Implementation of the Martini-Ericson-Chanfray-Marteau RPA-based neutrino and antineutrino cross-section model in the GENIE neutrino event generator

This paper presents the first implementation and validation of the Martini-Ericson-Chanfray-Marteau RPA-based model for quasielastic and multinucleon neutrino and antineutrino interactions within the GENIE event generator, demonstrating reasonable agreement with experimental data from T2K and MicroBooNE.

The Gist

Imagine you are trying to predict exactly how a billiard ball will bounce when it hits a cluster of other balls glued together on a table. In the world of physics, this is similar to trying to predict what happens when a neutrino (a tiny, ghost-like particle) smashes into an atomic nucleus (a cluster of protons and neutrons).

For decades, scientists have struggled to get this math right. The nucleus isn't just a static pile of balls; it's a chaotic, quantum "dance floor" where particles interact in complex ways. If you get the math wrong, you can't accurately measure the properties of the neutrino, which is crucial for understanding the universe.

Here is what this paper does, broken down simply:

1. The Problem: A Missing Piece of the Puzzle

Scientists use computer programs called "event generators" (like GENIE) to simulate these neutrino collisions. Think of GENIE as a video game engine that tries to predict the outcome of every crash.

However, for a long time, these programs were missing a key rule of the game. When a neutrino hits a nucleus, it doesn't just knock out one particle (like a single billiard ball). Sometimes, it knocks out a team of particles at once. The paper calls these "multinucleon" excitations (specifically 2p2h and 3p3h, which just means 2 or 3 protons/neutrons getting kicked out together).

Previous models ignored this "team kick" or handled it poorly. This led to big errors in predicting how much energy the neutrino had, which messed up experiments trying to study neutrino oscillations (how they change types).

2. The Solution: Installing a New "Physics Engine"

The authors of this paper took a very sophisticated mathematical model created by a team in Lyon, France (the Martini-Ericson-Chanfray-Marteau model) and successfully installed it into the GENIE computer program.

Think of the GENIE program as a car. Before this paper, the car had an engine that was good at driving on straight roads (simple collisions) but struggled on bumpy terrain (complex collisions). The authors took a brand-new, high-performance engine (the Lyon model) and bolted it into the car.

• What the new engine does: It calculates the probability of the neutrino hitting the nucleus and knocking out either a single particle or a whole group of them. It uses a method called "Random Phase Approximation" (RPA), which is like a highly detailed map of how the particles inside the nucleus wiggle and react to the hit.

3. The Test Drive: Does It Run Smoothly?

Before letting this new engine drive on the highway, the authors had to make sure it actually worked.

• The Check: They compared the computer's output against the original, hand-calculated math from the Lyon team.
• The Result: It was a perfect match. The new "Martini" engine in GENIE produced the exact same numbers as the original theoretical calculations.

4. The Road Test: Real-World Experiments

Next, they took the car out to see how it performed against real data from two major experiments: T2K (in Japan) and MicroBooNE (in the US).

• The T2K Test: They looked at collisions with Carbon and Oxygen nuclei. The new model predicted the results very well, matching the real-world data better than many other existing models. It correctly accounted for the "team kicks" that other models missed.
• The MicroBooNE Test: They looked at collisions with Argon (used in a different type of detector). Again, the new model fit the data incredibly well, even better than the other models currently in use.

5. The Limitations (The "Fine Print")

The paper is honest about where the new engine still has some rough edges:

• The Map is Incomplete: The new engine only works well for specific types of nuclei (Carbon, Oxygen, and Calcium/Argon). If you try to use it for heavier metals like Iron, the computer has to guess based on math tricks, which isn't perfect.
• The "Ghost" Particles: The model is great at predicting the total energy and number of particles, but it doesn't perfectly simulate the chaotic aftermath (like how the remaining nucleus shakes or how particles bounce off each other after the crash). It's like the engine predicts the crash perfectly, but the simulation of the debris field is still a bit rough.
• Missing Pieces: The model can technically handle other types of collisions (like creating pions), but for this specific paper, the authors only installed the parts for "quasielastic" and "multinucleon" hits. The rest is left for future updates.

The Bottom Line

This paper is a major upgrade to the software scientists use to study neutrinos. By installing this specific, highly accurate mathematical model into the GENIE program, they have given researchers a better tool to understand how neutrinos interact with matter. This helps reduce the "systematic errors" (the fog in the data) that currently limit our understanding of the universe.

In short: They took a complex, theoretical recipe for neutrino collisions, cooked it up inside the world's most popular neutrino simulation software, and proved that it tastes exactly like the real thing.

