu}{2} = Q e^Y \frac{n^\mu}{2}, \qquad q_b^\mu= x_b E_{\mathrm{cm}} \frac{\bar n^\mu}{2} = Q e^{-Y} \frac{\bar n^\mu}{2}, \tag*{(2)}

where we have defined the light-like vectors

nμ=(1,0,0,1),nˉμ=(1,0,0,−1)(3)n^\mu= (1,0,0,1), \qquad\bar n^\mu= (1,0,0,-1) \tag*{(3)}

which are aligned with the proton beam directions, and the Born momentum fractions

xa=QeYEcm,xb=Qe−YEcm.(4)x_a = \frac{Q e^Y}{E_{\mathrm{cm}}}, \qquad x_b = \frac{Q e^{-Y}}{E_{\mathrm{cm}}}. \tag*{(4)}

The normalization factors Qa,bQ_{a,b} in eq. (1) can be used to obtain different versions of the 0-jettiness resolution variable. Two natural definitions can be obtained by performing the minimization in the hadronic or in the leptonic/color singlet center of mass frame [27, 80].

This leads to the following definitions

leptonic:Qa=Qb=Q,T0lep=∑imin⁡{xaEcmQn⋅ki,xbEcmQnˉ⋅ki}=∑imin⁡{eYn⋅ki,e−Ynˉ⋅ki}hadronic:Qa,b=xa,bEcm,T0cm=∑imin⁡{n⋅ki,nˉ⋅ki}.(5)\begin{aligned} \text{leptonic:}\quad&Q_a=Q_b=Q,\qquad\mathcal{T}_0^{\mathrm{lep}}=\sum_i\min\left\{\frac{x_a E_{\mathrm{cm}}}{Q}n\cdot k_i,\frac{x_b E_{\mathrm{cm}}}{Q}\bar{n}\cdot k_i\right\}\\ &=\sum_i\min\left\{e^Y n\cdot k_i,e^{-Y}\bar{n}\cdot k_i\right\}\\ \text{hadronic:}\quad&Q_{a,b}=x_{a,b}E_{\mathrm{cm}},\qquad\mathcal{T}_0^{\mathrm{cm}}=\sum_i\min\left\{n\cdot k_i,\bar{n}\cdot k_i\right\}. \tag*{(5)} \end{aligned}

It is by now well understood [45, 46, 47] that the hadronic definition suffers from large power corrections that grow exponentially with rapidity. This can be understood as a manifestation of the fact that the hadronic definition forces an RPI-III [81] (or boost) charge on τ\tau, hence introducing a large dependence on YY in its power expansion coefficients. On the other hand, in the leptonic definition, the observable τ\tau becomes a boost invariant quantity, thanks to the explicit factor of the color singlet rapidity exactly compensating for the RPI charge of the plus or minus component of final state hadronic momenta. For this reason, it does not come as a surprise that an expansion in this variable has a much more stable dependence on YY. Given this behavior, modern implementations of 0-jettiness subtraction utilize the leptonic definition, which will thus also be adopted in the remainder of this paper.

The cross section can be organized as

σ=∫0τcutdτdσdτ+∫τcutτmaxdτdσdτ=∫0τcutdτ[dσ(0)dτ+∑i>0dσ(i)dτ]+∫τcutτmaxdτdσdτ=σsub(τcut)+Δσ(τcut)+∫τcutτmaxdτdσdτ.(6)\begin{aligned} \sigma&=\int_0^{\tau_{\mathrm{cut}}}\mathrm{d}\tau\frac{\mathrm{d}\sigma}{\mathrm{d}\tau}+\int_{\tau_{\mathrm{cut}}}^{\tau_{\mathrm{max}}}\mathrm{d}\tau\frac{\mathrm{d}\sigma}{\mathrm{d}\tau}\\ &=\int_0^{\tau_{\mathrm{cut}}}\mathrm{d}\tau\left[\frac{\mathrm{d}\sigma^{(0)}}{\mathrm{d}\tau}+\sum_{i>0}\frac{\mathrm{d}\sigma^{(i)}}{\mathrm{d}\tau}\right]+\int_{\tau_{\mathrm{cut}}}^{\tau_{\mathrm{max}}}\mathrm{d}\tau\frac{\mathrm{d}\sigma}{\mathrm{d}\tau}\\ &=\sigma_{\mathrm{sub}}(\tau_{\mathrm{cut}})+\Delta\sigma(\tau_{\mathrm{cut}})+\int_{\tau_{\mathrm{cut}}}^{\tau_{\mathrm{max}}}\mathrm{d}\tau\frac{\mathrm{d}\sigma}{\mathrm{d}\tau}. \tag*{(6)} \end{aligned}

For τ>0\tau>0, by construction of the observable, the process has at least one resolved emission on top of the Born configuration of the process. Therefore, the contribution in the region above the cut (τ>τcut>0\tau>\tau_{\mathrm{cut}}>0) can be calculated using a lower loop result for a process with an extra jet in the final state. All the infrared (IR) singularities related to the Born configuration are completely accounted for by the region below the cut, i.e. τ<τcut\tau<\tau_{\mathrm{cut}}. In eq. (6) we have introduced the notation for the power expansion of the differential distribution in τ\tau

dσdτ=dσ(0)dτ+∑i>0dσ(i)dτ,(7)\frac{\mathrm{d}\sigma}{\mathrm{d}\tau}=\frac{\mathrm{d}\sigma^{(0)}}{\mathrm{d}\tau}+\sum_{i>0}\frac{\mathrm{d}\sigma^{(i)}}{\mathrm{d}\tau}, \tag*{(7)}

where1

dσ(i)dτ/dσ(0)dτ∼O(τi).(8)\frac{\mathrm{d}\sigma^{(i)}}{\mathrm{d}\tau}\bigg/\frac{\mathrm{d}\sigma^{(0)}}{\mathrm{d}\tau}\sim\mathcal{O}\left(\tau^i\right). \tag*{(8)}

By construction of the NN-jettiness variable, the leading term of the expansion around τ=0\tau= 0, dσ(0)dτ\frac{\mathrm{d}\sigma^{(0)}}{\mathrm{d}\tau}, contains all the singular terms of the distribution. The remaining powers in the expansion yield a vanishing contribution for τcut→0\tau_{\mathrm{cut}} \to0 after integration. Consequently, we have divided the contributions from the region below the cut into two parts: a subtraction term, σsub\sigma_{\mathrm{sub}}, which must be calculated exactly, and a residual term, Δσ(τcut)\Delta\sigma(\tau_{\mathrm{cut}}), that will be neglected

σsub(τcut)=∫0τcutdτdσsubdτ,Δσ(τcut)=∫0τcutdτ[dσdτ−dσsubdτ].(9)\sigma_{\mathrm{sub}}(\tau_{\mathrm{cut}}) = \int_{0}^{\tau_{\mathrm{cut}}} \mathrm{d}\tau\frac{\mathrm{d}\sigma^{\mathrm{sub}}}{\mathrm{d}\tau}, \qquad\Delta\sigma(\tau_{\mathrm{cut}}) = \int_{0}^{\tau_{\mathrm{cut}}} \mathrm{d}\tau\left[\frac{\mathrm{d}\sigma}{\mathrm{d}\tau} - \frac{\mathrm{d}\sigma^{\mathrm{sub}}}{\mathrm{d}\tau}\right]. \tag*{(9)}

Calculating σ\sigma as the sum of the subtraction term and the integral above the cut introduces an error due to the omission of the terms in Δσ(τcut)\Delta\sigma(\tau_{\mathrm{cut}}), which can be minimized by taking τcut\tau_{\mathrm{cut}} smaller and smaller. 2 The main challenge in the application of NN-jettiness subtraction is the fact that for the precision requirement of state-of-the-art calculations, very small values of τcut\tau_{\mathrm{cut}} are necessary. This leads to severe numerical instabilities in the calculation of the cross section with an extra emission which, in all practical applications, is obtained via Monte Carlo integration techniques. At high perturbative order, and for very small values of τcut\tau_{\mathrm{cut}}, the required level of precision for the above-the-cut result either requires an enormous amount of computing time or it is just beyond the reach of currently available codes.

σ=∫τcutτmax⁡dτ[dσdτ−dσsubdτθ(τ<τoff)]+∫0τoffdτdσsubdτ+Δσ(τcut).(10)\sigma= \int_{\tau_{\mathrm{cut}}}^{\tau_{\max}} \mathrm{d}\tau\left[\frac{\mathrm{d}\sigma}{\mathrm{d}\tau} - \frac{\mathrm{d}\sigma^{\mathrm{sub}}}{\mathrm{d}\tau}\theta(\tau< \tau_{\mathrm{off}})\right] + \int_{0}^{\tau_{\mathrm{off}}} \mathrm{d}\tau\frac{\mathrm{d}\sigma^{\mathrm{sub}}}{\mathrm{d}\tau} + \Delta\sigma(\tau_{\mathrm{cut}}). \tag*{(10)}

Here τoff\tau_{\mathrm{off}} determines the region where the subtraction is active. Ignoring Δσ(τcut)\Delta\sigma(\tau_{\mathrm{cut}}) in eq. (10) introduces a τcut\tau_{\mathrm{cut}}-dependent error, similar to eq. (6). However, the final result remains independent of τoff\tau_{\mathrm{off}}’s value because its contribution is both added and subtracted identically. The main benefit of eq. (10) is that the cancellation of singular terms between dσdτ\frac{\mathrm{d}\sigma}{\mathrm{d}\tau} and dσsubdτ\frac{\mathrm{d}\sigma^{\mathrm{sub}}}{\mathrm{d}\tau} happens point by point in τ\tau. This in principle allows for improving numerical stability and therefore push τcut\tau_{\mathrm{cut}} to smaller values, see ref. [8] for more details.

