TL;DR
This paper proposes a powerful new method to identify dark matter (DM) particles by combining data from the Large Hadron Collider (LHC) and future direct detection experiments. By making a simple but physically motivated assumption about how the local density of DM relates to its overall abundance, the authors show that we can significantly improve our understanding of DM properties, breaking degeneracies that neither experiment can resolve alone.
What Problem Does This Paper Address?
The nature of dark matter remains one of the most significant mysteries in physics. While theories like Supersymmetry (SUSY) offer compelling candidates, such as the lightest supersymmetric particle (LSP) often referred to as a WIMP (Weakly Interacting Massive Particle), definitively identifying it and proving it accounts for the observed cosmic abundance is a major challenge.
The LHC is poised to discover new particles, potentially including WIMPs. However, even if WIMPs are found, determining their precise properties (like mass and composition) and confirming they are the dominant DM component can be difficult. Accelerator data alone might lead to multiple possible solutions for the DM's relic abundance or composition.
Direct detection experiments aim to observe WIMPs scattering off atomic nuclei in underground detectors. While experiments like DAMA/LIBRA have reported potential signals, these have not been universally confirmed, and interpreting them requires careful analysis. Future direct detection experiments are expected to become much more sensitive.
The core problem this paper addresses is how to combine the complementary information from these two powerful, yet individually limited, approaches (LHC and direct detection) to achieve a robust and unambiguous identification of dark matter.
Methods
The authors employ a Bayesian statistical framework to combine data from different experimental probes. They utilize a 24-parameter supersymmetric (SUSY) model as a concrete example, focusing on a scenario where the neutralino LSP is in the "stau co-annihilation region." This region is particularly interesting because the neutralino's mass is very close to that of the lightest stau, leading to co-annihilation processes that can naturally produce the observed DM relic abundance.
The key methodological innovation lies in the proposed "scaling Ansatz" for the local DM density ($\rho_\chi$). Instead of assuming a fixed local density (which can be difficult to determine independently), they assume that the local density scales proportionally with the cosmological relic abundance:
$\rho_\chi / \rho_{DM} = \Omega_\chi / \Omega_{DM}$
This Ansatz is justified if DM particles trace the large-scale structure of the universe, which is expected for cold, collisionless DM. This assumption is crucial for breaking degeneracies.
They simulate future LHC data by assuming a set of precise measurements for various sparticle masses and mass splittings, representative of what a future LHC run might provide. For direct detection, they simulate data from a future ton-scale experiment (akin to SuperCDMS) with a specific exposure. The combination of LHC and direct detection data is achieved by updating the prior probability distribution of the SUSY parameters using Bayes' theorem, incorporating the likelihood functions from both types of experiments.
Datasets
The paper uses simulated datasets representing:
- LHC Data: A set of precise measurements for sparticle masses and mass splittings within a 24-parameter SUSY model, specifically for a benchmark point in the stau co-annihilation region. This is based on previous studies and assumes an integrated luminosity of 300 fb$^{-1}$.
- Direct Detection Data: Simulated events from a future ton-scale Germanium detector (e.g., SuperCDMS) with an exposure of 300 ton-days. This includes both simulated WIMP signal events and background events, with a conservative assumption of a constant background recoil spectrum.
The paper does not introduce new experimental datasets but rather simulates realistic future data based on established models and experimental designs.
Key Insights and Ideas
- The Power of Combined Analysis: The central idea is that combining LHC and direct detection data is significantly more powerful than either alone for identifying DM. LHC provides information about the underlying particle physics model (e.g., SUSY parameters), while direct detection probes the DM particle's interaction strength and local density.
- The "Scaling Ansatz" is Game-Changing: The simple assumption that the local DM density scales with its cosmological abundance ($\Omega_\chi / \Omega_{DM}$) is remarkably effective. It allows the analysis to cut through parameter space in a way that directly links the observed DM signal to its fundamental properties and cosmological origin.
- Breaking Degeneracies: Without the scaling Ansatz, LHC data alone can lead to multiple, degenerate solutions for the neutralino's composition and relic density. Direct detection data, even with a fixed local density, helps, but the scaling Ansatz truly resolves these ambiguities.
- Probing the DM's Identity: The combined analysis allows for a much better reconstruction of the neutralino's composition (e.g., whether it's primarily a Bino or Wino state) and its relic density, moving beyond just confirming its existence.
- Consistency Check for Cosmological History: If the combined analysis yields a relic density consistent with cosmological observations (like WMAP), it also validates the standard cosmological history of DM freeze-out. A mismatch could hint at multi-component DM or non-standard cosmological scenarios.
Key Results
- Improved Relic Density Reconstruction: Figure 1 shows that LHC data alone results in a broad probability density function (PDF) for the neutralino relic density with multiple peaks. Adding direct detection data with a fixed local density slightly narrows this, but the scaling Ansatz dramatically sharpens the PDF, yielding a much more precise determination.
- Breaking Compositional Degeneracies: Figure 2 illustrates the improvement in the $\sigma_\chi^{SI}$ vs. $\Omega_\chi h^2$ plane. LHC data alone shows large, degenerate regions. The scaling Ansatz effectively rules out regions corresponding to a low local density, which in turn disfavors certain neutralino compositions (like mostly Wino) that would require a larger scattering cross-section to produce the observed signal.
- Quantitative Improvement: The paper demonstrates that the scaling Ansatz allows for a significantly better reconstruction of the neutralino's composition and relic density compared to using LHC data alone or direct detection data with a fixed local density assumption.
Limitations and Future Work
- Model Dependence: The analysis is performed within a specific 24-parameter SUSY model and a particular benchmark point (stau co-annihilation region). While the authors argue the method is general, its application to other models or regions might require adjustments.
- Assumptions on Direct Detection Data: The simulation assumes a specific detector type (Germanium), exposure, and a conservative background model. Real-world experimental uncertainties and systematic errors could affect the precision.
- Future Work: The authors suggest that a more refined analysis should account for uncertainties in the DM velocity distribution and potential halo triaxiality, which could slightly widen the posterior distributions. They also plan to explore the constraints from DM annihilation searches in a forthcoming work. The general applicability of this method to other DM candidates beyond WIMPs is also an open question.
Summary updated