🔭 astrophysics

Weighted Webs: Morphology-Informed Marked Fields

This paper demonstrates that while density-dependent marks outperform morphology-based marks individually in constraining cosmological parameters, combining both types of marks yields superior constraints by probing complementary aspects of the cosmic web's structure.

Original authors: Mikel Martin Barandiaran, Jessica A. Cowell, David Alonso, Javier Carrón Duque, Juan García-Bellido

Published 2026-07-29

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Mikel Martin Barandiaran, Jessica A. Cowell, David Alonso, Javier Carrón Duque, Juan García-Bellido

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the universe not as a smooth, empty void, but as a giant, three-dimensional spiderweb made of invisible matter. This is the "cosmic web," a vast network where galaxies are strung along thick, glowing filaments, clustered together in dense knots, and separated by enormous, empty bubbles called voids. For decades, scientists have tried to understand how this web formed and what it tells us about the rules of the universe. To do this, they usually look at the "density" of the web—basically, counting how much stuff is in a specific spot. It's like trying to understand a forest just by counting the number of trees in a square meter. While helpful, this method misses the shape of the forest: is it a dense jungle, a sparse grove, or a winding line of trees?

Recently, physicists have developed a smarter way to look at this web called "marked power spectra." Instead of just counting trees, they assign a special "weight" or "mark" to every spot in the universe based on its local environment. Think of it like giving a high score to a tree if it's standing in a dense crowd, or a low score if it's alone in a clearing. By studying how these weighted spots connect to each other, scientists can unlock hidden secrets about the universe's history, including how much dark matter exists and how heavy neutrinos (tiny, ghostly particles) are. But what if the shape of the neighborhood matters just as much as the crowd size? What if the fact that a tree is part of a long, winding line tells us something a simple count cannot?

This is exactly what the paper "Weighted Webs: Morphology-Informed Marked Fields" investigates. The authors, a team of cosmologists, asked a simple question: Can we get better answers by weighting the cosmic web based on its shape (morphology) rather than just its density? They tested a variety of new "marks" that act like filters, highlighting specific geometric features of the universe, such as whether a region is a flat sheet, a long filament, or a round cluster. They ran thousands of computer simulations to see if these shape-based weights could reveal more about the universe's fundamental ingredients than the traditional density-based weights.

The results were a mix of "not quite" and "yes, but together." The team found that the old-school method—weighting based purely on how dense a region is—still remains the champion on its own. It provides the strongest clues about the universe's composition. The new shape-based marks, which tried to highlight filaments or flat sheets, didn't beat the density method when used alone; they offered only modest improvements. However, the real magic happened when they combined them. When the scientists used both the density weights and the shape weights together, the results were significantly better than either method could achieve alone.

It's as if the density method is a great map of where the cities are, while the shape method is a great map of how the roads connect them. Using just one gives you a good picture, but using both gives you a complete navigation system. The study suggests that the universe's shape holds extra information that density alone misses, specifically regarding the "tidal forces" that stretch and squeeze matter into filaments and walls. While the specific mathematical formulas used to create these shape marks didn't matter much (the shape itself was the key), the discovery confirms that looking at the geometry of the cosmic web is a powerful new tool. The authors conclude that while we haven't found a single "magic" shape filter that beats the density filter, combining geometry with density is the winning strategy for decoding the universe's deepest secrets.

Technical Summary: Weighted Webs: Morphology-Informed Marked Fields

Problem Statement

The large-scale structure (LSS) of the Universe contains cosmological information encoded in the non-Gaussian features of the late-time matter density field, which standard two-point statistics (the power spectrum) fail to fully capture. While higher-order statistics (HOSs) such as polyspectra and Minkowski functionals exist, they often face challenges regarding theoretical predictability and computational efficiency. Marked power spectra offer a computationally efficient alternative by applying spatially varying weights (marks) to the density field, thereby encoding higher-order information into two-point statistics.

However, existing applications of marked power spectra predominantly rely on mark functions dependent solely on the local smoothed density ( $\delta_R$ ). This approach overlooks the fact that the gravitational evolution of the cosmic web is governed by the full tidal tensor (the Hessian of the gravitational potential), not just its trace (the density). Consequently, the morphology of the cosmic web—characterized by anisotropic structures like filaments and walls versus isotropic clusters and voids—may contain additional cosmological information that density-dependent marks fail to access. This work addresses the question of whether marks sensitive to the local morphology (specifically tidal shear and fractal dimension) can extract cosmological information complementary to standard density-based marks.

Methodology

The authors utilize the Quijote simulation suite, comprising over 43,000 full $N$ -body simulations, to generate density fields for a wide range of cosmological parameters. The analysis focuses on the cold-dark-matter-plus-baryons density field evolved to redshift $z=0$ in a periodic box of $1 \, h^{-1}\text{Gpc}$ .