In what follows, we wish to further investigate the size of Δσ(τcut)\Delta\sigma(\tau_{\mathrm{cut}}). In all common implementations of slicing in the literature, the subtraction term only contains the leading power term in the distribution, i.e. dσsubdτ=dσ(0)dτ∼1τ\frac{\mathrm{d}\sigma^{\mathrm{sub}}}{\mathrm{d}\tau} = \frac{\mathrm{d}\sigma^{(0)}}{\mathrm{d}\tau} \sim\frac{1}{\tau}. This implies that Δσ(τcut)\Delta\sigma(\tau_{\mathrm{cut}}) is the cumulant of the τ\tau distribution beyond leading power. Order by order in perturbation theory, we have that these terms take the form

dσ(i)dτ∼τi−1∑n=1∞∑m=02n−1(αs4π)nln⁡mτ cn,m(i),(11)\frac{\mathrm{d}\sigma^{(i)}}{\mathrm{d}\tau} \sim\tau^{i-1} \sum_{n=1}^{\infty} \sum_{m=0}^{2n-1} \left(\frac{\alpha_s}{4\pi}\right)^n \ln^m \tau\, c_{n,m}^{(i)}, \tag*{(11)}

with coefficients cn,m(i)c_{n,m}^{(i)}. Consequently, at O(αsn)\mathcal{O}(\alpha_s^n), the residual error naively scales as

Δσ(τcut)∼τcut∑m=02n−1ln⁡mτcut+O(τcut2).(12)\Delta\sigma(\tau_{\mathrm{cut}}) \sim\tau_{\mathrm{cut}} \sum_{m=0}^{2n-1} \ln^m \tau_{\mathrm{cut}} + \mathcal{O}(\tau_{\mathrm{cut}}^2). \tag*{(12)}

Hence, at each order in αs\alpha_s, high powers of logarithms are present and they can dramatically enhance the size of these residual uncertainties as one takes τcut\tau_{\mathrm{cut}} to small values. As one

approaches τcut→0\tau_{\mathrm{cut}} \to0, of course eventually the power suppression will dominate the logarithmic enhancements. However, at N3N^{3}LO (n=3n = 3), enhanced power corrections up to ln⁡5τcut\ln^{5}\tau_{\mathrm{cut}} demand very low values of τcut\tau_{\mathrm{cut}} to minimize the impact of Δσ(τcut)\Delta\sigma(\tau_{\mathrm{cut}}) on the final result.

Before we continue, let us make some technical remarks. In general, the cn,m(i)c_{n,m}^{(i)} are functions of the Born kinematics, and may depend on parton distribution functions in the case of hadron collisions. The fact that only integer powers of τcut\tau_{\mathrm{cut}} will appear beyond leading power is not guaranteed. It is in fact known that, for example, in the presence of fiducial cuts, arbitrary fractional powers of τ\tau will appear [82]. For example, pTp_{T} cuts on the leptonic final states of a color singlet decay will induce power corrections of O(τcut)O(\sqrt{\tau_{\mathrm{cut}}}), in analogy to the power corrections of O(∣qTcut∣)O(\lvert q_{T}^{\mathrm{cut}}\rvert) appearing in the qTq_{T}-subtraction formalism when such cuts are applied. However, these fractional power corrections vanish for sufficiently inclusive observables. Moreover, they are purely kinematic in nature. Although they may be difficult to account for analytically in general, we will demonstrate in § 2 how to disentangle them from other sources of power corrections by calculating them numerically. Therefore, we can disregard these fractional power corrections and consider eq. (12) as an appropriate starting point for understanding the scaling of the slicing error. In particular, since the focus of this work is the application of NN-jettiness subtraction at N3N^{3}LO, we will focus on the case n=3n = 3. In this case,

ΔσN3LO(τcut)∼(αs4π)3∫0τcutdτ(c3,5NLPln⁡5τ+c3,4NLPln⁡4τ+c3,3NLPln⁡3τ+…).(13)\Delta\sigma^{\mathrm{N3LO}}(\tau_{\mathrm{cut}}) \sim\left(\frac{\alpha_s}{4\pi}\right)^3 \int_0^{\tau_{\mathrm{cut}}} \mathrm{d}\tau\left(c_{3,5}^{\mathrm{NLP}}\ln^5\tau+ c_{3,4}^{\mathrm{NLP}}\ln^4\tau+ c_{3,3}^{\mathrm{NLP}}\ln^3\tau+ \ldots\right). \tag*{(13)}

As calculations beyond leading power at N3N^{3}LO are very challenging, it is interesting to understand what impact may have on the slicing error the calculation of individual terms in the logarithmic expansion of the NLP series. As a matter of fact, having analytic control on them would allow for their inclusion in the subtraction term σsub(τcut)\sigma^{\mathrm{sub}}(\tau_{\mathrm{cut}}), therefore removing them from contributing to the residual error ΔσN3LO(τcut)\Delta\sigma^{\mathrm{N3LO}}(\tau_{\mathrm{cut}}). In order to estimate such effect, one may look at the size ΔσN3LO(τcut)\Delta\sigma^{\mathrm{N3LO}}(\tau_{\mathrm{cut}}) with or without the ln⁡5τ\ln^5\tau making the assumption that c3,jNLP∼c3NLPc_{3,j}^{\mathrm{NLP}} \sim c_{3}^{\mathrm{NLP}} for j≤4j \leq4, i.e. that the numerical coefficients in front of the logarithms are of comparable size3. To get an idea of the size of the numerical improvement due to the analytic control on the LL term, we can compare it to the size of the slicing error ΔσN3LO(τcut′)\Delta\sigma^{\mathrm{N3LO}}(\tau'_{\mathrm{cut}}) for τcut′<τcut\tau'_{\mathrm{cut}} < \tau_{\mathrm{cut}}. In Figure 1, we show this analysis with some samples values τcut′=τcut/10,τcut/100\tau'_{\mathrm{cut}} = \tau_{\mathrm{cut}}/10, \tau_{\mathrm{cut}}/100. We see that, due to the large power of the logarithm, analytically calculating the leading logarithm at this order, may improve the numerical accuracy of the result by an order of magnitude which translates to obtaining the same accuracy with a cut that is almost 2 order of magnitudes larger than in the case with no power corrections. Although this is just a qualitative picture, as the real quantitative impact will necessarily depend on the relative size and sign of the coefficients of the logs (which will be functions of the Born kinematic and the partonic channel), the scaling of the improvement is very promising.

Figure 1. Size of next-to-leading power (NLP) logarithmic contributions to the cumulant cross section. We see that, for this region of τcut\tau_{\mathrm{cut}}, going from a ln⁡5τ\ln^5 \tau to a ln⁡4τ\ln^4 \tau dependence for the distribution has a numerical impact comparable to a reduction of O(50)\mathcal{O}(50) in the slicing parameter. As a consequence, gaining analytic control on single logarithmic powers of the NLP correction may provide improvements on the slicing errors comparable to orders of magnitude of additional computing time.

In this work we will calculate analytically the leading logarithmic contribution to Δσ(τcut)\Delta\sigma(\tau_{\mathrm{cut}}) at O(αs3)\mathcal{O}(\alpha_s^3).

Projection to Born Improved Subtractions

In realistic experimental measurements a set of kinematic selection cuts on the final state particles is required. If these selection cuts are IRC safe, they do not spoil the singular structure of the cross section. However, they may induce large power corrections to the τ\tau distribution, which often times are significantly bigger than the ones due to subleading power QCD dynamics. In the case of qTq_T subtraction for color singlet production, a boost prescription can be implemented to capture these fiducial power corrections at the level of the factorization theorem [82, 83, 84, 85, 86], but this method fails in the presence of photon isolation cuts. Unfortunately, in the more general cases, accounting for these cuts analytically in the subtraction term is very challenging, since they consist of non-analytic constraints on the phase space of particles. Here, following a method proposed in ref. [82], we will show how one can separate fiducial power corrections from other sources of power corrections4 and obtain them numerically. For concreteness, we focus on the case of Higgs production in gluon fusion at N3N^3LO with di-photon decay. Given the desired final state measurement O\mathcal{O}, which we allow for including fiducial cuts, instead of computing the cross section σ(O)\sigma(\mathcal{O}), we use 0-jettiness subtraction to compute the cross section for another observable σ(O~)\sigma(\widetilde{\mathcal{O}}), with O~\widetilde{\mathcal{O}} being a Born projection of the measurement O\mathcal{O}. The difference σ(O)−σ(O~)\sigma(\mathcal{O})-\sigma(\widetilde{\mathcal{O}}) vanishes for τ→0\tau\to0 for any infrared safe observable O\mathcal{O}. This implies that σ(O)−σ(O~)\sigma(\mathcal{O})-\sigma(\widetilde{\mathcal{O}}) can be calculated using only information about the NNLO Higgs + jet process. In formulas this takes the form

σh,N3LO(O)=σh,N3LO(O~)+σh+j,NNLO(O−O~)(14)\sigma_{h,N^3\mathrm{LO}}(\mathcal{O})=\sigma_{h,N^3\mathrm{LO}}(\widetilde{\mathcal{O}})+\sigma_{h+j,\mathrm{NNLO}}(\mathcal{O}-\widetilde{\mathcal{O}}) \tag*{(14)}
=∫0τcutdτ dσh,N3LOsubdτ(O~)+∫τ>τcutdσh+j,NNLOfull(O~)(Slicing for O~)+∫0τcutdτ[dσh,N3LOfulldτ−dσh,N3LOsubdτ](O~)(Slicing residual for O~ at τ<τcut)+∫dσh+j,NNLOfull(O−O~)(P2B correction factor for O vs O~)\begin{aligned} ={}&\int_{0}^{\tau_{\mathrm{cut}}}\mathrm{d}\tau\,\frac{\mathrm{d}\sigma^{\mathrm{sub}}_{h,N^3\mathrm{LO}}}{\mathrm{d}\tau}(\widetilde{\mathcal{O}})+\int_{\tau>\tau_{\mathrm{cut}}}\mathrm{d}\sigma^{\mathrm{full}}_{h+j,\mathrm{NNLO}}(\widetilde{\mathcal{O}}) && \text{(Slicing for }\widetilde{\mathcal{O}}\text{)}\\ &+\int_{0}^{\tau_{\mathrm{cut}}}\mathrm{d}\tau\left[\frac{\mathrm{d}\sigma^{\mathrm{full}}_{h,N^3\mathrm{LO}}}{\mathrm{d}\tau}-\frac{\mathrm{d}\sigma^{\mathrm{sub}}_{h,N^3\mathrm{LO}}}{\mathrm{d}\tau}\right](\widetilde{\mathcal{O}}) && \text{(Slicing residual for }\widetilde{\mathcal{O}}\text{ at }\tau<\tau_{\mathrm{cut}}\text{)}\\ &+\int\mathrm{d}\sigma^{\mathrm{full}}_{h+j,\mathrm{NNLO}}(\mathcal{O}-\widetilde{\mathcal{O}}) && \text{(P2B correction factor for }\mathcal{O}\text{ vs }\widetilde{\mathcal{O}}\text{)} \end{aligned}

