Anomaly in the Cosmic Microwave Background

by George Cristache, Ph.D. student

24.10.2025

Credit: Sullivan, R. M. & Scott, D. The CMB Dipole: Eppur Si Muove, arXiv:2111.12186(2021)

The Cosmic Microwave Background (CMB) is the “fossil light” of the Big Bang, arriving from all directions in the sky. It is extremely uniform, but it shows a clear variation of the dipole type: in one direction it appears slightly warmer, and in the opposite direction, slightly cooler.

This effect arises because we are moving at a speed of about 370 km/s relative to the CMB. Due to the Doppler effect, the radiation seems warmer in the direction of our motion and cooler in the opposite direction. Our motion through the Universe also produces secondary effects in measurements associated with the CMB such as the modulation of anisotropies, optical aberration, and changes in the distribution of galaxies. Researchers have shown that, in principle, these secondary effects can be measured. However, they point out that none of these measurements can guarantee with absolute certainty that the observed dipole is caused only by our motion. In theory, if the Universe itself were “tilted” in one direction (meaning it had an intrinsic dipole), the observed signal would look almost identical. From this, the researchers concluded that while the current data are consistent with the kinematic explanation, they cannot completely rule out a fundamental component suggesting that the Universe may be more complex than we think.

Complementary to these CMB studies, another team of researchers proposed examining the distribution of radio sources in the Universe. Under the cosmological principle, the Universe should look the same in all directions (this is called isotropy). Yet, because we move relative to the rest of the cosmos, we observe a small asymmetry: in one direction we see slightly more sources (because we are moving toward them), and in the opposite direction, fewer. The researchers combined three major radio surveys(NVSS, RACS-low, and LoTSS-DR2) and developed a more precise statistical method to estimate this asymmetry, accounting for the fact that radio sources are not completely random in their distribution (some tend to cluster, increasing the variance). Their result shows that the observed dipole is about 3.7 times stronger than the one expected from our motion through the Universe. This discrepancy is large enough to be considered a statistically significant anomaly.

If this effect is real, it could point to a new cosmological component of the dipole meaning that the Universe might possess a slight intrinsic directional preference. However, researchers remain cautious, emphasizing the need for careful verification of selection effects and possible systematic errors in radio instruments.

References:

Böhme, L. & Schwarz, D. J. Overdispersed Radio Source Counts and Excess Radio Dipole Detection, arXiv:2509.16732 (2025)

Sullivan, R. M. & Scott, D. The CMB Dipole: Eppur Si Muove, arXiv:2111.12186(2021)

New Tool Brings the Universe Into Reach Faster Than Ever

by Ana Caramete, Ph.D.

17.10.2025

Credit: "Effort.jl: a fast and differentiable emulator for the Effective Field Theory of the Large Scale Structure of the Universe" by Marco Bonici, Guido D'Amico, Julien Bel and Carmelita Carbone (2025, JCAP)

October 2025 — A team of cosmologists and computational scientists has unveiled Effort.jl, a breakthrough open-source software package that promises to transform how researchers model the large-scale structure of the Universe. Effort.jl uses modern machine-learning techniques to predict how galaxies and dark matter are distributed across the cosmos. Developed in the Julia programming language, it makes complex cosmological calculations millions of times faster than traditional simulations, while remaining precise and fully open-source.

The Universe isn’t random — galaxies are arranged in a vast, interconnected “cosmic web” of filaments and voids. To understand how this structure formed, scientists run enormous computer simulations that follow the movement of matter and energy over billions of years.

Projects like IllustrisTNG, Millennium, and Abacus have revealed breathtaking detail in how galaxies form, but they require supercomputers running for weeks or months at a time. For cosmologists trying to test different models of dark matter, dark energy, or gravity, that’s a major bottleneck.

To bridge this gap, scientists have turned to emulators: machine-learning models trained on suites of simulations that can predict observables, such as galaxy power spectra, in milliseconds rather than hours or days. Emulators make it possible to carry out high-precision inference and forecasting at the scale of modern datasets.

That’s where Effort.jl comes in. Instead of rerunning full simulations each time, Effort.jl acts as a “smart shortcut” a lightning-fast emulator that learns from detailed simulations and can instantly predict what the Universe should look like under different physical conditions.

Because it’s built in Julia, Effort.jl is not only fast but also differentiable — meaning it can automatically track how small changes in cosmic parameters affect the final predictions. That makes it ideal for exploring how sensitive our Universe is to quantities like the amount of dark energy or the strength of gravity.

The tool’s open architecture allows researchers worldwide to adapt it for new surveys, extend it to more complex models, or combine it with other simulation codes. Future work will expand Effort.jl’s reach, incorporating more detailed physics, testing against real survey data, and connecting with other emulators for cosmic microwave background or gravitational lensing studies making a step forward toward a “digital twin” of the Universe

As cosmology enters the precision era, tools like Effort.jl will be essential for extracting the maximum information from vast astronomical datasets — turning terabytes of survey data into sharper insights about dark matter, dark energy, and the fundamental physics of the Universe.

Reference:

"Effort.jl: a fast and differentiable emulator for the Effective Field Theory of the Large Scale Structure of the Universe" by Marco Bonici, Guido D'Amico, Julien Bel and Carmelita Carbone (2025, JCAP).

Space mission proposal selected to next stage in ESA’s F3 mission call

09.10.2025

by Laurentiu Caramete, PhD

Photo credits: GUEST Consortium

The GUEST proposal (Gravitational Universe Exploration with Satellite Tracking), led by Prof. Diego Blas from IFAE Spain, and with contribution from nine countries, has been selected to advance to the next phase in the European Space Agency’s 2025 Call for Missions (Fast / F-class category).

The project aims to design, build, and launch two or more satellites to detect gravitational waves through laser orbital monitoring, opening a new window to explore the gravitational universe. GUEST brings together a strong international consortium in which the Institute of Space Science - INFLPR subsidiary plays an substantial role. The Romanian contribution is lead by Dr. Laurentiu-Ioan Caramete, head of the Astrophysics,  Space  Exploration and Quantum Computing Group – ASPACE-Q, and, for the moment, consist of providing astrophysical catalogs of potential sources emitting gravitational waves and data analysis solutions for the detection and characterization of gravitational waves. Another future objective in the context of this mission is to design, build and operate a system for laser ranging measurements from the ground.

In the next phase of the ESA call (2026, https://www.cosmos.esa.int/web/call-for-missions-2025/), GUEST will compete with other mission concepts across different fields of space science to determine which projects will move forward to implementation.

Sign of a black hole falling into a star

by Andrei Militaru

03.10.2025

Photo source: https://scitechdaily.com/hawking-stars-what-happens-if-you-put-a-black-hole-into-the-sun/

The most luminous electromagnetic events we had observed in the universe are gamma-ray bursts. On july 2, 2025, The Fermi Gamma-ray Burst Monitor and many others X-ray and Gamma-ray monitors have identified the longest gamma-ray burst that has ever been seen. 

The common source for such events are collapsing stars, but the duration of this burst is much longer than what this explanation allows. After going through other possible scenario, including neutron star mergers, supermassive black holes merger, and a white dwarf tidal disruption by a black hole, Eliza Neights and Eric Burns conclude that the likely cause of this burst is the fall of a black hole inside a massive helium star.

When the back hole reaches the core of the star, the core is turned into an accretion disk, and the resulting magnetic field causes the star to explode and create a supernova.

Source: https://arxiv.org/abs/2509.22792