🔭 astrophysics

CHIME/Slow overview and pilot survey: A new backend to search for second-duration radio transients with the CHIME telescope

This paper introduces the CHIME/Slow backend for detecting second-duration radio transients, reports the discovery of nine bursts (including a new non-repeating source) from a pilot survey, and derives an all-sky rate for such transients based on the pipeline's sensitivity and performance.

Original authors: Sujay Mate, Kevin Luke, Yash Bhusare, Arvind Balasubramanian, Ziggy Pleunis, Paul Scholz, Shriharsh P. Tendulkar, Mohit Bhardwaj, Charanjot Brar, Fengqiu Adam Dong, Emmanuel Fonseca, B. M. Gaensler, J

Published 2026-07-03

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Sujay Mate, Kevin Luke, Yash Bhusare, Arvind Balasubramanian, Ziggy Pleunis, Paul Scholz, Shriharsh P. Tendulkar, Mohit Bhardwaj, Charanjot Brar, Fengqiu Adam Dong, Emmanuel Fonseca, B. M. Gaensler, Jason Hessels, Jeff Huang, Naman Jain, Ronniy C. Joseph, Victoria M. Kaspi, Afrokk Khan, Robert Main, Bradley W. Meyers, Nikola Milutinovic, Kenzie Nimmo, Kaitlyn Shin, David Spear, Ingrid Stairs, Chia Min Tan

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

The Big Picture: Looking for the "Slow" Radio Bursts

Imagine the universe is a giant, noisy radio station. For years, astronomers have been tuning in to catch "Fast Radio Bursts" (FRBs)—these are like lightning bolts of radio energy that flash for just a millisecond (a thousandth of a second). We have great tools to catch these quick flashes.

However, there is a gap in our knowledge. What if there are radio bursts that last longer? Think of them not as lightning bolts, but as slow-burning campfires or long, drawn-out drumbeats that last for seconds. For a long time, our telescopes were like high-speed cameras set to only catch lightning; they were too fast and too sensitive to short bursts to notice these slower, longer events.

This paper introduces a new tool called CHIME/Slow. It is a new "backend" (a piece of software) designed specifically to catch these second-long radio transients using the CHIME telescope in Canada.

The New Tool: CHIME/Slow

The CHIME telescope is like a giant, curved mirror made of many smaller mirrors (antennas) that listens to the sky. The standard software (CHIME/FRB) is excellent at finding quick flashes, but it has a blind spot for things that last longer than a tenth of a second.

CHIME/Slow is like putting on a pair of "slow-motion glasses."

How it works: Instead of looking at the data in tiny, fast slices, it groups the data into larger chunks (16 milliseconds, 128 milliseconds, and even up to 512 milliseconds).
The Trade-off: To do this, it simplifies the data slightly (like compressing a high-resolution photo into a slightly smaller file) so the computer doesn't get overwhelmed. This allows it to hunt for signals that last from 16 milliseconds up to 5 seconds.

The Pilot Survey: A Test Run

Before running this new tool 24/7, the team did a "pilot survey" (a test run) using old data from late 2022 and early 2023. They looked at about 17 days of sky time.

The Results:
They found nine bursts in total.

Eight of them came from a known "hyperactive" source called FRB 20220912A. This source is like a firework that keeps going off repeatedly. Interestingly, two of these bursts were only found by the new CHIME/Slow tool. The old tool missed them because they were too long and "scattered" (blurred out) for the fast camera to see clearly.
One of them was a brand new discovery: FRB 20230204C. This was a "one-off" event (it didn't repeat). It was a very "scattered" burst, meaning the signal got smeared out as it traveled through space, making it look like a long, fuzzy tail. The new tool caught it perfectly, while the old tool only saw a tiny, faint glimpse of it in the side of its view.

Why This Matters: Filling the Gap

The paper explains that for a long time, we didn't know if these "slow" radio bursts existed in the numbers we expected.

The Analogy: Imagine you are trying to count fish in a pond. You have a net with very small holes. You catch thousands of tiny minnows (fast bursts), but you miss the big, slow-moving whales because they slip right through the holes or you aren't looking for them.
The Discovery: By using the new "net" (CHIME/Slow), they found that these "whales" (second-long bursts) are likely much more common than we thought.

The Numbers: How Common Are They?

Based on finding just one new, non-repeating burst in their test, the team calculated how many of these events might be happening across the entire sky every day.

They estimate that there are between 184 and 4,556 of these second-long radio bursts happening every day across the whole sky.
This range is wide because they only found one new example, but it proves that these events are not rare anomalies; they are a significant part of the radio universe.

What's Next?

The paper concludes that this new tool works. It successfully found bursts that the old tools missed, especially those that are "scattered" (blurred) by the space they travel through.

The team is currently upgrading the system to run in real-time.
Once fully operational (expected by mid-to-late 2026), the CHIME telescope will be able to scan the entire sky for these slow bursts as they happen, alerting astronomers immediately so they can study these mysterious cosmic events.

