Hunting Dark Matter’s Echo at the LHC

Author: Denis Avetisyan

New analysis techniques could reveal subtle, time-varying signals at the Large Hadron Collider hinting at interactions with ultralight dark matter particles.

The distribution of particle masses over time distinguishes a sought-after signal from background noise-the latter remaining constant while the signal exhibits temporal variation and statistical fluctuations-and a further analysis of likelihood ratios confirms this separation, suggesting a viable method for isolating the desired phenomenon.

This review details how time-dependent signatures and improved statistical methods can enhance searches for new physics beyond the Standard Model at the LHC.

Current searches for new physics at the Large Hadron Collider typically assume time-invariant signals, potentially overlooking crucial information. This paper, ‘Time-dependent signals of new physics at the LHC’, investigates the sensitivity gains achievable by explicitly considering time-dependent signatures arising from interactions between ultralight dark matter and Standard Model particles. We demonstrate that incorporating timing information into resonance searches can improve sensitivity by up to a factor of two, leveraging novel statistical techniques and background modeling strategies. Could this approach unlock previously hidden signals and reshape our understanding of physics beyond the Standard Model?

The Universe’s Missing Pieces: Beyond the Standard Model

Despite its extraordinary predictive power and comprehensive description of known fundamental particles and forces, the Standard Model of particle physics encounters a significant challenge: it cannot account for the vast amount of unseen matter that constitutes approximately 85% of the universe’s total mass – dark matter. Numerous astrophysical observations, from galactic rotation curves to the cosmic microwave background, strongly suggest the existence of this non-luminous substance, yet no Standard Model particle possesses the necessary properties to explain it. This discrepancy indicates that the Standard Model is an incomplete picture of reality, compelling physicists to explore theoretical extensions and search for new particles and interactions that lie beyond its current framework. The search for dark matter, therefore, represents a crucial frontier in modern physics, promising to reveal fundamental truths about the universe’s composition and evolution.

The limitations of the Standard Model of particle physics demand a broadening of the search for fundamental constituents and forces. Current investigations aren’t simply refining existing theories, but actively proposing and testing entirely new particles and interactions that lie beyond its predictive power. These explorations encompass a wide range of possibilities, from weakly interacting massive particles (WIMPs) to axions and sterile neutrinos, each requiring novel experimental approaches for detection. Furthermore, physicists are considering scenarios involving extra spatial dimensions and supersymmetry, concepts that, while mathematically complex, could potentially resolve the Standard Model’s shortcomings and explain phenomena like dark matter and dark energy. This pursuit of “new physics” isn’t just about filling gaps in knowledge; it represents a fundamental shift in understanding the very fabric of reality, potentially rewriting the rules governing the universe at its most basic level.

An intriguing solution to the dark matter puzzle proposes a shift in perspective – considering ultralight dark matter not as a collection of particles, but as a classical field, akin to an electromagnetic field. This concept, diverging from the standard ‘particle dark matter’ model, suggests that dark matter’s effects arise from the collective behavior of waves rather than individual interactions. Instead of searching for weakly interacting massive particles (WIMPs), this framework predicts phenomena like interference patterns and the formation of “solitons” – stable, localized waves – within the dark matter distribution. These solitons could potentially explain the density profiles of dwarf galaxies and even act as seeds for the formation of larger structures in the universe, offering a radically different explanation for the observed cosmic web. The wave-like nature also allows for resonant amplification of signals, potentially making ultralight dark matter detectable through precision measurements of gravitational waves or the subtle effects it might have on light traveling through galactic halos.

Hunting for Ephemeral Signals: Time-Varying Evidence

Ultralight dark matter, with masses between $10^{-{22}}$ and $10^{-{18}}$ eV, is theorized to form coherent oscillations due to its wave-like nature. These oscillations manifest as time-dependent variations in the dark matter density, creating a detectable signal. The frequency of these variations is directly related to the ultralight dark matter particle mass, with lower masses corresponding to longer oscillation periods. These variations are expected to induce measurable effects in systems interacting with the dark matter, such as those used in direct detection experiments, presenting as time-varying forces or energy depositions. The amplitude of the signal is dependent on the local dark matter density and the coupling strength between the dark matter and the detector.

