main concepts scientists sections
SimplePopularAdvanced

How to Spot the Invisible Backbone of Galaxies ✛ ✎

Original: "Polar coordinate transformations for machine learning based dark matter subhalo detection in strong gravitational lenses"
· Jacob M Campbell, Simon Dye, Emma Chapman, Adam Moss
arXiv:2607.02663v1 · 2026-07-02 · arXiv non-exclusive · independent analysis · 1 min
Galaxies Cosmology
A simple twist on space images improves the hunt for dark matter clumps.

Modern science describes the Universe with the standard model, where it all started with the Big Bang and continues to expand thanks to dark energy. But the most mysterious part of this story is dark matter – an invisible scaffold that keeps galaxies from flying apart. You can't touch it, but its gravity bends light like a lens – that's gravitational lensing. Sometimes, this distortion turns a distant galaxy into a bright ring, with tiny clumps of dark matter hidden inside.

Back in the 1930s, Fritz Zwicky noticed galaxies moving too fast and suggested invisible mass. Later, Vera Rubin proved it by studying star rotation.

Researchers trained a program to find these clumps in images from the Hubble telescope. Usually, a computer sees an image like a sheet of graph paper. But a lens ring is round. So the scientists simply cut it and straightened it into a line, like a carousel ribbon. On a straight path, any bumps are immediately noticeable. This twist of the data let the program spot even the tiniest clumps more often.

The Hubble telescope, named after Edwin Hubble, has been operating in orbit for over 30 years and has captured numerous images of gravitational lenses.

Confidently detecting these tiny dark matter clumps will help us understand what it's made of – like guessing a soup recipe from its crumbs. The number of clumps depends on the properties of its particles. This new method will speed up solving the biggest cosmic mystery.

Whole graph →

By some estimates, billions of dark matter particles zip through an ordinary room every second — you just don't notice them.

In Liu Cixin's sci-fi novel 'The Dark Forest,' gravitational lenses are used as colossal antennas for interstellar communication — almost like in real cosmology.

\rho(r) = \frac{M_0}{4\pi r (r + r_s)^2} \left( \frac{r_t^2}{r^2 + r_t^2} \right)
\rho — dark matter density at radius r, r_s — scale radius, r_t — truncation radius, M_0 — mass scale.
L = \frac{1}{N} \sum_{i=1}^{N} \left[ \frac{(\hat{y}_i - \mu_i)^2}{2\sigma_i^2} + \log \sigma_i \right]
L — loss value for N images, \hat{y}_i — true logarithmic mass, \mu_i and \sigma_i^2 — predicted mean and variance.
Scientists
Fritz ZwickyVera RubinJacob BekensteinKarl SchwarzschildJoseph WeberEmmy Noether
Concepts
neural network polarimetry Convolutional Neural Networks weak lensing gravitational lensing analysis generalization dark matter subhalo Large Language Models Machine Learning deep learning in medical imaging dark matter gravitational waves microlensing Lorentz transformations ai astrophysics training data polarimetric remote sensing image classification computer vision molecular property prediction polarimetry Natural Language Processing