Theoretical Framework

The study constructs marked fields $\tau(x)$ by weighting the density field $(1+\delta)$ with a mark function $m(x)$ :
$\tau(x) = m(x)(1+\delta(x)) - \langle m(x)(1+\delta(x)) \rangle$
Unlike previous works where $m$ depends only on $\delta_R$ , this paper explores marks dependent on:

Tidal Invariants: Specifically the shear amplitude $s = \sqrt{s_2}$ , where $s_2$ is the second invariant of the traceless tidal tensor (shear tensor). The authors also consider the cubic invariant $s_3$ (prolateness) but find it less effective.
Local Fractal Dimension (LFD): A geometric descriptor ( $D_R$ ) quantifying how enclosed mass scales with radius, serving as a continuous alternative to discrete tidal-tensor classifications (voids, walls, filaments, clusters).

Mark Families

The authors define several families of morphology-inspired marks:

Shear Amplitude: $s_R(x)$ , the raw shear field.
Isotropy Mark: Upweights regions with low shear (isotropic environments), defined as $[1 + b s_R^2]^{-p}$ .
Filament Mark: Targets anisotropic, intermediate-density regions using a shear term modulated by Gaussian density suppression: $s_R^2 \exp(-\delta_R^2 / 2\delta_*^2)$ .
Gaussian Shear Mark: Selects a specific range of shear values via a Gaussian window.
Gaussian LFD (GLFD) Mark: Uses the local fractal dimension derived from a Gaussian smoothing kernel.

Statistical Analysis

The constraining power of these marks is quantified using the Fisher matrix formalism. The data vector includes the auto-power spectrum of the unmarked field ( $P_{\delta\delta}$ ), the cross-power spectrum ( $P_{\delta\tau}$ ), and the auto-power spectrum of the marked field ( $P_{\tau\tau}$ ).

Parameters: The analysis targets $\Omega_m$ , $\sigma_8$ , and the neutrino mass sum $M_\nu$ .
Metric: Performance is measured by the Error Improvement Factor (EIF), defined as the ratio of the uncertainty on a parameter using only the standard power spectrum to the uncertainty using the marked statistics.
Covariance: Covariance matrices are estimated from 2,100 fiducial realizations using the unbiased sample covariance estimator with a Hartlap correction, rather than relying on Gaussian approximations which underestimate mode coupling at small scales.

Key Contributions

Systematic Exploration of Morphology: The paper provides the first systematic assessment of marked power spectra where the mark depends on the local tidal shear and fractal dimension, rather than just density.
Perturbative Interpretation: The authors outline the perturbative structure of shear-based marks in Fourier space, demonstrating that they access bispectrum-like (via cross-spectra) and trispectrum-like (via auto-spectra) information through non-linear mode coupling.
Complementarity Demonstration: The work establishes that morphology-based marks and density-based marks probe complementary aspects of the matter distribution.
Robustness of Shear Information: The study finds that the specific functional form of the shear-based mark (e.g., targeting filaments vs. isotropy) has a minor impact on constraints, suggesting the information is primarily encoded in the shear field itself rather than the specific non-linear transformation.

Results

Shear vs. Density: At the fiducial smoothing scale ( $R = 10 \, h^{-1}\text{Mpc}$ ), the shear amplitude mark ( $s_R$ ) alone yields modest improvements (EIF $\approx 1.27$ ) compared to the standard smoothed density mark (EIF $\approx 1.35-1.43$ ). However, at smaller scales ( $R = 5 \, h^{-1}\text{Mpc}$ ), the shear mark outperforms the density mark, indicating that tidal anisotropy retains more non-linear information on small scales.
Synergy: The most significant result is that combining density-based and morphology-based marks consistently yields the highest constraints. For instance, combining the standard White mark (density-based) with the shear amplitude mark improves the EIF for $\Omega_m$ from $\approx 2.24$ to $\approx 2.46$ (a $\sim 10\%$ improvement). This confirms that shear fields contain information not captured by density weighting.
Targeted Marks: Elaborate marks designed to specifically isolate filaments or isotropic regions do not significantly outperform the simple shear amplitude mark. The detailed functional form of the mark appears secondary to the underlying shear field.
Fractal Dimension: The GLFD mark provides a clear improvement over the standard power spectrum (EIF $\approx 1.5-1.6$ ) and performs better than using multiple smoothed density fields as separate marks. However, it does not surpass the empirically optimized White mark on its own. When combined with the White mark, it yields a modest but significant improvement, though less than the shear combination, likely due to the similar qualitative dependence of both marks on underdense regions.

Significance and Claims

The authors claim that this work establishes a physically motivated framework for exploring optimal marks beyond simple density weighting. The primary significance lies in demonstrating that:

Morphology matters: The geometric properties of the cosmic web (anisotropy, dimensionality) encode cosmological information distinct from local density.
Complementarity: While density-dependent marks remain the most powerful individual observables, morphology-informed marks provide complementary information that enhances constraints when combined.
Efficiency: The shear field itself is a robust carrier of this information, suggesting that complex, targeted mark functions may not be necessary to access the bulk of the morphological signal.

The paper concludes that while an optimized density-dependent mark will likely remain the strongest individual probe, the inclusion of morphology-aware statistics opens a new avenue for extracting additional cosmological information from the non-linear large-scale structure, particularly for parameters like the neutrino mass and matter abundance. The work serves as a foundation for future investigations into perturbative connections between these marks and higher-order correlation functions, as well as their application to redshift-space galaxy catalogs with realistic systematics.