Detection of time-varying signals from ultralight dark matter necessitates the application of frequency domain analysis techniques, prominently including Fourier Transforms. These transforms decompose a time-series signal into its constituent frequencies, allowing researchers to identify periodic variations that may indicate the presence of dark matter. The resulting frequency spectrum reveals the amplitude of each frequency component; a peak at a specific frequency suggests a corresponding oscillation in the original data. The frequency resolution of the analysis, determined by the duration of the observed data stream, directly impacts the ability to resolve closely spaced frequency components. Furthermore, the choice of window function applied prior to the Fourier Transform influences spectral leakage and the accuracy of amplitude measurements, requiring careful consideration during data processing to minimize artifacts and maximize sensitivity.

CATHODE employs machine learning techniques to improve the detection of time-varying signals potentially indicative of ultralight dark matter. This is achieved by training algorithms to identify and characterize temporal features within the data, effectively discriminating between genuine signals and background noise. Specifically, the machine learning models learn the time-dependent characteristics of the expected signal, allowing for a more sensitive search than traditional methods. Simulations indicate this approach could improve limits on signal rates by up to a factor of 7 in scenarios where the signal is heavily masked by noise, representing a substantial enhancement in detection capability.

CATHODE successfully learns the temporal modulation of a signal <span class="katex-eq" data-katex-display="false"> ext{(solid blue)}</span> despite being trained without prior knowledge of its period or phase, achieving performance comparable to the ground truth <span class="katex-eq" data-katex-display="false"> ext{(dashed orange)}</span> and demonstrating robustness to limitations in the number of detectable signal events. — CATHODE successfully learns the temporal modulation of a signal $ext{(solid blue)}$ despite being trained without prior knowledge of its period or phase, achieving performance comparable to the ground truth $ext{(dashed orange)}$ and demonstrating robustness to limitations in the number of detectable signal events.

Pinpointing the Invisible: Resonance Searches and Statistical Rigor

Resonance searches identify potential new particles by analyzing Invariant Mass Distributions (IMDs) derived from interaction events. These distributions are constructed by calculating the $m_{inv} = \sqrt{(E_{i} + E_{j})^2 - (p_{i} + p_{j})^2}$ for all possible particle pairs (i, j) produced in a collision, where $E$ and $p$ represent energy and momentum, respectively. A peak in the IMD suggests the existence of a resonance – a short-lived particle decaying into the observed pair. The position of the peak indicates the mass of the new particle, and its width relates to its decay rate. Statistical analysis is then performed on the observed distribution to determine if the peak is a genuine signal or a result of random fluctuations in the background.

Missing transverse momentum, denoted as $\vec{p}_T^{miss}$ , arises from the vector sum of all detected particle momenta. A significant $\vec{p}_T^{miss}$ signature indicates the presence of undetected particles, typically those that interact weakly or are otherwise difficult to detect with the experimental apparatus. This occurs because momentum is a conserved quantity; if detected particles do not account for the total initial momentum, the imbalance is attributed to these ‘missing’ particles. The magnitude of $\vec{p}_T^{miss}$ is calculated in the plane perpendicular to the beam axis to maximize sensitivity, as longitudinal momentum is often poorly reconstructed. Consequently, searches for physics beyond the Standard Model, such as supersymmetry or extra dimensions, frequently utilize $\vec{p}_T^{miss}$ as a key signature to identify events containing these weakly interacting particles.

PyHF is a statistical framework employed in high-energy physics to establish upper limits on signal rates for hypothesized new particles, particularly in scenarios where a statistically significant excess of events above background is not observed. This framework utilizes a frequentist approach, enabling the calculation of confidence intervals on signal strengths. Importantly, PyHF allows for the incorporation of time-dependent effects into the analysis; specifically, analyses leveraging time-varying backgrounds or signal efficiencies have demonstrated an improvement in sensitivity – up to a 50% increase – compared to traditional methods that assume static rates throughout the data-taking period. This enhanced sensitivity stems from a more accurate modeling of the observed data and a reduction in systematic uncertainties related to time-dependent variations.