It is easy to recognize that the Projection to Born (P2B) correction factor precisely corresponds to the subtracted real emission term of a P2B subtraction in pppp [18, 87], with all the nice numerical features that P2B subtraction carries. The main advantage of this P2B improved slicing from the point of view of the power corrections is that the slicing calculation is performed for a Born observable. This means that the residual corrections of the slicing only depend on Born kinematics and therefore they are insensitive to effects of the fiducial cuts on the full phase space (which are the ones responsible for the presence of large fiducial power corrections). The physical effect of the presence of the fiducial cuts is accounted for by the P2B correction factor that enjoys the numerical stability of the P2B subtraction, namely the fact that, point by point in the phase space of the real radiation, the counterterm is given by the full real matrix element. Notice that using projection to Born improved subtractions would not change the fact that these cuts may severely affect the perturbative convergence of the fixed order distributions in certain regions [18, 19], as this is an issue of fixed order predictions and not of the way those are obtained. Moreover, even in the case of different fiducial cuts, such as the quadratic-cuts from ref. [88], one would still benefit from employing the P2B improved subtraction compared to the traditional one. This is the case because the main effect of those cuts is to remove the linear power corrections (which in the case of 0-jettiness would be the ones scaling as O(τcut)\mathcal{O}(\sqrt{\tau_{\mathrm{cut}}})), but they would modify the power corrections beyond that. Therefore, in order to analytically improve the subtraction without employing the P2B improved subtraction, one would need to analytically calculate both the dynamical power corrections as well as the fiducial ones, now entering at O(τcut)O(\tau_{\mathrm{cut}}).

Leading Logarithmic Power Corrections at N3^{3}LO

In this section we give the results for the leading logarithmic power corrections for color singlet production at the LHC at N3^{3}LO using 0-jettiness subtraction. We start by using consistency relations to understand what is the minimal set of ingredients needed to obtain the result. We will then present the results for the case of Drell-Yan and Higgs production in gluon fusion.

Consistency Relations and Differential Color Singlet Production Beyond Leading Power

The cancellation of IR singularities order by order in perturbation theory imposes constraints on the structure of the differential distribution in τ\tau. We can leverage these constraints to derive consistency conditions between different contributions entering at NLP [45]. Here, we work out the consistency conditions to N3^{3}LO. The modes necessary to describe the process to the order we are working are hard, (ultra)soft, and collinear. Note that we do not distinguish between nn and nˉ\bar{n} collinear modes in this counting as they will behave identically for the purposes of this section. In the same spirit we will use the word loop loosely, indicating both loop corrections as well as phase space integrals, since, mode by mode, they both contribute in dimensional regularization with the same power in the ϵ\epsilon expansion to the cross section. Beyond leading power, we always need at least one collinear or soft mode, since otherwise we are confined to the Born configuration whose contributions are always proportional to δ(τ)\delta(\tau), which clearly constitutes a leading power term. The next-to-leading-power term of the differential distribution in τ\tau at bare level takes the form

dσNLPdτ=∑n=1∞(αs4π)n×[∑t=0n−1∑κ∈Mn−t∑i=02n−1cκ,itϵi(μ2Q2τm(κ)/(n−t))(n−t)ϵ](15)\frac{\mathrm{d}\sigma^{\mathrm{NLP}}}{\mathrm{d}\tau} = \sum_{n=1}^{\infty} \left(\frac{\alpha_s}{4\pi}\right)^n \times \left[ \sum_{t=0}^{n-1} \sum_{\kappa\in\mathcal{M}_{n-t}} \sum_{i=0}^{2n-1} \frac{c_{\kappa,i}^{t}}{\epsilon^i} \left(\frac{\mu^2}{Q^2\tau^{m(\kappa)/(n-t)}}\right)^{(n-t)\epsilon} \right] \tag*{(15)}

where Ml\mathcal{M}_l is the set of all possible sets of modes of length ll excluding purely hard contributions, m(κ)m(\kappa) is the scaling of the list of modes κ\kappa and t>0t>0 represents terms that will be removed by the interference of UV counterterms (in the full theory) with lower order results, such as coupling renormalization terms. To make this definitions more concrete, at O(αs3)O(\alpha_s^3), we have contributions involving up to 3 modes, so

M3={{s,s,s},{c,c,c},{c,c,s},{h,h,c},{h,c,s},…},{h,h,h}∉M3.(16)\mathcal{M}_3 = \left\{ \{s,s,s\},\{c,c,c\},\{c,c,s\},\{h,h,c\},\{h,c,s\},\ldots \right\}, \qquad \{h,h,h\}\notin\mathcal{M}_3. \tag*{(16)}

The τϵm(κ)\tau^{\epsilon m(\kappa)} scaling for a set of modes is determined by the characteristic scales associated with the soft modes μs2∼Q2τ2\mu_s^2\sim Q^2\tau^2, collinear modes μc2∼Q2τ\mu_c^2\sim Q^2\tau, and the hard modes μh2∼Q2\mu_h^2\sim Q^2. To give some examples of scaling at three loops:

m({s,s,s})=6,m({c,s,s})=5,m({c,c,s})=4,m(\{s,s,s\})=6,\qquad m(\{c,s,s\})=5,\qquad m(\{c,c,s\})=4,
m({h,c,s})=3,m({h,h,s})=2,m({h,h,c})=1.(17)m(\{h,c,s\})=3,\qquad m(\{h,h,s\})=2,\qquad m(\{h,h,c\})=1. \tag*{(17)}

Due to the requirement of having at least one cc or ss mode contributing at NLP, we always have m(κ)>0m(\kappa)>0. Equipped with this notation, we can expand eq. (15) at each order in αs\alpha_s up to O(ϵ0)\mathcal{O}(\epsilon^0) and impose the cancellation of all poles. We work out the constraints at N3^3LO and we write the N3^3LO distribution at NLP in terms of the minimal number of coefficients

1σ0dσNLPdτ∣O(αs3)=chhc,5ln⁡5τ(18)\left.\frac{1}{\sigma_0}\frac{\mathrm{d}\sigma^{\mathrm{NLP}}}{\mathrm{d}\tau}\right|_{\mathcal{O}(\alpha_s^3)} = c_{hhc,5}\ln^5\tau \tag*{(18)}
+(chhc,4+csss,4−dss,4)ln⁡4τ+dss,4ln⁡μ2Q2ln⁡3τ{}+ \left(c_{hhc,4}+c_{sss,4}-d_{ss,4}\right)\ln^4\tau+d_{ss,4}\ln\frac{\mu^2}{Q^2}\ln^3\tau
+(4chhc,3+chhs,3+chcc,3−csss,3+2dhc,3+dss,3−es,3)ln⁡3τ+…{}+\left(4c_{hhc,3}+c_{hhs,3}+c_{hcc,3}-c_{sss,3}+2d_{hc,3}+d_{ss,3}-e_{s,3}\right)\ln^3\tau+\ldots

where the dots include terms with lower powers of ln⁡τ\ln\tau and constants, and for ease of notation we used ck,i0≡ck,ic^0_{k,i}\equiv c_{k,i}, ck,i1≡dk,ic^1_{k,i}\equiv d_{k,i}, ck,i2≡ek,ic^2_{k,i}\equiv e_{k,i}. From eq. (18) we see that for the coefficient of the leading log, ln⁡5τ\ln^5\tau, one only needs to calculate the leading, 1/ϵ51/\epsilon^5, pole of the contribution coming from two hard loops and a single collinear emission.

For this reason, we consider the production of a color singlet hh in proton-proton collisions in association with a single collinear emission of momentum kk

p(pa)+p(pb)→h(q)+k.(19)p(p_a)+p(p_b)\to h(q)+k. \tag*{(19)}

We want to retain all information on the Born kinematics of the color-singlet such that the calculation of our power corrections can be applied to fully differential calculations. Following the common practice, we parametrize the Born phase space using the rapidity and the invariant mass of the color singlet

Q2=q2,Y=12ln⁡(pb⋅qpa⋅q).(20)Q^2=q^2,\qquad Y=\frac{1}{2}\ln\left(\frac{p_b\cdot q}{p_a\cdot q}\right). \tag*{(20)}

The contribution from a single collinear emission along the nn-direction at NLP in the 0-jettiness distribution, differential in the color singlet rapidity and invariant mass and using the leptonic definition of 0-jettiness, reads [47]

dσnNLPdQ2dYdτ=∫xa1dzazaQ2xaxbEcm4(Q2τ/μ2)−ϵ(4π)2(1−za)ϵ{fa(xaza)fb(xb)A(2)(k[τ,za],ϵ)+τA(0)(k[τ,za],ϵ)[−fa(xaza)fb(xb)+fa(xaza)xbfb′(xb)]},(21)\frac{\mathrm{d}\sigma_n^{\mathrm{NLP}}}{\mathrm{d}Q^2\mathrm{d}Y\mathrm{d}\tau} =\int_{x_a}^{1}\frac{\mathrm{d}z_a}{z_a}\frac{Q}{2x_ax_bE_{\mathrm{cm}}^4}\frac{(Q^2\tau/\mu^2)^{-\epsilon}}{(4\pi)^{2}(1-z_a)^\epsilon} \left\{f_a\left(\frac{x_a}{z_a}\right)f_b(x_b)A^{(2)}(k[\tau,z_a],\epsilon) +\tau A^{(0)}(k[\tau,z_a],\epsilon)\left[-f_a\left(\frac{x_a}{z_a}\right)f_b(x_b)+f_a\left(\frac{x_a}{z_a}\right)x_bf_b'(x_b)\right]\right\}, \tag*{(21)}

where kk is fixed by the measurement constraints to be

kμ[τ,za]=xaEcm1−zazanμ2+τEcmxanˉμ2+Qτ1−zazan⊥μ,(22)k^\mu[\tau,z_a]=x_aE_{\mathrm{cm}}\frac{1-z_a}{z_a}\frac{n^\mu}{2}+\frac{\tau E_{\mathrm{cm}}}{x_a}\frac{\bar n^\mu}{2}+Q\sqrt{\tau}\sqrt{\frac{1-z_a}{z_a}}n_\perp^\mu, \tag*{(22)}

and A(2)∼τA(0)A^{(2)}\sim\tau A^{(0)} is the subleading power expansion of the matrix element. In eq. (21) we see that we get two contributions: The first term accounts for the contribution from subleading power matrix elements integrated over the leading power collinear phase space.

