🔭 astrophysics

A Census of Variable Radio Sources at $3\,$ GHz

Using data from the first two epochs of the Very Large Array Sky Survey (VLASS), this study identifies approximately 3,600 compact variable radio sources, revealing that while blazars dominate the population and produce the largest absolute flux changes, galactic sources exhibit the most extreme fractional brightness variations.

Original authors: Yjan A. Gordon, Peter S. Ferguson, Michael N. Martinez, Eric J. Hooper

Published 2026-05-20

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Yjan A. Gordon, Peter S. Ferguson, Michael N. Martinez, Eric J. Hooper

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 night sky as a giant, bustling city. Most of the lights in this city are steady streetlamps—they shine with the same brightness night after night. But every now and then, a light flickers, dims, or suddenly flares up like a neon sign going haywire. In the world of astronomy, these "flickering lights" are called variable radio sources.

This paper is essentially a census (a headcount) of these flickering lights, conducted by astronomers using a massive radio telescope array called the Very Large Array (VLA). They looked at the sky at a specific radio frequency (3 GHz) over a period of about 2.5 years to see which lights changed their brightness.

Here is a breakdown of what they found, using simple analogies:

1. The Great Radio Sky Count

The researchers looked at data from two different "snapshots" of the sky (called Epoch 1 and Epoch 2), taken roughly 32 months apart.

The Result: They found about 3,600 compact radio sources that changed their brightness significantly between these two snapshots.
The "Missing" Lights: They estimate there are likely more than 10,000 additional flickering lights that were so faint or changed so drastically that they appeared in only one snapshot and vanished in the other. They focused on the ones they could see in both pictures to be sure.

2. How Bright is "Bright"?

Not all flickering lights are created equal. The team looked at how much the brightness changed based on how bright the source was to begin with.

The Rule of Thumb: If a radio source is moderately bright (brighter than 20 milliJanskys, a unit of radio brightness), about 5% of them flicker by more than 30%.
The Brighter, The More Active: As the sources get even brighter (over 300 milliJanskys), the percentage of flickering lights jumps to 9%. It seems the "loudest" radio stations are the ones most likely to change their tune.

3. Who Are These Flickering Lights?

The astronomers didn't just count the lights; they tried to figure out what they were. They used a "multi-wavelength" approach, which is like checking a person's ID card, their car, and their social media to figure out who they really are. They looked at infrared light (heat) and gamma rays (high-energy radiation).

The Main Characters (Blazars and Quasars): The vast majority of these flickering lights turned out to be Blazars and Quasars.
- Analogy: Imagine a lighthouse. A normal galaxy is like a lighthouse shining its beam sideways. A Blazar is a lighthouse that is pointed directly at your face. Because the beam is aimed right at us, it looks incredibly bright and changes intensity wildly as the "engine" inside the black hole at its center sputters and surges.
- The study found that Blazars are overrepresented among the flickering lights. In fact, the biggest absolute changes in brightness (the biggest jumps in volume) were caused by these Blazars.
The Small Actors (Galactic Sources): While Blazars caused the biggest total changes, the sources that showed the biggest percentage changes (the most dramatic relative flickers) were actually stars within our own Milky Way galaxy. These are often stars with strong magnetic fields that act like cosmic flares.

4. The "Flicker" vs. The "Fade"

The team noticed an interesting pattern in how these lights changed:

Flaring (Getting Brighter): When a source got brighter, it tended to do so with a slightly larger "kick" in brightness than when it got dimmer.
The Analogy: Think of a car engine. It might take a sudden, sharp burst of gas to speed up (flare), but it takes a longer, slower coasting period to slow down (fade). The data suggests radio sources brighten faster than they fade.

5. The "Quick Look" Challenge

The data they used came from "Quick Look" images.

Analogy: Imagine taking a photo with a smartphone immediately after a concert. It's good enough to see the crowd and the stage, but the colors might be slightly off, and there might be some blur. It's not a professional studio photo.
The astronomers had to be very careful because these "Quick Look" photos had some known quality issues (like slight errors in measuring brightness). They developed a strict set of rules and even had humans visually inspect the "flickering" candidates to make sure they weren't just camera glitches. After this rigorous check, they were confident in their list of 3,618 reliable variable sources.

6. The "Missing" Gamma Rays

Blazars are known to emit high-energy gamma rays. The researchers checked if their flickering radio sources also showed up in gamma-ray catalogs.

The Finding: Only about 1 in 12 of their variable radio sources had a matching gamma-ray source.
The Reason: Many of these Blazars are likely too faint or too far away (high redshift) for current gamma-ray telescopes to see, even though they are loud and bright in radio waves. This suggests that radio surveys are finding a "hidden population" of Blazars that other telescopes miss.

Summary

In short, this paper is a map of the "flickering" radio universe. It tells us that while the radio sky is mostly stable, there is a significant population of changing lights. Most of these are Blazars—supermassive black holes shooting jets of energy directly at Earth—acting like cosmic strobe lights. The study confirms that radio telescopes are excellent tools for finding these dynamic objects, even those that are too faint for other types of telescopes to detect.

Technical Summary: A Census of Variable Radio Sources at 3 GHz

Problem and Motivation
Radio variability is a ubiquitous phenomenon across diverse astronomical timescales, ranging from milliseconds in pulsars to decades in Active Galactic Nuclei (AGN). While dedicated monitoring of known variables and narrow-field surveys have provided insights, the broader population of variable radio sources remains poorly characterized due to limited sampling. Previous wide-area surveys, such as FIRST and NVSS, identified variability but were often limited by sensitivity or cadence. At lower frequencies, variability is rare, but at higher frequencies (above a few GHz), it becomes more common, yet a near-all-sky census at 3 GHz has been lacking. This work addresses the need to characterize the rate, scale, and nature of radio variability across the sky to better understand dynamic processes such as accretion mechanics, jet propagation, and stellar evolution.