In summary: The paper introduces a new way of listening to the universe that isn't just looking for lightning-fast flashes, but is also tuned to catch the slower, longer radio "drumbeats" that were previously hiding in plain sight.

Technical Summary: CHIME/Slow Overview and Pilot Survey

Problem Statement
While the construction of radio interferometers with large mapping speeds has led to a proliferation of Fast Radio Burst (FRB) detections, a significant gap remains in the parameter space of radio transients. Most instruments, including the standard CHIME/FRB backend, are optimized for millisecond-duration bursts and lack sensitivity to the widest and most highly scattered events (timescales $\gtrsim 100$ ms). This limitation arises from two primary factors: the lack of gain stability in radio telescopes at second-timescales due to Radio Frequency Interference (RFI) and bright sources transiting through sidelobes, and the computational challenges of performing image-plane searches at very high time resolutions. Consequently, the parameter space for second-duration radio transients (ranging from hundreds of milliseconds to seconds) remains largely unexplored, despite theoretical predictions that compact binary mergers and Gamma-Ray Bursts (GRBs) could produce such flashes.

Methodology
To address this gap, the authors developed CHIME/Slow, a new real-time transient search backend for the Canadian Hydrogen Intensity Mapping Experiment (CHIME) telescope. The pipeline is designed to search for radio transients with pulse widths between 16 ms and 5 s.

Data Processing: CHIME/Slow utilizes a $\sim$ 3-bit Huffman compressed intensity data stream derived from the CHIME/FRB 8-bit stream. To reduce computational costs, the pipeline downsamples the frequency resolution from 16,384 channels to 1,024 channels and employs three distinct time-sampling streams: 16 ms, 128 ms, and 512 ms.
Search Pipeline: The pipeline integrates established tools including PRESTO (for dedispersion and single pulse search), HDBSCAN (for clustering events in Dispersion Measure (DM)-time space), and FETCH (for classifying candidates as astrophysical or RFI).
- RFI Cleaning: Performed using rfifind with parameters optimized for long-duration transients.
- Dedispersion: Data are dedispersed from 50 to 3,000 pc cm $^{-3}$ .
- Search: Single pulse searches are conducted independently on each time-sampling stream with detection thresholds set at a Signal-to-Noise (S/N) ratio of 10.
Pilot Survey: The prototype pipeline was tested on archival data collected in December 2022, January 2023, and February 2023. This dataset provided 17 days of on-sky time with an instantaneous field of view of $\sim$ 13 deg $^2$ .
Completeness Analysis: To estimate sensitivity, the authors injected $\sim$ 10,000 simulated pulses across the three data streams with varying DMs, pulse widths, scattering measures, and S/N values. Completeness was defined as the fraction of injected pulses recovered by the pipeline after clustering.

Key Results

Detection: The pilot survey detected nine bursts. One burst, FRB 20230204C, was identified as a new non-repeating source. The remaining eight bursts originated from the known hyperactive repeater FRB 20220912A.
Comparison with CHIME/FRB: Two of the bursts from FRB 20220912A were detected only by CHIME/Slow, not by the standard CHIME/FRB pipeline. Conversely, the non-repeating FRB 20230204C was detected by CHIME/FRB only in a beam sidelobe; in that detection, it exhibited a shorter width and narrower bandwidth compared to the CHIME/Slow detection.
Source Properties:
- FRB 20230204C: Exhibited an intrinsic width of $3.49 \pm 1.32$ ms but a scattering timescale of $98.31 \pm 8.33$ ms at 400 MHz, making it one of the most highly scattered FRBs detected by CHIME.
- FRB 20220912A: Intrinsic widths ranged from 10.22 to 25.48 ms, with scattering timescales between 9.59 and 75.06 ms for the scattered bursts.
All-Sky Rate Estimation: Using the single detection of FRB 20230204C and the completeness analysis, the authors estimated the all-sky rate of radio transients with burst widths between 16 ms and 5 s, fluence $> 5$ Jy ms, and observing frequency of 600 MHz. The estimated rate is between 184 and 4,556 bursts sky $^{-1}$ day $^{-1}$ (95% credible region).

Significance and Claims
The paper claims that CHIME/Slow successfully demonstrates the capability to detect intrinsically wide and highly scattered radio transients that are missed by standard millisecond-optimized pipelines. The detection of FRB 20230204C, one of the most highly scattered FRBs known, validates the hypothesis that current surveys tuned for short durations are missing a population of scattered events.

The estimated all-sky rate is consistent with rates from other surveys (such as ASKAP/CRACO-PS and Parkes-70cm) that searched for wider pulse widths, suggesting that the population of second-duration transients is significant. The authors emphasize that while the pilot survey was a proof-of-concept, the results support the existence of an undetected population of highly scattered events in existing CHIME/FRB data. The work outlines a path toward a fully operational real-time backend expected by mid-to-late 2026, which will systematically search the entire CHIME sky for these transients and provide low-latency alerts to the community. The authors note that future work with a larger sample of detections is necessary to better constrain the dependence of the rate on Dispersion Measure and pulse width.