Figure 2. Example SCET cut diagrams for the off-diagonal channels contributing to leading logarithm at next-to-leading-power. In blue are depicted collinear fields, while in black the color singlet particles. The red blobs represent subleading hard scattering operators whose Wilson coefficients will encode the virtual loop corrections to the process. For our N3^{3}LO analysis, one needs to consider the square of the 1-loop Wilson coefficients and the interference of the 2-loop and tree-level Wilson coefficients.

The second term in the braces corresponds to the terms involving the leading power matrix element integrated over the subleading power expansion of the phase space, which in this case includes the Born measurement functions defining QQ and YY as well as τ\tau in terms of the momenta pa,pbp_a,p_b, and kk of the color charged particles. The leading power matrix elements can be obtained using two loop splitting amplitudes, while the amplitudes at NLP for a single collinear emission at two loops were derived in ref. [62, 65] from the full kinematic results of ref. [89, 90, 91].

The less obvious results are the ones for the off-diagonal subleading power matrix element related to the soft quark Sudakov derived in ref. [65] and we report them here for completeness. The one for Higgs production reads

Agq,H(2),LL(k,ϵ)=Agg→HLO(Q)×8παsCF(1−za)za12[(αs4π)4CF+4(CA−CF)(1−za)−ϵϵ2]2(23)A^{(2),\mathrm{LL}}_{gq,H}(k,\epsilon)=A^{\mathrm{LO}}_{gg\to H}(Q)\times\frac{8\pi\alpha_s C_F}{(1-z_a)z_a}\frac{1}{2}\left[\left(\frac{\alpha_s}{4\pi}\right)\frac{4C_F+4(C_A-C_F)(1-z_a)^{-\epsilon}}{\epsilon^2}\right]^2 \tag*{(23)}

while for Drell-Yan we have

Aqg,DY(2),LL(k,ϵ)=Aqqˉ→VLO(Q)×4παsTF(2+za2−2za1−za)[(αs4π)4CF+4(CA−CF)(1−za)−ϵϵ2]2(24)A^{(2),\mathrm{LL}}_{qg,\mathrm{DY}}(k,\epsilon)=A^{\mathrm{LO}}_{q\bar q\to V}(Q)\times4\pi\alpha_s T_F\left(\frac{2+z_a^2-2z_a}{1-z_a}\right)\left[\left(\frac{\alpha_s}{4\pi}\right)\frac{4C_F+4(C_A-C_F)(1-z_a)^{-\epsilon}}{\epsilon^2}\right]^2 \tag*{(24)}

These subleading power matrix elements have no τ\tau dependence since they scale as τ0\tau^0, which is suppressed compared the 1/τ1/\tau behavior of the matrix element squared at LP. Notice the divergent za→1z_a\to1 behaviour of these subleading power matrix elements which is responsible for promoting to leading logarithm the contributions of these off-diagonal channels.

Analytic Results

We now present the explicit results for the leading logarithmic correction for 0-jettiness at N3^{3}LO for Drell-Yan and Higgs production. In both cases we can factor out the leading order partonic cross section

σ^LO(Q)=∣MLO∣22xaxbEcm2={4παem23NcQ2Ecm2[Qq2+(vq2+aq2)(vl2+al2)−2Qqvqvl(1−mZ2/Q2)(1−mZ2/Q2)2+mZ2ΓZ2/Q4]Drell-Yan2πδ(Q2−mH2)2Q2Ecm2αs2Q4576π2v2Higgs(25)\hat{\sigma}^{\mathrm{LO}}(Q)=\frac{|\mathcal{M}^{\mathrm{LO}}|^2}{2x_a x_b E_{\mathrm{cm}}^2}= \begin{cases} \frac{4\pi\alpha_{\mathrm{em}}^2}{3N_cQ^2E_{\mathrm{cm}}^2}\left[Q_q^2+\frac{(v_q^2+a_q^2)(v_l^2+a_l^2)-2Q_qv_qv_l(1-m_Z^2/Q^2)}{(1-m_Z^2/Q^2)^2+m_Z^2\Gamma_Z^2/Q^4}\right] & \text{Drell-Yan}\\ \frac{2\pi\delta(Q^2-m_H^2)}{2Q^2E_{\mathrm{cm}}^2}\frac{\alpha_s^2Q^4}{576\pi^2v^2} & \text{Higgs} \end{cases} \tag*{(25)}

and write the cross section at NLP as

dσNLP(n)dQ2dYdτ=σ^LO(Q)(αs4π)n∫xa1dzaza∫xb1dzbzb[fi(xaza)fj(xbzb)CasfifjNLP(za,zb,τ)+xazafi′(xaza)fj(xbzb)CasfifjNLP(za,zb,τ)+fi(xaza)xbzbfj′(xbzb)CasfifjNLP(za,zb,τ)].(26)\frac{\mathrm{d}\sigma^{\mathrm{NLP}(n)}}{\mathrm{d}Q^2\mathrm{d}Y\mathrm{d}\tau} =\hat{\sigma}^{\mathrm{LO}}(Q)\left(\frac{\alpha_s}{4\pi}\right)^n\int_{x_a}^1\frac{\mathrm{d}z_a}{z_a}\int_{x_b}^1\frac{\mathrm{d}z_b}{z_b}\left[f_i\left(\frac{x_a}{z_a}\right)f_j\left(\frac{x_b}{z_b}\right)C_{a_s f_i f_j}^{\mathrm{NLP}}(z_a,z_b,\tau)+\frac{x_a}{z_a}f_i'\left(\frac{x_a}{z_a}\right)f_j\left(\frac{x_b}{z_b}\right)C_{a_s f_i f_j}^{\mathrm{NLP}}(z_a,z_b,\tau)+f_i\left(\frac{x_a}{z_a}\right)\frac{x_b}{z_b}f_j'\left(\frac{x_b}{z_b}\right)C_{a_s f_i f_j}^{\mathrm{NLP}}(z_a,z_b,\tau)\right]. \tag*{(26)}

Drell-Yan

For the case of Drell-Yan, we find for the diagonal channel

CasfqfqˉNLP(za,zb,τ)=64CF3log⁡5(τ)δ(1−za)δ(1−zb),C_{a_s f_q f_{\bar q}}^{\mathrm{NLP}}(z_a,z_b,\tau)=64C_F^3\log^5(\tau)\delta(1-z_a)\delta(1-z_b),
Casfq′fqˉNLP(za,zb,τ)=Casfqfqˉ′NLP(za,zb,τ)=−12CasfqfqˉNLP(za,zb,τ).(27)C_{a_s f_q' f_{\bar q}}^{\mathrm{NLP}}(z_a,z_b,\tau)=C_{a_s f_q f_{\bar q}'}^{\mathrm{NLP}}(z_a,z_b,\tau)=-\frac{1}{2}C_{a_s f_q f_{\bar q}}^{\mathrm{NLP}}(z_a,z_b,\tau). \tag*{(27)}

The fact that the result is proportional to δ\delta-functions in {za,zb}\{z_a,z_b\} is expected due to the soft-collinear consistency of the leading logarithmic term. As a matter of fact, hard and soft modes are confined to za=zb=1z_a=z_b=1, therefore every soft pole of hard and soft nature will be accompanied by a corresponding delta function in these splitting variables. In this work we obtained the results from the collinear limit, which in general has the full phase space available in za,bz_{a,b}, however, since the leading divergence will necessarily need to cancel against the soft one, the leading logarithmic contribution needs to be proportional to δ(1−za)δ(1−zb)\delta(1-z_a)\delta(1-z_b).

At NLP the leading logarithmic correction receives contributions also from the off-diagonal channel qgqg. For this channel we obtain

CasfqfgNLP(za,zb,τ)=−163TF(CA2+CACF+CF2)log⁡5(τ)δ(1−za)δ(1−zb),C_{a_s f_q f_g}^{\mathrm{NLP}}(z_a,z_b,\tau)=-\frac{16}{3}T_F(C_A^2+C_AC_F+C_F^2)\log^5(\tau)\delta(1-z_a)\delta(1-z_b),
Casfq′fgNLP(za,zb,τ)=Casfqfg′NLP(za,zb,τ)=0(28)C_{a_s f_q' f_g}^{\mathrm{NLP}}(z_a,z_b,\tau)=C_{a_s f_q f_g'}^{\mathrm{NLP}}(z_a,z_b,\tau)=0 \tag*{(28)}

These leading logarithmic off-diagonal contributions are generate by subleading power matrix elements and encode contributions due to soft quark emissions and subleading power quark-gluon splitting functions, for details see [65]. Note again the presence of delta functions in za,bz_{a,b}. This may seem surprising at first, since at leading power off-diagonal channels do not have contributions from soft kinematics, but that is not the case beyond leading power.

These results encode the full LL contributions at NLP at this order. The other channels start to contribute only beyond the LL at NLP.

Higgs Production in Gluon Fusion

For the case of Higgs in gluon fusion we get leading logarithmic corrections to both the diagonal channel gggg as well as for gqgq. At NLP for the gggg channel, there are contributions both from the expansion of the phase space as well as from the matrix element. Combining both together we get

Cas3fgfgNLP(za,zb,τ)=64CA3log⁡5(τ)δ(1−za)δ(1−zb),C_{a_s^3 f_g f_g}^{\mathrm{NLP}}(z_a,z_b,\tau)=64C_A^3\log^5(\tau)\delta(1-z_a)\delta(1-z_b),
Cas3fg′fgNLP(za,zb,τ)=Cas3fgfg′NLP(za,zb,τ)=−12Cas3fgfgNLP(za,zb,τ).(29)C_{a_s^3 f'_g f_g}^{\mathrm{NLP}}(z_a,z_b,\tau)=C_{a_s^3 f_g f'_g}^{\mathrm{NLP}}(z_a,z_b,\tau)=-\frac{1}{2}C_{a_s^3 f_g f_g}^{\mathrm{NLP}}(z_a,z_b,\tau). \tag*{(29)}

As for the case of Drell-Yan we get PDF derivatives for the diagonal channel. This is due to the kinematic expansion of the momentum fraction beyond leading power and it is one of the contributions coming from the subleading power expansion of the phase space.

