🔭 astrophysics

Galactic Constellations in DESI DR1 and the Scales of Cosmological Homogeneity

This paper introduces "galactic constellations" as a novel, public-facing tool for exploring DESI DR1 data, demonstrating that the size distribution of these user-identified patterns is consistent with the cosmological principle and the Lambda-CDM model.

Original authors: Claire Lamman

Published 2026-04-01

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Claire Lamman

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 vast, empty void, but as a giant, three-dimensional ocean filled with billions of glowing islands (galaxies). For thousands of years, humans have looked up at the night sky and connected the dots between stars to create pictures of bears, hunters, and swans. These are our stellar constellations.

This paper is about doing the exact same thing, but with galaxies instead of stars, using a massive new telescope map called DESI.

Here is the story of the paper, broken down into simple, everyday concepts:

1. The New Map: A High-Resolution Google Earth for the Universe

Think of the Dark Energy Spectroscopic Instrument (DESI) as a super-powered camera on a telescope in Arizona. Instead of taking a blurry photo, it takes a "3D scan" of millions of galaxies, telling us exactly where they are and how far away they are.

The authors looked at the first batch of data from this camera (called DR1). They didn't just look at the numbers; they looked for shapes. Just like your brain instantly sees a face in a cloud, the human brain is amazing at spotting patterns in messy data.

2. The "Galactic Constellations"

The team found some funny and cool shapes in the galaxy map. They named them:

Pisces Grandis: A giant fish swimming toward us. It's huge—about 900 million light-years long. If you were inside it, you'd see it growing bigger every second as the universe expands.
The DESI Stick Woman: A stick-figure lady. She is the female cousin of the famous "CfA Stick Man," a galaxy shape discovered back in 1985.
W: A shape that looks like the letter W (or the constellation Cassiopeia).

The Analogy: Imagine you are looking at a crowd of 10,000 people from a helicopter. Most people are just a blur. But if you squint, you might see a group of people holding hands that looks like a heart, or a line of people that looks like a snake. These "galactic constellations" are those shapes, but made of entire galaxies.

3. The "Crowdsourcing" Experiment

The authors realized that if they just looked for shapes, they might miss the funny ones. So, they built a website (like a digital art gallery) where anyone could log in, look at random slices of the galaxy map, draw lines between galaxies to make their own constellations, and name them.

The Result: Regular people found 93 new constellations.
The Winner: The most popular one was named "Cow Tools." This is a reference to an old internet meme about a cartoon cow making tools that look like... well, cow tools. It's a silly name, but it proves that constellations aren't just about geometry; they are about culture and stories. We project our own ideas onto the sky.

4. The Serious Science: Is the Universe "Smooth"?

Here is the twist: While this sounds like a fun art project, the authors used it to test a very serious rule of physics called the Cosmological Principle.

The Rule: This principle says that if you zoom out far enough, the universe looks the same everywhere. It's "homogeneous" (smooth) and "isotropic" (looks the same in all directions). It's like looking at a bowl of oatmeal up close: you see individual oats and lumps. But from a mile away, it just looks like a smooth, brown mush.
The Test: The scientists asked: "At what size do these galaxy shapes stop looking like distinct 'objects' and start blending into the smooth mush?"
The Finding: Everyone's drawings were smaller than a certain size (about 400 million light-years). No one found a "Giant Galaxy Dragon" that stretched across the whole universe.
The Conclusion: This confirms that the universe does smooth out at large scales, just like the Cosmological Principle predicts. The "lumps" (galaxies and clusters) exist, but they don't form giant, coherent structures that break the rules of the universe.

5. Why This Matters (and Why AI Failed)

The paper makes a funny but important point about Artificial Intelligence. The authors tried to use a super-smart AI to find these shapes, but the AI failed miserably. It couldn't see the "Stick Woman" or the "Fish."

The Metaphor: AI is like a robot that counts pixels. Humans are like artists who see stories. Our brains are uniquely wired to find meaning in chaos. We can look at a random scattering of dots and say, "That looks like a dog!" AI just sees dots.

Summary

This paper is a celebration of human curiosity. It says:

Look at the data: The universe is full of hidden shapes.
Play with it: We built a game so anyone can find their own cosmic pictures.
Learn from it: Even though we were just "playing," our game accidentally proved that the universe is smooth and orderly on the biggest scales.

It's a reminder that science isn't just about cold equations; it's also about looking up at the dark, connecting the dots, and telling stories about the cosmos. And yes, there is a giant fish in the universe, and it's getting bigger every day.

Based on the provided manuscript, here is a detailed technical summary of the paper "Galactic Constellations in DESI DR1 and the Scales of Cosmological Homogeneity."

1. Problem Statement

The paper addresses two primary challenges in modern cosmology:

Accessibility and Intuition: While cosmologists use increasingly complex statistics to analyze the Large-Scale Structure (LSS) of the universe, these methods are often inaccessible to the general public and difficult to intuitively grasp. There is a need for engaging ways to visualize the cosmic web and help audiences develop an intuition for characteristic patterns and their scales.
Probing Homogeneity: The Cosmological Principle (CP) posits that the universe is homogeneous and isotropic on large scales. However, defining the exact transition scale (where the universe becomes homogeneous) remains non-trivial, with theoretical and observational estimates ranging from 60 to 500 $h^{-1}$ Mpc. The authors propose an unconventional method to probe these scales by analyzing the physical sizes of human-identified patterns.