Methodology
The authors utilize data from the first two epochs of the Karl G. Jansky Very Large Array Sky Survey (VLASS), a multi-epoch survey operating at $\nu \sim 3$ GHz. The analysis relies on the VLASS Quick Look images and catalogs, which, while subject to known quality limitations (e.g., flux density scaling errors and artifacts), cover the entire sky north of $-40^\circ$ declination.

Data Selection: The study focuses on compact sources detected in both Epoch 1 (2017–2019) and Epoch 2 (2020–2022), separated by a median time of 974 days. Sources were filtered for quality flags indicating real detections and resolved/unresolved status based on the ratio of integrated to peak flux density ( $S_{total}/S_{peak}$ ) versus signal-to-noise ratio.
Systematic Correction: To account for non-homogeneous flux calibration across the survey footprint, the authors analyzed the ratio of median peak flux densities between epochs on a per-tile basis. They identified a bimodality in flux ratios dependent on the sub-epoch of observation and corrected the Epoch 2 flux densities to ensure a median ratio of unity within each tile.
Variability Identification: Candidate variables were identified using two parameters: the Variability Statistic ( $V_s$ ), representing the absolute flux change relative to measurement uncertainty, and the modulation index ( $m$ ), representing the fractional change. Sources were selected if $|V_s| > 4\sigma_{V_s, Tile}$ and $|m| > 0.26$ .
Quality Assurance: A visual inspection of 4,124 candidates was conducted using a three-panel image format (Epoch 1, Epoch 2, and difference image) with convolved resolution to mitigate Point Spread Function (PSF) variations. A consensus flag ( $f_{vi}$ ) was assigned based on independent classifications by three authors. The final robust sample consists of 3,618 sources where all classifiers agreed the source was a real variable.
Multiwavelength Characterization: The variable sample was cross-matched with SDSS, LS DR8 (for redshifts), AllWISE (for infrared colors), and the 4FGL Fermi-LAT catalog (for $\gamma$ -ray counterparts) to determine the nature of the variable sources.

Key Results

Variability Rate: Approximately 3,600 compact sources were found to vary significantly. For sources with a mean flux density ( $\mu_S$ ) brighter than 20 mJy, 4.9% exhibit variability ( $|m| > 0.26$ ). This fraction increases with brightness, reaching 9.0% for sources with $\mu_S > 300$ mJy.
Flux Changes: The study distinguishes between flaring and fading sources. While the number of flaring and fading sources is roughly equal, flaring sources exhibit slightly larger absolute flux changes (median $|\Delta S| = 15.6$ mJy) compared to fading sources (median $|\Delta S| = 13.4$ mJy), a difference found to be statistically significant ( $p = 4 \times 10^{-4}$ ).
Source Nature: Multiwavelength analysis reveals that the majority of variable radio sources are extragalactic.
- Blazars: Infrared color analysis indicates that blazars are significantly overrepresented among variable sources compared to the general radio population. Approximately 68% of variables with WISE colors are consistent with blazars (56% BL Lacs, 40% FSRQs), compared to 42% in the parent sample.
- Galactic Sources: The largest fractional changes in brightness are exhibited by galactic sources, specifically radio stars (e.g., RS CVn systems) and microquasars.
- $\gamma$ -ray Correlation: Variable radio sources are more likely to have a $\gamma$ -ray counterpart in the 4FGL catalog (8%) compared to the general radio population (2%). Furthermore, these counterparts show higher $\gamma$ -ray variability indices, suggesting a correlation between radio and high-energy variability.
Redshift Distribution: The redshift distribution of variable sources mirrors that of the non-variable parent population, suggesting variables exist throughout the same epochs of cosmic history. However, variables with $\gamma$ -ray detections tend to be at lower redshifts ( $z \sim 0.64$ ) than those without ( $z \sim 1.22$ ).
Extreme Examples: The paper identifies specific extreme cases, including LB 2449 (a BL Lac with a five-fold brightness increase) and PKS 1502+106 (a Flat Spectrum Radio Quasar with a nearly 50% drop). Galactic examples include Cygnus X-3 (a microquasar showing a 5.4-fold increase) and $\sigma$ Gem (an RS CVn star showing a large fractional decrease).

Significance and Claims
The authors claim that this work provides the first near-all-sky census of variable radio sources at 3 GHz, establishing a baseline for the population of variables on $\sim$ 2.5-year timescales. The study demonstrates that:

Blazars dominate the variable population: The overrepresentation of blazars suggests that deep radio time-domain surveys are effective tools for identifying faint, high-redshift blazars that may be missed by current $\gamma$ -ray surveys.
Methodological robustness: Despite the use of Quick Look images with known systematic errors, the per-tile statistical approach and visual inspection successfully isolated real variables, with contamination from spurious detections appearing minimal.
Galactic utility: The survey is effective at identifying galactic radio stars and microquasars, particularly those exhibiting extreme fractional variability.
Future potential: The authors note that the addition of a third VLASS epoch will allow for three-point variability statistics, improving the detection of stochastic variables and identifying sources with consistent long-term trends (e.g., newly switched-on jets). They also highlight the potential of identifying "quasi-stable" faint point sources for use as phase calibrators for next-generation radio telescopes like the ngVLA and SKA.

The paper concludes that radio variability surveys are a powerful, complementary method for studying AGN physics and stellar evolution, capable of revealing populations of objects that are faint or at high redshifts relative to those found in other wavebands.