We also have a leading logarithmic contribution for the off-diagonal channel gqgq. At LP, this channel does not contribute to the LL since the matrix element is not singular in the soft and collinear limit. At NLP, there is a singularity associated with the emitted quark going soft and collinear which promotes this contribution to a leading logarithm [54, 65]. This can be clearly seen from the za→1z_a \to1 behaviour of eq. (23). Therefore, the leading logarithmic contribution for this channel comes from the singular behaviour of this subleading power amplitude, while the expansion of the phase space to subleading power only generates subleading logarithmic contributions. For this channel we get

Cas3fgfqNLP(za,zb,τ)=−163CF(CA2+CACF+CF2)log⁡5(τ)δ(1−za)δ(1−zb),C_{a_s^3 f_g f_q}^{\mathrm{NLP}}(z_a,z_b,\tau)=-\frac{16}{3}C_F(C_A^2+C_A C_F+C_F^2)\log^5(\tau)\delta(1-z_a)\delta(1-z_b),
Cas3fg′fqNLP(za,zb,τ)=Cas3fgfq′NLP(za,zb,τ)=0.(30)C_{a_s^3 f'_g f_q}^{\mathrm{NLP}}(z_a,z_b,\tau)=C_{a_s^3 f_g f'_q}^{\mathrm{NLP}}(z_a,z_b,\tau)=0. \tag*{(30)}

The absence of PDF derivatives in this channel at this order is expected and it is one manifestation of the lack of contributions due to subleading power phase space expansion discussed above. Note that in general, beyond leading power, one may encounter PDF derivatives even integrating a subleading power matrix element against the leading power phase space. This is due to the fact that subleading power matrix elements may develop power law divergences [64, 92, 93]. In dimensional regularization these divergences generate higher-order plus distributions which involve derivatives of delta functions or plus distributions acting on the leading power phase space and therefore on the collinear PDFs in the case of hadronic collisions. For beam thrust this is not the case at NLP, however note that for pTp_T distributions beyond leading power, and therefore for power corrections needed for improving qTq_T subtraction, this indeed happens, see ref. [64] for details.

Numerical Results

In this section we analyze the numerical size of the power corrections calculated in the previous sections. For concreteness, we will focus on Higgs production in gluon fusion in the infinite top mass approximation and Drell-Yan in proton-proton collision. We will consider a collider center of mass energy Ecm=13.6E_{\mathrm{cm}} = 13.6 TeV and employ PDF4LHC15_nnlo_mc PDFs, with αs(mZ)=0.118\alpha_s(m_Z) = 0.118. We do not expect any of the conclusions in this section to depend in significant way from these choices. Following the discussion in § 1 the residual slicing error, in the usual setup where the subtraction term consists of only the leading power distribution, can be written as

ΔσN3LO(τcut,Y,Q)=∫0τcutdτ[dσdτdQdY−dσLPdτdQdY]O(αs3)=(αs4π)3∫0τcutdτ(C3,5NLP(Q,Y)ln⁡5τ+C3,4NLP(Q,Y)ln⁡4τ+C3,3NLP(Q,Y)ln⁡3τ+…)(31)\begin{aligned} \Delta\sigma^{N^3LO}(\tau_{\mathrm{cut}},Y,Q) = \int_{0}^{\tau_{\mathrm{cut}}} \mathrm{d}\tau\left[ \frac{\mathrm{d}\sigma}{\mathrm{d}\tau\mathrm{d}Q\mathrm{d}Y} - \frac{\mathrm{d}\sigma^{\mathrm{LP}}}{\mathrm{d}\tau\mathrm{d}Q\mathrm{d}Y} \right]_{\mathcal{O}(\alpha_s^3)} \\ &= \left(\frac{\alpha_s}{4\pi}\right)^3 \int_{0}^{\tau_{\mathrm{cut}}} \mathrm{d}\tau\left( C_{3,5}^{\mathrm{NLP}}(Q,Y)\ln^5\tau+ C_{3,4}^{\mathrm{NLP}}(Q,Y)\ln^4\tau+ C_{3,3}^{\mathrm{NLP}}(Q,Y)\ln^3\tau+ \ldots\right) \tag*{(31)} \end{aligned}

In this section we will present the results for the cumulant of the power corrections calculated in secs. 3.2.1 and 3.2.2, differential in rapidity, and Q2Q^2. This shows the numerical impact of our analytic calculation and it is a proxy for ΔσN3LO(τcut,Y,Q)\Delta\sigma^{N^3LO}(\tau_{\mathrm{cut}},Y,Q) since, for τcut≪10−2\tau_{\mathrm{cut}} \ll10^{-2}, the leading logarithmic term is expected to be the dominant contribution. For convenience, from now on we will fix Q=91.2Q = 91.2 GeV for Drell-Yan and Q=125.5Q = 125.5 GeV for Higgs production and drop the dependence on QQ for the rest of the discussion in this section. As ΔσN3LO(τcut,Y)\Delta\sigma^{N^3LO}(\tau_{\mathrm{cut}},Y) strongly depends on τcut\tau_{\mathrm{cut}}, we will present results for three reference values: τcut={10−3,10−4,10−5}\tau_{\mathrm{cut}} = \{10^{-3},10^{-4},10^{-5}\}.

Drell-Yan

In figs. 3 and 4 we show the cumulant of the power corrections calculated in § 3.2.1, normalized by the leading order rapidity spectrum for different initial state channels. To get a quantitative understanding of the size of such NLP terms, we compare them to two interesting quantities. In dark yellow is the scale uncertainty of the N3^3LO Drell-Yan rapidity spectrum from ref. [20]. This provides a reference for the perturbative uncertainty of this class of processes at N3^3LO and one would like the numerical error of the N3^3LO correction, including the slicing one, to be significantly smaller compared to such uncertainty. The other quantity we decided to plot for numerical comparison is the size of the three loop hard function boundary Hqqˉ(Q)H_{q\bar{q}}(Q) [31], which is one of the N3^3LO corrections to the singular distribution and it is shown in dashed gray. Note that Hqqˉ(Q)H_{q\bar{q}}(Q) enters the singular distribution with a δ(τ)\delta(\tau). As we are comparing objects at the cumulant level, it is therefore important to point out that its functional dependence on τ\tau always integrates to 1 irrespectively of the value of τcut\tau_{\mathrm{cut}}. Overall, Hqqˉ(Q)H_{q\bar{q}}(Q) gives a rough idea of the size of the boundary ingredients that enter in the subtraction term at this order. In fig. 3 we present the result for the sum of all the channels while in fig. 4 we provide a breakdown of the contributions for the two production channels that enter at LL NLP. We see that the LL NLP term gives contributions at the order of percent for τcut=10−3\tau_{\mathrm{cut}} = 10^{-3} and starts to become negligible compared to the size of N3^3LO scale variation uncertainties and hard function contribution only starting from τcut=10−5\tau_{\mathrm{cut}} = 10^{-5}. It is interesting to notice in the channel by channel comparison of fig. 4, that the power corrections coming from the off-diagonal qqq q channel are noticeably larger than the ones for the diagonal qqˉq\bar{q} channel. The fact that for Drell-Yan at N3^{3}LO large corrections beyond leading power are coming from the qgqg channel is not completely surprising. As a matter of fact, in the applications of qTq_T-subtraction to these processes it was empirically observed that the most challenging channel for taming the effect of NLP terms was indeed the qgqg channel [20]. On the one hand, this is due to the fact that at LHC energies there is an abundance of gluons from the PDFs, but it is important to recall that at NLP the off-diagonal qgqg channel is not logarithmically suppressed. Although our calculation focuses only on 0-jettiness and therefore it cannot be considered an analytic demonstration of the behavior seen in the applications of qTq_T-slicing, the feature of the PDF enhancement and that this off-diagonal channel will contribute at leading log beyond leading power, we expect to generalize to the case of qTq_T-slicing very naturally, since it was already shown analytically at lower order [64].

Given the experience at NNLO, where already tens of thousands of CPU hours are necessary to obtain reliable NLO above-the-cut results for τcut<10−3\tau_{\mathrm{cut}} < 10^{-3}, see for example refs. [94, 95], it seems unrealistic to expect that current NNLO codes for color singlet plus jet will be able to push to τcut∼10−5\tau_{\mathrm{cut}} \sim10^{-5} without requiring an outrageous amount of computing resources. Hence, these power corrections will have significant impact for realistic applications of 0-jettiness at N3^3LO. In this regard, in Figure 5 we try the exercise of estimating the size of the slicing error if the leading logarithmic corrections at NLP are included in the subtraction term. Clearly, the real size of the residual error will ultimately depend on the actual numerical coefficients of the subleading logarithmic (and power) terms, so some assumptions are necessary for this estimate. In the following we look at the case where c3,4NLP(Y)=c3,5NLP(Y)c_{3,4}^{\mathrm{NLP}}(Y) = c_{3,5}^{\mathrm{NLP}}(Y), i.e. we use for the next-to-leading logarithmic coefficient the same numerical value we obtained for the leading log. We neglect further subleading logs and powers as we expect that for such small values of τcut\tau_{\mathrm{cut}} they will not play a significant role.

Estimate of the slicing error after removal of the leading logarithmic terms

Figure 5. Estimate of the slicing error Δσ(τcut,Y)\Delta\sigma(\tau_{\mathrm{cut}},Y) after the removal of the leading logarithmic terms analytically calculated in this work (i.e. after their inclusion in the subtraction term) for various values of τcut\tau_{\mathrm{cut}}. The result is normalized by the leading order rapidity spectrum and plotted for comparisons are the scale variation uncertainties of the N3^{3}LO spectrum from [20] and the N3^{3}LO hard function contribution to the subtraction term.

Figure 4. Same as Figure 3 but for different production channels. We show the diagonal channel qqˉq\bar q on the left and qgqg on the right. Both channels contribute at leading log at NLP. It is evident that the power corrections to qgqg are substantially larger than those for qqˉq\bar q. Note that the plot shows the absolute value of the LL NLP cumulant, however there is a relative sign between these two channels.

