Two-loop leading-color QCD corrections for Higgs plus two-jet production in the heavy-top limit

This paper presents the leading-color two-loop QCD corrections for Higgs-plus-two-jet production in the heavy-top limit, providing analytic expressions for finite helicity amplitudes derived via numerical unitarity and a novel partial fraction decomposition algorithm, alongside a stable numerical implementation and an analysis of their singularity structure.

The Gist

Imagine the Large Hadron Collider (LHC) as a giant, high-speed particle smasher. Physicists use it to smash protons together to find the Higgs boson, a particle that gives other particles mass. But finding the Higgs is like finding a needle in a haystack, because the "haystack" is full of other messy debris from the collisions.

One of the biggest sources of this messy debris is a process where two jets of particles (like tiny, fast-moving streams of debris) are created alongside the Higgs boson. To distinguish the "real" Higgs signal from this background noise, physicists need to predict exactly how often this background happens. To do that, they need to perform incredibly complex math calculations called "two-loop corrections."

This paper is about a team of physicists successfully solving a massive, previously unsolvable piece of this math puzzle. Here is how they did it, explained simply:

1. The Problem: A Math Mountain Too High to Climb

Calculating how particles interact is like trying to predict the exact path of a million billiard balls bouncing off each other at the speed of light.

• The "Heavy Top" Shortcut: In the real world, the Higgs talks to a heavy particle called the "top quark." But because the top quark is so heavy, the physicists decided to pretend it doesn't exist as a separate particle in their calculations. Instead, they replaced the top quark with a "magic button" (an effective interaction) that the Higgs can press to create gluons (the particles that make up jets). This is called the "Heavy Top Limit." It's a simplification, but a very good one for the energies they are studying.
• The "Leading Color" Filter: Quantum physics has a property called "color" (unrelated to visual color). There are billions of ways these interactions can happen. The team decided to focus only on the most dominant, "loudest" versions of these interactions (the "leading color" ones) and ignore the quiet, rare ones. This made the mountain of math slightly smaller, but it was still a towering peak.

2. The Method: Building a Map from Dots

The team didn't try to solve the whole mountain of math at once. Instead, they used a clever strategy called Numerical Unitarity.

• The "Black Box" Approach: Imagine you have a complex machine (the particle collision) and you want to know how it works, but you can't see inside. Instead of taking it apart, you feed it thousands of different inputs (different collision speeds and angles) and record the outputs.
• The "Finite Field" Trick: To do this math on a computer, they didn't use normal numbers (like 1, 2, 3). They used "finite fields," which are like a clock system where numbers wrap around after a certain point (like a clock that only goes up to 17). This prevents the computer from getting confused by messy decimals and allows for incredibly fast, precise calculations.
• The "Reconstruction" Puzzle: They generated millions of these "dots" (numerical results). Then, they used a computer algorithm to look at the pattern of the dots and ask: "What single, clean mathematical formula connects all these dots?" This is like looking at a scatter plot of points and drawing the perfect curve through them.

3. The Breakthrough: New Tools for a New Puzzle

Usually, when you try to draw that perfect curve through millions of points, the formula becomes so huge and messy that no human could ever read it. The team invented new tools to keep the formula clean:

• The "Simplification Filter": They realized that many parts of the math were redundant. They developed a new algorithm to strip away the unnecessary clutter, finding the simplest possible version of the formula. It's like taking a tangled ball of yarn and finding the single straight thread running through it.
• The "Slice" Technique: Instead of trying to understand the whole 3D shape of the math at once, they sliced it into 2D pieces (like slicing a loaf of bread). They figured out the pattern of each slice and then glued the slices back together to form the whole loaf. This allowed them to reconstruct the full, complex formula without getting overwhelmed.

4. The Discovery: A Hidden "Bump" in the Road

While solving the math, they found something surprising.

• The "Anomalous Threshold": In physics, there are usually "thresholds" where things change behavior, often because a heavy particle is being created. The team found a new kind of threshold that happens even when no heavy particles are being created. It's like driving down a smooth road and suddenly hitting a tiny, invisible speed bump that changes how the car vibrates, even though the road looks perfectly flat.
• Why it matters: This "bump" (a mathematical discontinuity in the slope of the curve) was predicted by theory but had never been seen in this specific type of particle collision before. It confirms that our understanding of how these particles interact is deep and subtle.

5. The Result: A Ready-to-Use Tool

The team didn't just write down the answer; they built a software library (a C++ tool) that other scientists can use immediately.

• Speed: They made sure the tool is fast. It can calculate the result for a single collision scenario in about 1 to 10 seconds.
• Stability: They tested it to make sure it doesn't crash or give wrong answers, even in tricky situations.

Summary

In short, this paper is about a team of physicists who:

1. Simplified a complex particle physics problem by using smart approximations.
2. Used a "guess-and-check" method with millions of computer-generated numbers to find the hidden mathematical formula.
3. Invented new ways to keep that formula simple and readable.
4. Discovered a strange, new "speed bump" in the math of particle collisions.
5. Packaged all of this into a fast, reliable software tool so other scientists can use it to predict what will happen at the Large Hadron Collider with extreme precision.