Technical Summary

Technical Summary: Implementation of the Martini-Ericson-Chanfray-Marteau RPA-based Model in GENIE

Problem Statement
The success of next-generation long-baseline neutrino oscillation experiments, such as Hyper-Kamiokande and DUNE, is contingent upon reducing systematic uncertainties to the percent level. Currently, the dominant source of these uncertainties is the modeling of neutrino-nucleus cross sections. The nucleus is a complex quantum many-body system, and accurately modeling its dynamics—particularly the interplay between quasielastic (QE) scattering and multinucleon (2p2h and 3p3h) excitations—remains a significant challenge. While various theoretical models exist, none currently describes all available experimental data across the full neutrino energy range and diverse nuclear targets (e.g., $^{12}$C, $^{16}$O, $^{40}$Ar) simultaneously. Furthermore, existing implementations in Monte Carlo event generators often rely on reweighting procedures using cross-section tables rather than direct implementation of the underlying hadron tensors, limiting their flexibility for exclusive measurements.

Methodology
This paper details the first implementation of the Martini-Ericson-Chanfray-Marteau (Martini et al.) model, a Random Phase Approximation (RPA)-based theoretical framework, into the GENIE neutrino event generator. The implementation focuses on charged-current (anti)neutrino interactions, specifically the quasielastic (1p1h) and multinucleon (npnh, comprising 2p2h and 3p3h) excitation channels.

The core of the methodology involves inserting new hadron tensor look-up tables directly into GENIE, rather than relying on cross-section reweighting. These tables, derived from the original theoretical calculations in Ref. [5] (including relativistic corrections from Refs. [48, 49]), cover energy transfers ($\omega$) from 5 to 995 MeV and momentum transfers ($q$) from 1 to 2000 MeV/c. The tensors are provided for $^{12}$C, $^{16}$O, and $^{40}$Ca. For targets like $^{40}$Ar, the model employs scaling procedures based on nucleon number and Fermi momentum ratios to adapt the $^{40}$Ca tensors.

The implementation utilizes a factorization scheme to generate hadron kinematics from the inclusive input model, a standard approach in GENIE for models of this type. The 3p3h contribution is treated as an additional effective 2p2h channel within the existing GENIE hadron kinematics generator, resulting in two nucleons in the final state rather than three, a limitation noted for future extension.

Key Contributions

1. Direct Hadron Tensor Implementation: The paper introduces the first direct implementation of the Martini et al. model's hadron tensor into a Monte Carlo generator. This allows for a unified description of 1p1h and npnh channels without the need for external reweighting.
2. Component Separation: The implementation preserves the model's ability to separate distinct 2p2h contributions: nucleon-nucleon (NN) correlations, meson-exchange currents (MEC), and their interference, as well as 3p3h excitations. This facilitates detailed comparisons between models and helps avoid double-counting when combining channels from different theoretical frameworks.
3. Validation: The authors performed rigorous validation by comparing GENIE outputs directly against the original theoretical calculations for $^{12}$C, $^{16}$O, and $^{40}$Ca. This included comparisons of total cross sections, flux-integrated double differential cross sections, and fixed-kinematics double differential cross sections.
4. Experimental Comparison: The model's predictions were compared against experimental data from the T2K and MicroBooNE collaborations, specifically focusing on "CC0$\pi$" (no pions) and "CC1p0$\pi$" (one proton, no pions) topologies on Carbon, Oxygen, and Argon targets.

Results

• Validation: The GENIE implementation shows perfect agreement with the original theoretical calculations for total cross sections and differential distributions across the tested kinematic ranges. The implementation successfully reproduces the distinct phase-space behaviors of the 1p1h and npnh channels, including the presence of two branches in the multinucleon response (one below and one above the quasielastic region).
• T2K Comparison: For T2K flux-integrated $\nu_\mu$ CC0$\pi$ cross sections on Carbon and Oxygen, the Martini et al. model yields a $\chi^2$ of 76.6 for 58 degrees of freedom ($p=0.05$), indicating reasonable agreement. The model performs better than several other implementations (including NEUT, NuWro, and GENIE's SuSAv2) in the forward scattering region, likely due to its explicit inclusion of RPA suppression effects.
• MicroBooNE Comparison: For MicroBooNE flux-integrated $\nu_\mu$ CC1p0$\pi$ cross sections on Argon, the model achieves a $\chi^2$ of 6.4 for 18 degrees of freedom ($p=0.994$), providing the best fit among the models considered in the referenced MicroBooNE analysis. This superior performance is attributed to the model's higher predicted 2p2h cross section compared to other models based on previous versions of the Nieves et al. approach.

Significance
The paper concludes that this implementation represents a significant step toward reducing systematic uncertainties in neutrino-nucleus cross-section modeling for future precision experiments. By providing a unified, hadron-tensor-based description of quasielastic and multinucleon excitations within GENIE, the work enables more robust comparisons between theory and the increasingly exclusive measurements being performed by current and future experiments. The authors note that while the current implementation is limited to charged-current interactions and specific nuclear targets, and relies on scaling for non-isoscalar nuclei, it provides a reliable framework for outgoing lepton kinematics and serves as a foundation for future extensions to neutral-current channels and single pion production.