Figure 3. Size of the leading logarithmic power corrections at N3^{3}LO for Drell-Yan differential in rapidity. This result accounts for all the contributions entering at leading logarithmic accuracy, which at next-to-leading power include the off-diagonal channel qgqg due to contributions of soft quarks and fermion number changing collinear splitting. For comparison, the scale variation uncertainties of the N3^{3}LO spectrum from [20] and the N3^{3}LO hard function contribution to the subtraction term are plotted, showing that these power corrections are sizable for τcut≳10−5\tau_{\mathrm{cut}} \gtrsim10^{-5}.

As we can observe from Figure 5, in this situation the power corrections would be noticeably smaller than the scale variation already starting from τcut∼10−3\tau_{\mathrm{cut}} \sim10^{-3} and completely negligible for τcut<10−4\tau_{\mathrm{cut}} < 10^{-4}. Being able to keep the slicing error under control while loosening the cuts so significantly would be a very important improvement in the feasibility and efficiency of slicing calculations at this order.

Higgs Production in Gluon Fusion

We now turn our attention to the case of Higgs production in gluon fusion. As we have seen in § 3.2.2, we have contributions from two channels. The numerical result for their sum is displayed in Figure 6, together with the N3^3LO inclusive K-factor from [17] and the N3^3LO hard function contribution [31] to the subtraction term for comparison.

Figure 6. Size of the leading logarithmic power corrections at N3^{3}LO for Higgs production in gluon fusion differential in rapidity. This result accounts for all the contributions entering at leading logarithmic accuracy, which at next-to-leading power include the off-diagonal channel gqgq due to contributions of soft quarks and fermion number changing collinear splitting. For comparison, the N3^{3}LO inclusive K-factor from [17] and the N3^{3}LO hard function [31] contribution to the subtraction term are plotted, showing that these power corrections are sizable for τcut≳10−5\tau_{\mathrm{cut}} \gtrsim10^{-5}. Note that the behaviour at very large rapidities is mainly due to the choice of normalization and will have a negligible effect in absolute terms.

For Drell-Yan we decided to show the results normalized by the leading order rapidity spectrum, since the higher order corrections are not large and therefore they do not change significantly the picture. However, in the case of Higgs production in gluon fusion it is very well known that there are sizable higher order corrections, although they are pretty flat in rapidity [10, 18, 96, 97]. For this reason, normalizing N3^{3}LO results by the LO rapidity spectrum would make all the N3^{3}LO ingredients appear unnaturally large. While this equally affects both the calculated power corrections and the ingredients we are comparing them to, and thus wouldn’t distort any numerical comparison between them, we preferred to avoid this. Hence, we present the results in this section normalized by the product of the LO rapidity spectrum and the NNLO inclusive K-factor, which gives a good proxy for the size of the NNLO rapidity distribution we are computing corrections to. Regarding the size of the LL power corrections, we see that, similarly to what we have seen in the previous section for Drell-Yan, they are quite sizable compared to the typical size of the N3^{3}LO corrections, unless one considers τcut\tau_{\mathrm{cut}} values significantly below 10−410^{-4}, which is quite challenging numerically. We also see that the dependence in rapidity is rather flat. Note that the behaviour at large rapidities is mainly caused by the vanishing of the normalization which is determined by the gluon PDF at large xx. This has an irrelevant effect in absolute terms since in that region there is very little cross section.

For completeness, in Figure 7 we show the channel decomposition of our result. We see that the contribution from the diagonal gggg channel is by far the most dominant one being more than an order of magnitude larger than the gqgq channel. This is in contrast with what we have seen in § 4.1 in the case of Drell-Yan where the off-diagonal channel was larger. Note that the gqgq channel enters at the same logarithmic order as the diagonal channel and their relative size is most likely explained by the gluon PDF being large in the typical kinematic regions probed in Higgs production at the LHC.

Figure 7. Same as Figure 3 but for different production channels. We show the diagonal channel qqˉq\bar{q} on the left and qgqg on the right. Both channels contribute at leading log at NLP. It is evident that the power corrections to the diagonal channel gggg are the only one numerically relevant, since they constitute more than 95% of the total contribution. Note that the behaviour at very large rapidities is mainly due to the choice of normalization and will have a negligible effect in absolute terms.

Finally, in Figure 8 we repeat the exercise of estimating the size of the residual slicing uncertainties after including the results of § 3.2.2 in the subtraction term, assuming a similar size between the coefficients of the logarithms at NLP. We observe that the inclusion of these terms analytically alleviates the impact of slicing uncertainties and could allow for obtaining results with acceptable uncertainties with τcut\tau_{\mathrm{cut}} not too far from 10−310^{-3}. If that turns out to be the case, it will be then interesting to see if this could roughly compete in terms of numerical efficiency with calculations purely based on P2B such as the one of ref. [18].

Figure 8. Estimate of the slicing error Δσ(τcut,Y)\Delta\sigma(\tau_{\mathrm{cut}},Y) after the removal of the leading logarithmic terms analytically calculated in this work (i.e. after their inclusion in the subtraction term) for various values of τcut\tau_{\mathrm{cut}}. Plotted for comparison are the inclusive N3^3LO K-factor and the N3^3LO hard function contribution to the subtraction term.

Conclusions

In this work, we calculated the leading logarithmic dynamical power corrections to the 0-jettiness distribution at next-to-next-to-next-to-leading order, differential in rapidity and invariant mass, for Drell-Yan and Higgs production in gluon fusion. We have presented a method to disentangle dynamical power corrections, tied to QCD dynamics beyond leading power, from those arising from fiducial cuts using Projection to Born-improved subtractions. From the calculation of these power corrections, we can estimate that if only the leading power terms in the 0-jettiness distribution are included in the subtraction term, it would be necessary to push the slicing parameter τcut\tau_{\mathrm{cut}} to values below 10−510^{-5} to obtain results with slicing errors at few per-mille level. On the other hand, including these leading logarithmic terms in the subtraction can substantially improve the situation. We estimate that removing this leading source of power corrections can yield a numerical improvement at N3^3LO comparable to reducing the slicing parameter by O(50)O(50) in the cases where there are no accidental large differences between the LL NLP coefficients and the ones multiplying the subleading logarithms. This is the case at NLO and NNLO for Drell-Yan and gluon-fusion Higgs production at the LHC [45, 46, 47]. Therefore, we believe that with the inclusion of the leading logarithmic corrections and using the projection to Born improvement for the treatment of fiducial power corrections, it should be possible to perform fully differential N3^3LO calculations with τcut∼10−4−10−3\tau_{\mathrm{cut}} \sim10^{-4} - 10^{-3} with slicing uncertainties under control. This advancement opens the door to N3^3LO predictions based on slicing with significantly enhanced efficiency, crucial for broadening the availability and utility of these predictions in precision phenomenology studies at current and future colliders.

Acknowledgments

We thank Alexander Huss for numerous discussions throughout this work, feedback on the manuscript, and for providing the numerical results from ref. [20]. We are grateful to Iain Stewart, HuaXing Zhu, and Ian Moult for discussions which initiated this project, to Alessandro Broggio and Simone Alioli for insightful discussions, and to Pier Monni for helpful comments on this work and feedback on the manuscript. We thank the Erwin-Schrödinger International Institute for Mathematics and Physics at the University of Vienna for partial support during the Programme “Quantum Field Theory at the Frontiers of the Strong Interactions”. We thank the Galileo Galilei Institute for Theoretical Physics for the hospitality and the INFN for partial support during the completion of this work.