Weighting by the likelihood ratio (LR) effectively separates signal (red) from background (black) in mass distribution and improves the accuracy of expected limits versus <span class="katex-eq" data-katex-display="false">T_{\rm Period}</span> when utilizing time-dependent information. — Weighting by the likelihood ratio (LR) effectively separates signal (red) from background (black) in mass distribution and improves the accuracy of expected limits versus $T_{\rm Period}$ when utilizing time-dependent information.

Expanding the Dark Sector: A More Complex Universe

Current dark matter research isn’t limited to simple models; theories involving Non-Abelian Pseudo-Nambu-Goldbosons (pNGBs) propose a richer, more interactive dark sector. These pNGBs arise from spontaneously broken symmetries within extensions to the Standard Model, suggesting dark matter particles aren’t isolated but engage in complex interactions with each other. This framework expands upon the ultralight dark matter paradigm – which posits extremely low-mass dark matter particles – by introducing self-interactions mediated by these new particles. Consequently, the distribution and behavior of dark matter become far more nuanced, potentially explaining discrepancies observed in galactic structure formation and offering avenues for detection beyond simple gravitational effects. Exploring these models necessitates considering a wider range of dark matter particle properties and interaction strengths, ultimately painting a more detailed picture of the universe’s hidden mass.

The pursuit of dark matter candidates increasingly explores scenarios beyond simple, weakly interacting particles, with models incorporating spontaneously broken symmetries as a compelling avenue. A common feature of these theoretical frameworks is the introduction of new force carriers, most notably the ‘dark photon’ – a hypothetical particle analogous to the photon but mediating interactions within the dark sector. This dark photon arises from a broken U(1)’ symmetry, suggesting a hidden force distinct from electromagnetism, and potentially coupling weakly to standard model particles. Detecting these dark photons represents a significant challenge, but ongoing experiments are designed to search for subtle anomalies in electromagnetic interactions or the decay of dark matter into detectable photons, offering a potential window into the previously unseen realm of dark interactions and a more complete understanding of the universe’s missing mass.

Characterizing the elusive nature of dark matter demands a detailed understanding of its velocity distribution, a task approached through the lens of probability distributions like the Rayleigh distribution. This isn’t simply about average speeds; it’s about the likelihood of dark matter particles possessing specific velocities within a given population. Current analyses are pushing the boundaries of detection by examining potential oscillations in dark matter signals over extraordinarily long coherence times-reaching up to $10^6$ months. Such extended observation periods are crucial for discerning subtle variations and confirming the presence of these faint signals, as shorter durations might obscure genuine oscillations within the noise. The ability to resolve these long-term patterns hinges on precise modeling of the velocity distribution and the capacity to filter out extraneous influences, offering a promising pathway toward unveiling the properties of this mysterious substance.

The sensitivity to dark matter oscillation periods is significantly improved by accounting for time-dependent effects on the signal, as demonstrated by the limits on the number of signal events (<span class="katex-eq" data-katex-display="false"> N_s </span>) derived from both a model-independent analysis and a specific model explored in Ref. [15], with the constraining power primarily stemming from a single kinematic bin and being influenced by the phase offset δ for long oscillation periods. — The sensitivity to dark matter oscillation periods is significantly improved by accounting for time-dependent effects on the signal, as demonstrated by the limits on the number of signal events ( $N_s$ ) derived from both a model-independent analysis and a specific model explored in Ref. [15], with the constraining power primarily stemming from a single kinematic bin and being influenced by the phase offset δ for long oscillation periods.