2. Methodology

The study utilizes data from the Dark Energy Spectroscopic Instrument (DESI) First Data Release (DR1) and employs a hybrid approach of human visual pattern recognition and interactive web-based data exploration.

Data Selection:
- Source: DESI DR1, containing 14 million extragalactic spectra.
- Region: A high-completeness patch of sky ( $125^\circ < \alpha < 250^\circ$ , $-7^\circ < \delta < 7^\circ$ ), primarily within the Virgo constellation.
- Redshift: Limited to $z < 0.8$ , focusing on the Bright Galaxy Survey (BGS) and Luminous Red Galaxy (LRG) samples.
- Processing: Data was converted to comoving coordinates using Planck 2018 cosmological parameters. The authors intentionally retained Redshift-Space Distortions (RSD) as they are a historical and scientific feature of the maps, noting that reconstructed maps would lack accuracy for individual structures.
Visualization Strategy:
- The 3D data was sliced into 2D declination slices of 2-degree thickness to balance structural clarity with field density.
- Points were plotted with semi-transparency and brightness scaled by declination to emphasize overdensities and depth.
- Images were generated in square formats ranging from 100 to 1000 $h^{-1}$ Mpc.
Pattern Discovery Mechanism:
- Initial Search: Conducted by a "biological neural network" (the human author, Claire Lamman) trained on 29 years of data, identifying three primary constellations.
- Interactive Platform: A public website (cmlamman.github.io/galactic-constellations) was developed allowing users to:
  - View random 2-degree slices of DESI data.
  - Rotate images and draw connections between galaxies to form constellations.
  - Submit, name, and share their discoveries.
- AI Comparison: The authors tested AI models (specifically Claude's Opus 4.6) on the same data, finding them unable to identify recognizable shapes even with detailed prompting, highlighting human superiority in visual pattern recognition.

3. Key Contributions

Discovery of New Galactic Constellations: The paper introduces the first three distinct galactic constellations found in DESI data:
1. Pisces Grandis: A massive fish-like structure 270 $h^{-1}$ Mpc long (~900 million light-years).
2. The DESI Stick Woman: A counterpart to the historic "CfA Stick Man."
3. W: A shape resembling the stellar constellation Cassiopeia.
Public Engagement Tool: The creation of a functional website that democratizes access to DESI data, resulting in 93 user-submitted constellations from early testers.
Novel Cosmological Probe: The proposal to use the maximum physical span of human-identified constellations as a statistical probe for the scale of cosmological homogeneity.

4. Results

Constellation Sizes: Analysis of the 93 submitted constellations revealed that the maximum distance between points in any drawing was consistently less than 400 $h^{-1}$ Mpc.
Homogeneity Scale: This upper limit aligns with the theoretical transition scales (60–500 $h^{-1}$ Mpc) where the universe is expected to become homogeneous and isotropic. The absence of recognizable, coherent human-identified patterns above ~400 $h^{-1}$ Mpc is qualitatively consistent with the Cosmological Principle and the $\Lambda$ CDM model.
AI Limitations: The failure of AI to replicate human pattern recognition in this context underscores the unique capability of human visual processing in identifying complex, non-linear structures in noisy data.
Statistical Tension (Satire/Warning): The authors note that no "Pisces Grandis" was found in 25 mock DESI datasets from the Abacus simulation suite. While a frequentist interpretation might suggest tension with $\Lambda$ CDM at 96% confidence, the authors conclude that frequentist statistics should not be applied to individual structures in cosmological analysis, as unique, high-order patterns are not expected to be reproduced exactly in simulations.

5. Significance

Scientific Communication: The work bridges the gap between complex cosmological data and public understanding, using "fun" pattern recognition to teach concepts of the cosmic web and LSS.
Validation of the Cosmological Principle: By demonstrating that human-identified structures do not persist beyond the expected homogeneity scale, the study provides an intuitive, albeit unconventional, confirmation of the $\Lambda$ CDM model.
Future Prospects: The paper highlights that the current DESI sample covers only 10% of the final sky area and <0.01% of its volume. Future work will expand to 3D slices and right ascension, potentially revealing new structures and refining the statistical power of this "constellation" method.
Cultural Impact: It reinvigorates the ancient tradition of constellation mapping, extending it from the local stellar neighborhood to the extragalactic scale, creating a new "mythology" for the universe.

Note on Tone: The paper is written with a distinct blend of rigorous cosmological analysis and playful satire (e.g., the "biological neural network" description of the author, the "Cow Tools" meme reference, and the humorous conclusion about frequentist statistics). While the methodology and data handling are technical, the narrative style is intentionally engaging and slightly irreverent, fitting the journal Acta Prima Aprilia (a known April Fools' Day journal). However, the core scientific premise regarding the scales of homogeneity and the utility of human pattern recognition is presented seriously.