References

  1. [1]S. Catani and M. Grazzini, An NNLO subtraction formalism in hadron collisions and its application to Higgs boson production at the LHC, Phys. Rev. Lett. 98 (2007) 222002, [hep-ph/0703012].DOI
  2. [2]F. Caola, K. Melnikov and R. Röntsch, Nested soft-collinear subtractions in NNLO QCD computations, Eur. Phys. J. C 77 (2017) 248, [arXiv:1702.01352].DOI
  3. [3]A. Gehrmann-De Ridder, T. Gehrmann and E. W. N. Glover, Antenna subtraction at NNLO, JHEP 09 (2005) 056, [hep-ph/0505111].
  4. [4]M. Czakon, A novel subtraction scheme for double-real radiation at NNLO, Phys. Lett. B 693 (2010) 259–268, [arXiv:1005.0274].DOI
  5. [5]R. Boughezal, K. Melnikov and F. Petriello, A subtraction scheme for NNLO computations, Phys. Rev. D 85 (2012) 034025, [arXiv:1111.7041].DOI
  6. [6]M. Czakon and D. Heymes, Four-dimensional formulation of the sector-improved residue subtraction scheme, Nucl. Phys. B 890 (2014) 152–227, [arXiv:1408.2500].DOI
  7. [7]R. Boughezal, C. Focke, W. Giele, X. Liu and F. Petriello, Higgs boson production in association with a jet at NNLO using jettiness subtraction, Phys. Lett. B 748 (2015) 5–8, [arXiv:1505.03893].arxiv.org/abs/1505.03893
  8. [8]J. Gaunt, M. Stahlhofen, F. J. Tackmann and J. R. Walsh, N-jettiness Subtractions for NNLO QCD Calculations, JHEP 09 (2015) 058, [arXiv:1505.04794].arxiv.org/abs/1505.04794
  9. [9]V. Del Duca, C. Duhr, A. Kardos, G. Somogyi, Z. Szőr, Z. Trócsányi et al., Jet production in the CoLoRFulNNLO method: event shapes in electron-positron collisions, Phys. Rev. D 94 (2016) 074019, [arXiv:1606.03453].DOI
  10. [10]F. Dulat, S. Lionetti, B. Mistlberger, A. Pelloni and C. Spe cchia, Higgs-differential cross section at NNLO in dimensional regularisation, JHEP 07 (2017) 017, [arXiv:1704.08220].DOI
  11. [11]L. Magnea, E. Maina, G. Pelliccioli, C. Signorile-Signorile, P. Torrielli and S. Uccirati, Local analytic sector subtraction at NNLO, JHEP 12 (2018) 107, [arXiv:1806.09570]. [Erratum: JHEP 06, 013 (2019)].DOI
  12. [12]F. Devoto, K. Melnikov, R. Röntsch, C. Signorile-Signorile and D. M. Tagliabue, A fresh look at the nested soft-collinear subtraction scheme: NNLO QCD corrections to N-gluon final states in qq̄ annihilation, arXiv:2310.17598.arxiv.org/abs/2310.17598
  13. [13]T. Gehrmann, E. W. N. Glover and M. Marcoli, The colourful antenna subtraction method, arXiv:2310.19757.DOI
  14. [14]C. Anastasiou, C. Duhr, F. Dulat, E. Furlan, T. Gehrmann, F. Herzog et al., High precision determination of the gluon fusion Higgs boson cross-section at the LHC, JHEP 05 (2016) 058, [arXiv:1602.00695].arxiv.org/abs/1602.00695
  15. [15]B. Mistlberger, Higgs boson production at hadron colliders at N³LO in QCD, JHEP 05 (2018) 028, [arXiv:1802.00833].DOI
  16. [16]C. Duhr, F. Dulat and B. Mistlberger, Higgs Boson Production in Bottom-Quark Fusion to Third Order in the Strong Coupling, Phys. Rev. Lett. 125 (2020) 051804, [arXiv:1904.09990].DOI
  17. [17]J. Baglio, C. Duhr, B. Mistlberger and R. Szafron, Inclusive production cross sections at N³LO, JHEP 12 (2022) 066, [arXiv:2209.06138].DOI
  18. [18]X. Chen, T. Gehrmann, E. W. N. Glover, A. Huss, B. Mistlberger and A. Pelloni, Fully Differential Higgs Boson Production to Third Order in QCD, Phys. Rev. Lett. 127 (2021) 072002, [arXiv:2102.07607].DOI
  19. [19]G. Billis, B. Dehnadi, M. A. Ebert, J. K. L. Michel and F. J. Tackmann, Higgs pT Spectrum and Total Cross Section with Fiducial Cuts at Third Resummed and Fixed Order in QCD, Phys. Rev. Lett. 127 (2021) 072001, [arXiv:2102.08039].DOI
  20. [20]X. Chen, T. Gehrmann, N. Glover, A. Huss, T.-Z. Yang and H. X. Zhu, Dilepton Rapidity Distribution in Drell-Yan Production to Third Order in QCD, Phys. Rev. Lett. 128 (2022) 052001, [arXiv:2107.09085].DOI
  21. [21]X. Chen, T. Gehrmann, E. W. N. Glover, A. Huss, P. F. Monni, E. Re et al., Third-Order Fiducial Predictions for Drell-Yan Production at the LHC, Phys. Rev. Lett. 128 (2022) 252001, [arXiv:2203.01565].DOI
  22. [22]X. Chen, T. Gehrmann, N. Glover, A. Huss, T.-Z. Yang and H. X. Zhu, Differential N3LO QCD corrections to charged current production at the LHC, PoS LL2022 (2022) 004, [arXiv:2207.08584].arxiv.org/abs/2207.08584
  23. [23]X. Chen, T. Gehrmann, N. Glover, A. Huss, T.-Z. Yang and H. X. Zhu, Transverse mass distribution and charge asymmetry in W boson production to third order in QCD, Phys. Lett. B 840 (2023) 137876, [arXiv:2205.11426].DOI
  24. [24]T. Neumann and J. Campbell, Fiducial Drell-Yan production at the LHC improved by transverse-momentum resummation at N4LLp+N3LO, Phys. Rev. D 107 (2023) L011506, [arXiv:2207.07056].DOI
  25. [25]J. Campbell and T. Neumann, Third order QCD predictions for fiducial W-boson production, JHEP 11 (2023) 127, [arXiv:2308.15382].DOI
  26. [26]E. Farhi, A QCD Test for Jets, Phys. Rev. Lett. 39 (1977) 1587–1588.
  27. [27]I. W. Stewart, F. J. Tackmann and W. J. Waalewijn, Factorization at the LHC: From PDFs to Initial State Jets, Phys. Rev. D 81 (2010) 094035, [arXiv:0910.0467].DOI
  28. [28]I. W. Stewart, F. J. Tackmann and W. J. Waalewijn, N-Jettiness: An Inclusive Event Shape to Veto Jets, Phys. Rev. Lett. 105 (2010) 092002, [arXiv:1004.2489].DOI
  29. [29]M.-x. Luo, T.-Z. Yang, H. X. Zhu and Y. J. Zhu, Quark Transverse Parton Distribution at the Next-to-Next-to-Next-to-Leading Order, Phys. Rev. Lett. 124 (2020) 092001, [arXiv:1912.05778].DOI
  30. [30]M. A. Ebert, B. Mistlberger and G. Vita, Transverse momentum dependent PDFs at N³LO, JHEP 09 (2020) 146, [arXiv:2006.05329].
  31. [31]T. Gehrmann, E. W. N. Glover, T. Huber, N. Ikizlerli and C. Studerus, Calculation of the quark and gluon form factors to three loops in QCD, JHEP 06 (2010) 094, [arXiv:1004.3653].DOI
  32. [32]Y. Li and H. X. Zhu, Bootstrapping Rapidity Anomalous Dimensions for Transverse-Momentum Resummation, Phys. Rev. Lett. 118 (2017) 022004, [arXiv:1604.01404].DOI
  33. [33]C. W. Bauer, S. Fleming and M. E. Luke, Summing Sudakov logarithms in B → X_sγ in effective field theory., Phys. Rev. D 63 (2000) 014006, [hep-ph/0005275].DOI
  34. [34]C. W. Bauer, S. Fleming, D. Pirjol and I. W. Stewart, An Effective field theory for collinear and soft gluons: Heavy to light decays, Phys. Rev. D 63 (2001) 114020, [hep-ph/0011336].DOI
  35. [35]C. W. Bauer and I. W. Stewart, Invariant operators in collinear effective theory, Phys. Lett. B 516 (2001) 134–142, [hep-ph/0107001].DOI
  36. [36]C. W. Bauer, D. Pirjol and I. W. Stewart, Soft collinear factorization in effective field theory, Phys. Rev. D 65 (2002) 054022, [hep-ph/0109045].DOI
  37. [37]C. W. Bauer, S. Fleming, D. Pirjol, I. Z. Rothstein and I. W. Stewart, Hard scattering factorization from effective field theory, Phys. Rev. D 66 (2002) 014017, [hep-ph/0202088].DOI
  38. [38]R. Brüser, Z. L. Liu and M. Stahlhofen, Three-Loop Quark Jet Function, Phys. Rev. Lett. 121 (2018) 072003, [arXiv:1804.09722].DOI
  39. [39]M. A. Ebert, B. Mistlberger and G. Vita, N-jettiness beam functions at N^3LO, JHEP 09 (2020) 143, [arXiv:2006.03056].DOI
  40. [40]D. Baranowski, A. Behring, K. Melnikov, L. Tancredi and C. Wever, Beam functions for N-jettiness at N^3LO in perturbative QCD, JHEP 02 (2023) 073, [arXiv:2211.05722].DOI
  41. [41]D. Baranowski, NNLO zero-jettiness beam and soft functions to higher orders in the dimensional-regularization parameter ε, Eur. Phys. J. C 80 (2020) 523, [arXiv:2004.03285].DOI
  42. [42]D. Baranowski, M. Delto, K. Melnikov and C.-Y. Wang, On phase-space integrals with Heaviside functions, JHEP 02 (2022) 081, [arXiv:2111.13594].DOI
  43. [43]D. Baranowski, M. Delto, K. Melnikov and C.-Y. Wang, Same-hemisphere three-gluon-emission contribution to the zero-jettiness soft function at N^3LO QCD, Phys. Rev. D 106 (2022) 014004, [arXiv:2204.09459].DOI
  44. [44]W. Chen, F. Feng, Y. Jia and X. Liu, Double-real-virtual and double-virtual-real corrections to the three-loop thrust soft function, JHEP 22 (2020) 094, [arXiv:2206.12323].
  45. [45]I. Moult, L. Rothen, I. W. Stewart, F. J. Tackmann and H. X. Zhu, Subleading Power Corrections for N-Jettiness Subtractions, Phys. Rev. D 95 (2017) 074023, [arXiv:1612.00450].DOI
  46. [46]I. Moult, L. Rothen, I. W. Stewart, F. J. Tackmann and H. X. Zhu, N-jettiness subtractions for gg → H at subleading power, Phys. Rev. D 97 (2018) 014013, [arXiv:1710.03227].DOI
  47. [47]M. A. Ebert, I. Moult, I. W. Stewart, F. J. Tackmann, G. Vita and H. X. Zhu, Power Corrections for N-Jettiness Subtractions at O(α_s), JHEP 12 (2018) 084, [arXiv:1807.10764].DOI
  48. [48]I. Moult, I. W. Stewart, F. J. Tackmann and W. J. Waalewijn, Employing Helicity Amplitudes for Resummation, Phys. Rev. D 93 (2016) 094003, [arXiv:1508.02397].DOI
  49. [49]D. Bonocore, E. Laenen, L. Magnea, S. Melville, L. Vernazza and C. D. White, A factorization approach to next-to-leading-power threshold logarithms, JHEP 06 (2015) 008, [arXiv:1503.05156].arxiv.org/abs/1503.05156