The Road Ahead: Synergies and Future Discoveries

The Large Hadron Collider (LHC) remains central to probing the fundamental constituents of the universe and validating theoretical models beyond the Standard Model. Through high-energy particle collisions, the LHC doesn’t just confirm known physics; it actively searches for deviations hinting at new particles and interactions. These investigations encompass a broad range, from supersymmetry-predicting partner particles for each known particle-to extra dimensions and the potential production of dark matter candidates directly. By precisely measuring the properties of known particles and seeking evidence of the unexpected, the LHC provides critical constraints on theoretical frameworks and guides the direction of future research in particle physics, effectively mapping the landscape of the unknown with each collision.

A comprehensive understanding of dark matter necessitates a multi-pronged approach, skillfully integrating data from seemingly disparate sources. Particle accelerators, like the Large Hadron Collider, probe the high-energy frontier, seeking to create and identify potential dark matter particles through collision events. Complementing this, direct detection experiments aim to observe the faint interactions between dark matter particles and ordinary matter within highly shielded detectors. Simultaneously, astrophysical observations-mapping galactic rotation curves, gravitational lensing effects, and the cosmic microwave background-provide crucial insights into the distribution and abundance of dark matter on a cosmic scale. By cross-correlating these datasets-collider results, direct detection signals, and astronomical surveys-scientists can build a more robust and complete picture of dark matter’s properties, verifying theoretical predictions and potentially revealing unexpected phenomena within the universe’s hidden sector.

The prevailing understanding of dark matter centers on Weakly Interacting Massive Particles (WIMPs), but a growing body of theoretical work suggests exploring alternative candidates is vital for a complete picture of the universe’s hidden sector. One promising avenue involves the Dilaton, a hypothetical particle arising from string theory that interacts with standard model particles through a unique mechanism tied to the Higgs boson. Investigating the Dilaton-and other beyond-WIMP candidates like axions or sterile neutrinos-requires innovative experimental approaches and refined theoretical models. These alternative dark matter forms often predict distinct signatures, potentially detectable through subtle variations in gravitational effects, or through precise measurements of particle interactions. By broadening the search beyond conventional WIMPs, scientists aim to unlock the true nature of dark matter and gain deeper insights into the fundamental laws governing the cosmos, potentially revealing connections between dark matter, dark energy, and the very structure of spacetime.

The pursuit of new physics, as detailed in this analysis of LHC data, necessitates a rigorous embrace of uncertainty. The paper’s focus on time-dependent signals and refined statistical methods underscores the inherent challenges in distinguishing subtle anomalies from background noise. It is a testament to the scientific method’s iterative nature, where repeated refinement – and potential disproof – strengthens conclusions. As Albert Camus observed, “The only way to deal with an unfree world is to become so absolutely free that your very existence is an act of rebellion.” This sentiment mirrors the study’s approach; a relentless questioning of established models, and a dedication to uncovering truth even amidst the complexities of resonance searches and background modeling.

Where Do We Go From Here?

The pursuit of ultralight dark matter, as subtly probed through LHC phenomenology, highlights a persistent tension. Statistical significance, even with refined background modeling, remains a slippery concept. The paper rightly emphasizes the need for novel techniques, but one suspects that if a single statistical approach were to unlock all mysteries, it would be a triumph of marketing rather than analysis. The true challenge isn’t merely finding a signal, but convincingly excluding every conventional explanation – a task inherently asymptotic.

Future work must grapple with the limitations of resonance searches. Assuming dark matter interacts with Standard Model particles, and further, that these interactions manifest as observable resonances, is a powerful, yet restrictive, assumption. Exploring genuinely time-dependent effects – variations in signal strength, phase shifts, or altered decay profiles – may require moving beyond traditional search strategies and embracing more complex event reconstruction techniques.

Predictive power, even if statistically compelling, is not causality. The LHC, as a tool, reveals correlations, not necessarily fundamental truths. The next step requires a broader theoretical framework – one that connects these potential dark matter interactions to other outstanding puzzles in particle physics. Until then, these tantalizing signals remain just that: suggestive hints demanding further, rigorous, and skeptical investigation.

Original article: https://arxiv.org/pdf/2605.11071.pdf

Contact the author: https://www.linkedin.com/in/avetisyan/