They didn't just solve a math problem; they built the engine that will help future experiments understand the Higgs boson better.

Technical Summary

Technical Summary: Two-loop Leading-Color QCD Corrections for Higgs plus Two-Jet Production

Problem Statement
The precise determination of Higgs boson properties at the Large Hadron Collider (LHC) requires accurate theoretical predictions for Higgs production in association with two jets ($Hjj$). While the Vector Boson Fusion (VBF) channel is well understood, the dominant irreducible QCD background—Higgs production via gluon fusion ($ggF Hjj$)—remains challenging to compute at high orders. This process is loop-induced, mediated by top-quark loops, and requires the calculation of two-loop five-point amplitudes with internal masses. These calculations are currently beyond the reach of standard computational methods for full mass dependence. Consequently, this work focuses on the heavy-top limit (HTL) effective field theory, where the top quark is integrated out, and the leading-color approximation (LCA), where $N_c \to \infty$. The goal is to compute the double-virtual Next-to-Next-to-Leading Order (NNLO) QCD corrections for $pp \to Hjj$ in this regime.

Methodology
The authors employ a hybrid numerical-analytic approach to compute the finite remainders of the helicity amplitudes:

1. Numerical Unitarity (Caravel): The numerical evaluation of the two-loop amplitudes is performed using the Caravel framework. This implements the numerical unitarity method, where generalized unitarity cuts are matched to an integrand parametrization of master and surface terms.

   • The authors extended Caravel to the heavy-top effective theory by incorporating effective $ggH$, $gggH$, and $ggggH$ vertices.
   • They utilized dimensional reduction to handle the state-counting parameter $D_s$ analytically, significantly reducing computational costs compared to dimensional reconstruction.
   • The calculation involves non-planar topologies (specifically HexaBox and Double-Pentagon families) which are unique to the HTL even in the leading-color limit.

2. Finite-Field Sampling: The amplitudes are evaluated numerically over finite fields. The results are expressed as coefficients of a basis of canonical five-point one-mass pentagon functions.

3. Analytic Reconstruction: The core innovation lies in the reconstruction of the analytic form of the rational coefficient functions from these numerical samples. The authors refined existing functional reconstruction techniques through several key algorithmic advances:

   • Improved Rational Bases: They identified linear dependencies among coefficient functions using univariate phase-space slices to find a basis with reduced numerator degrees.
   • Bivariate Slice Partial Fraction Decomposition (PFD): A new algorithm was introduced to construct multivariate PFDs. Instead of relying on empirical data or singular limits, they perform functional reconstruction on a generic bivariate slice of the phase space. This allows them to systematically determine ideal intersections and construct compact partial-fraction ansätze directly from numerical data.
   • Efficient Lifting: The reconstruction of the basis change matrix and the PFD coefficients is performed in a single finite field, with the final lift to rational numbers requiring only a minimal number of additional phase-space evaluations.

Key Contributions and Results

• First Analytic Results: The paper presents the first analytic expressions for the two-loop helicity amplitudes in all partonic channels contributing to $ggF Hjj$ in the heavy-top limit: the four-gluon, two-quark-two-gluon, and four-quark channels.
• Compact Representations: The finite remainders are expressed in terms of one-mass pentagon functions with spinor-helicity coefficients. The authors demonstrate that the rational coefficients possess a surprisingly simple structure: their denominator factors are identical (up to permutations) to those found in planar amplitudes, despite the presence of non-planar diagrams.
• Anomalous Threshold: The analysis reveals a new type of non-analytic behavior. While the finite remainders are continuous across the surface $\Sigma_5^{(3)} = 0$ (a variety associated with non-planar integrals), their first derivatives exhibit a cusp (a discontinuity in the second derivative). This constitutes an "anomalous" threshold occurring at non-degenerate physical momentum configurations, a phenomenon not previously reported in massless scattering.
• Numerical Implementation: The authors provide a stable and efficient C++ library for the numerical evaluation of these results. The implementation achieves evaluation times ranging from 1 to 10 seconds per phase-space point, with high numerical stability (99% of points in the most difficult channel evaluated to 2 digits of precision in double precision).
• Validation: The results are validated through internal consistency checks (UV/IR pole cancellation, scale dependence), comparison with independent one-loop calculations, and agreement with existing two-loop results for specific channels in the literature.

Significance
The authors state that these results provide the necessary ingredients for NNLO QCD predictions for Higgs plus two-jet production at hadron colliders. Furthermore, combined with existing three-loop results for inclusive Higgs production, these amplitudes enable N$^3$LO predictions for Higgs production in gluon fusion. The work also offers a rigorous testbed for the correspondence between five-point QCD amplitudes and four-point form factors in $\mathcal{N}=4$ Super Yang-Mills theory. The methodological advances, particularly the bivariate slice approach to partial fraction decomposition, are presented as process-independent tools that can be applied to other high-multiplicity two-loop amplitude calculations, potentially extending to subleading-color contributions.