  50. [50]D. Bonocore, E. Laenen, L. Magnea, L. Vernazza and C. D. White, Non-abelian factorisation for next-to-leading-power threshold logarithms, JHEP 12 (2016) 121, [arXiv:1610.06842].DOI
  51. [51]D. W. Kolodrubetz, I. Moult and I. W. Stewart, Building Blocks for Subleading Helicity Operators, JHEP 05 (2016) 139, [arXiv:1601.02607].DOI
  52. [52]M. Beneke, M. Garny, R. Szafron and J. Wang, Anomalous dimension of subleading-power N-jet operators, JHEP 03 (2018) 001, [arXiv:1712.04416].DOI
  53. [53]M. Beneke, M. Garny, R. Szafron and J. Wang, Subleading-power N-jet operators and the LBK amplitude in SCET, PoS RADCOR2017 (2017) 048, [arXiv:1712.07462].arxiv.org/abs/1712.07462
  54. [54]I. Moult, I. W. Stewart and G. Vita, A subleading operator basis and matching for gg → H, JHEP 07 (2017) 067, [arXiv:1703.03408].DOI
  55. [55]C.-H. Chang, I. W. Stewart and G. Vita, A Subleading Power Operator Basis for the Scalar Quark Current, JHEP 04 (2018) 041, [arXiv:1712.04343].DOI
  56. [56]I. Feige, D. W. Kolodrubetz, I. Moult and I. W. Stewart, A Complete Basis of Helicity Operators for Subleading Factorization, JHEP 11 (2017) 142, [arXiv:1703.03411].DOI
  57. [57]M. Beneke, M. Garny, R. Szafron and J. Wang, Anomalous dimension of subleading-power N-jet operators. Part II, JHEP 11 (2018) 112, [arXiv:1808.04742].DOI
  58. [58]M. Beneke, M. Garny, R. Szafron and J. Wang, Violation of the Kluberg-Stern-Zuber theorem in SCET, JHEP 09 (2019) 101, [arXiv:1907.05463].DOI
  59. [59]I. Moult, I. W. Stewart and G. Vita, Subleading Power Factorization with Radiative Functions, JHEP 11 (2019) 153, [arXiv:1905.07411].DOI
  60. [60]R. Boughezal, A. Isgrò and F. Petriello, Next-to-leading-logarithmic power corrections for N-jettiness subtraction in color-singlet production, Phys. Rev. D 97 (2018) 076006, [arXiv:1802.00456].DOI
  61. [61]R. Brüser, S. Caron-Huot and J. M. Henn, Subleading Regge limit from a soft anomalous dimension, JHEP 04 (2018) 047, [arXiv:1802.02524].DOI
  62. [62]I. Moult, I. W. Stewart, G. Vita and H. X. Zhu, First Subleading Power Resummation for Event Shapes, JHEP 08 (2018) 013, [arXiv:1804.04665].DOI
  63. [63]M. Beneke, A. Broggio, M. Garny, S. Jaskiewicz, R. Szafron, L. Vernazza et al., Leading-logarithmic threshold resummation of the Drell-Yan process at next-to-leading power, JHEP 03 (2019) 043, [arXiv:1809.10631].DOI
  64. [64]M. A. Ebert, I. Moult, I. W. Stewart, F. J. Tackmann, G. Vita and H. X. Zhu, Subleading power rapidity divergences and power corrections for qT, JHEP 04 (2019) 123, [arXiv:1812.08189].DOI
  65. [65]I. Moult, I. W. Stewart, G. Vita and H. X. Zhu, The Soft Quark Sudakov, JHEP 05 (2020) 089, [arXiv:1910.14038].DOI
  66. [66]M. Beneke, M. Garny, S. Jaskiewicz, R. Szafron, L. Vernazza and J. Wang, Leading-logarithmic threshold resummation of Higgs production in gluon fusion at next-to-leading power, JHEP 01 (2020) 094, [arXiv:1910.12685].DOI
  67. [67]L. Buonocore, M. Grazzini and F. Tramontano, The qT subtraction method: electroweak corrections and power suppressed contributions, Eur. Phys. J. C 80 (2020) 254, [arXiv:1911.10166].DOI
  68. [68]M. van Beekveld, W. Beenakker, E. Laenen and C. D. White, Next-to-leading power threshold effects for inclusive and exclusive processes with final state jets, JHEP 03 (2020) 106, [arXiv:1905.08741].DOI
  69. [69]M. Beneke, A. Broggio, S. Jaskiewicz and L. Vernazza, Threshold factorization of the Drell-Yan process at next-to-leading power, JHEP 07 (2020) 078, [arXiv:1912.01585].DOI
  70. [70]C. Anastasiou and A. Penin, Light Quark Mediated Higgs Boson Threshold Production in the Next-to-Leading Logarithmic Approximation, JHEP 07 (2020) 195, [arXiv:2004.03602]. [Erratum: JHEP 01, 164 (2021)].DOI
  71. [71]E. Laenen, J. Sinninghe Damsté, L. Vernazza, W. Waalewijn and L. Zoppi, Towards all-order factorization of QED amplitudes at next-to-leading power, Phys. Rev. D 103 (2021) 034022, [arXiv:2008.01736].DOI
  72. [72]G. Vita, QCD Beyond Leading Power. PhD thesis, MIT, 9, 2020. arXiv:2008.10606.arxiv.org/abs/2008.10606
  73. [73]Z. L. Liu, B. Mecaj, M. Neubert and X. Wang, Factorization at subleading power and endpoint divergences in h → γγ decay. Part II. Renormalization and scale evolution, JHEP 01 (2021) 077, [arXiv:2009.06779].DOI
  74. [74]A. Broggio, S. Jaskiewicz and L. Vernazza, Next-to-leading power two-loop soft functions for the Drell-Yan process at threshold, JHEP 10 (2021) 061, [arXiv:2107.07353].DOI
  75. [75]T. Liu, S. Modi and A. A. Penin, Higgs boson production and quark scattering amplitudes at high energy through the next-to-next-to-leading power in quark mass, JHEP 02 (2022) 170, [arXiv:2111.01820].DOI
  76. [76]Z. L. Liu, M. Neubert, M. Schnubel and X. Wang, Factorization at next-to-leading power and endpoint divergences in gg → h production, JHEP 06 (2023) 183, [arXiv:2212.10447].DOI
  77. [77]A. Broggio, S. Jaskiewicz and L. Vernazza, Threshold factorization of the Drell-Yan quark-gluon channel and two-loop soft function at next-to-leading power, JHEP 12 (2023) 028, [arXiv:2306.06037].DOI
  78. [78]I. W. Stewart, F. J. Tackmann and W. J. Waalewijn, The Beam Thrust Cross Section for Drell-Yan at NNLL Order, Phys. Rev. Lett. 106 (2011) 032001, [arXiv:1005.4060].DOI
  79. [79]T. T. Jouttenus, I. W. Stewart, F. J. Tackmann and W. J. Waalewijn, The Soft Function for Exclusive N-Jet Production at Hadron Colliders, Phys. Rev. D 83 (2011) 114030, [arXiv:1102.4344].DOI
  80. [80]C. F. Berger, C. Marcantonini, I. W. Stewart, F. J. Tackmann and W. J. Waalewijn, Higgs Production with a Central Jet Veto at NNLL+NNLO, JHEP 04 (2011) 092, [arXiv:1012.4480].DOI
  81. [81]A. V. Manohar, T. Mehen, D. Pirjol and I. W. Stewart, Reparameterization invariance for collinear operators, Phys. Lett. B539 (2002) 59–66, [hep-ph/0204229].DOI
  82. [82]M. A. Ebert and F. J. Tackmann, Impact of isolation and fiducial cuts on qT and N-jettiness subtractions, JHEP 03 (2020) 158, [arXiv:1911.08486].DOI
  83. [83]S. Catani, D. de Florian, G. Ferrera and M. Grazzini, Vector boson production at hadron colliders: transverse-momentum resummation and leptonic decay, JHEP 12 (2015) 047, [arXiv:1507.06937].DOI
  84. [84]M. A. Ebert, J. K. L. Michel, I. W. Stewart and F. J. Tackmann, Drell-Yan q_T resummation of fiducial power corrections at N^3LL, JHEP 04 (2021) 102, [arXiv:2006.11382].DOI
  85. [85]S. Camarda, L. Cieri and G. Ferrera, Fiducial perturbative power corrections within the q_T subtraction formalism, Eur. Phys. J. C 82 (2022) 575, [arXiv:2111.14509].DOI
  86. [86]L. Buonocore, S. Kallweit, L. Rottoli and M. Wiesemann, Linear power corrections for two-body kinematics in the q_T subtraction formalism, Phys. Lett. B 829 (2022) 137118, [arXiv:2111.13661].DOI
  87. [87]M. Cacciari, F. A. Dreyer, A. Karlberg, G. P. Salam and G. Zanderighi, Fully Differential Vector-Boson-Fusion Higgs Production at Next-to-Next-to-Leading Order, Phys. Rev. Lett. 115 (2015) 082002, [arXiv:1506.02660]. [Erratum: Phys.Rev.Lett. 120, 139901 (2018)].
  88. [88]G. P. Salam and E. Slade, Cuts for two-body decays at colliders, JHEP 11 (2021) 220, [arXiv:2106.08329].DOI
  89. [89]T. Gehrmann, M. Jaquier, E. W. N. Glover and A. Kou kout sakis, Two-Loop QCD Corrections to the Helicity Amplitudes for H → 3 partons, JHEP 02 (2012) 056, [arXiv:1112.3554].DOI
  90. [90]L. W. Garland, T. Gehrmann, E. W. N. Glover, A. Koukoutsakis and E. Remiddi, The Two loop QCD matrix element for e+ e- --> 3 jets, Nucl. Phys. B 627 (2002) 107–188, [hep-ph/0112081].DOI
  91. [91]L. W. Garland, T. Gehrmann, E. W. N. Glover, A. Koukoutsakis and E. Remiddi, Two loop QCD helicity amplitudes for e+ e- --> three jets, Nucl. Phys. B 642 (2002) 227–262, [hep-ph/0206067].DOI
  92. [92]A. Bhattacharya, I. Moult, I. W. Stewart and G. Vita, Helicity Methods for High Multiplicity Subleading Soft and Collinear Limits, JHEP 05 (2019) 192, [arXiv:1812.06950].DOI
  93. [93]R. Boughezal, A. Isgrò and F. Petriello, Next-to-leading power corrections to V + 1 jet production in N-jettiness subtraction, Phys. Rev. D 101 (2020) 016005, [arXiv:1907.12213].DOI
  94. [94]S. Alekhin, A. Kardos, S. Moch and Z. Trócsányi, Precision studies for Drell-Yan processes at NNLO, Eur. Phys. J. C 81 (2021) 573, [arXiv:2104.02400].DOI
  95. [95]J. M. Campbell, R. K. Ellis and S. Seth, Non-local slicing approaches for NNLO QCD in MCFM, JHEP 06 (2022) 002, [arXiv:2202.07738].DOI
  96. [96]C. Anastasiou, K. Melnikov and F. Petriello, Higgs boson production at hadron colliders: Differential cross sections through next-to-next-to-leading order, Phys. Rev. Lett. 93 (2004) 262002, [hep-ph/0409088].DOI
  97. [97]F. Dulat, B. Mistlberger and A. Pelloni, Differential Higgs production at N^3LO beyond threshold, JHEP 01 (2018) 145, [arXiv:1710.03016].DOI

Paper details

vlnc:W_qmefzjdx
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents