[Paper Review] Studying dark matter with MadDM 3.1: a short user guide

Chiara Arina, Jan Heisig+5 more|arXiv (Cornell University)|Dec 16, 2020

Dark Matter and Cosmic Phenomena27 references4 citations

TL;DR

MadDM 3.1 is an automated numerical tool integrated with MadGraph5_aMC@NLO that computes dark matter observables—such as relic density, direct detection recoil rates, and indirect detection spectra—for generic new physics models. It enables full differential cross-section calculations for 2→n annihilation processes using Pythia 8 or PPPC4DMID, supporting precise phenomenological studies of dark matter candidates across collider, direct, and indirect detection experiments.

ABSTRACT

MadDM is an automated numerical tool for the computation of dark-matter observables for generic new physics models. We announce version 3.1 and summarize its features. Notably, the code goes beyond the mere cross-section computation for direct and indirect detection. For instance, it allows the user to compute the fully differential nuclear recoil rates as well as the energy spectra of photons, neutrinos and charged cosmic rays for arbitrary $2 o n$ annihilation processes. This short user guide equips researchers with all the relevant information required to readily perform comprehensive phenomenological studies of particle dark-matter models.

Motivation & Objective

• To provide a comprehensive, automated tool for computing dark matter observables across direct, indirect, and relic density searches.
• To extend beyond cross-section calculations by enabling fully differential nuclear recoil rates and energy spectra for photons, neutrinos, and cosmic rays.
• To support arbitrary 2→n annihilation processes in generic new physics models via event-level simulation or precomputed libraries.
• To integrate updated experimental constraints from LUX, Xenon1T, and Fermi-LAT for robust phenomenological interpretation.
• To streamline the workflow for researchers through a plug-in architecture within the MG5_aMC@NLO framework, enabling model-agnostic studies.

Proposed method

• Integrates with MadGraph5_aMC@NLO to generate matrix elements for any model in the Universal FeynRules Output (UFO) format.
• Uses Pythia 8 for parton showering and hadronization in precise mode, enabling full simulation of 2→n annihilation processes.
• Leverages PPPC4DMID for fast, precomputed annihilation spectra in the 'fast' mode, reducing computational cost.
• Computes relic density via thermal freeze-out using the standard Boltzmann equation, with rescaling via the annihilation cross-section factor ξ².
• Applies Fermi-LAT likelihood analysis on gamma-ray fluxes from dwarf spheroidal galaxies to compute exclusion limits and p-values.
• Generates output files with parsed observables, including fluxes, cross-sections, and likelihoods, for both single-point and scan runs.

Experimental results

Research questions

• RQ1 How can the full differential nuclear recoil rate be computed for direct detection experiments across multiple target materials?
• RQ2 What are the energy spectra of prompt photons, neutrinos, and cosmic rays from 2→n dark matter annihilation processes?
• RQ3 How do experimental constraints from LUX, Xenon1T, and Fermi-LAT impact the viability of dark matter models?
• RQ4 What is the impact of using Pythia 8 versus PPPC4DMID on the accuracy and efficiency of indirect detection spectrum computation?
• RQ5 How can the relic density be computed and compared to the observed value in a model-independent way using MadDM 3.1?

Key findings

• MadDM 3.1 computes the fully differential nuclear recoil rate as a function of energy, scattering angle, and time, enabling detailed direct detection simulations.
• The code computes velocity-averaged annihilation cross-sections ⟨σv⟩₀ in cm³s⁻¹ for all annihilation channels, with values ranging from 1.67×10⁻³⁸ to 1.76×10⁻³⁰ cm³s⁻¹ depending on the channel.
• For the same parameter point, the prompt photon flux at Earth is calculated as 1.57×10⁻¹⁴ photons/(cm² sr), with neutrino fluxes at 8.89×10⁻¹⁸ to 9.71×10⁻¹⁸ photons/(cm² sr).
• Fermi-LAT constraints yield exclusion limits from 1.11×10⁻²⁵ to 2.85×10⁻²⁵ cm³s⁻¹, with the total cross-section limit computed via full likelihood analysis.
• The program correctly identifies and skips processes with zero cross-section, such as xrxr and y0y0, ensuring computational efficiency.
• Output files like MadDM_results.txt and scan_run_01.txt contain structured, machine-readable data including likelihoods, p-values, and spectral fluxes for downstream analysis.