Debiasing the Observed Fast Radio Burst Population with the CHIME/FRB Selection Function

Kyle McGregor, Jason W. T. Hessels, Victoria M. Kaspi, Kaitlyn Shin, Fengqiu Adam Dong, Naman Jain, Robert Main, Mawson Sammons, Michele Woodland, Daniel Amouyal, Derek Bingham, Charanjot Brar, Amanda M. Cook, Radu V. Craiu, Alice Curtin, Gwendolyn Eadie, B. M. Gaensler, Jeff Huang, Afrokk Khan, Calvin Leung, Kiyoshi W. Masui, Ayush Pandhi, Swarali S. Patil, Aaron B. Pearlman, Sachin Pradeep E. T, Paul Scholz, Seth Siegel, Kendrick M. Smith, David Stenning · 2026

Read Debiasing the Observed Fast Radio Burst Population with the CHIME/FRB Selection Function by Kyle McGregor, Jason W. T. Hessels, Victoria M. Kaspi, Kaitlyn Shin, Fengqiu Adam Dong, Naman Jain, Robert Main, Mawson Sammons, Michele Woodland, Daniel Amouyal, Derek Bingham, Charanjot Brar, Amanda M. Cook, Radu V. Craiu, Alice Curtin, Gwendolyn Eadie, B. M. Gaensler, Jeff Huang, Afrokk Khan, Calvin Leung, Kiyoshi W. Masui, Ayush Pandhi, Swarali S. Patil, Aaron B. Pearlman, Sachin Pradeep E. T, Paul Scholz, Seth Siegel, Kendrick M. Smith, David Stenning (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-26334.txt.

Introduction

Kyle McGregor and the CHIME/FRB team correct the census of fast radio bursts for the telescope's own blind spots. Fast radio bursts are millisecond flashes from deep space, and CHIME catalogues them by the hundreds. What the survey misses skews what astronomers infer. Hundreds of thousands of synthetic bursts were injected into the live search pipeline to map that selection effect honestly. Catalog 2, the largest sample yet, gets reanalysed with the map in hand.

Four burst properties get intrinsic distributions: fluence, scattering time, pulse width, and dispersion measure. Scattering times were the weak spot last time, starved of statistics at long delays. A logistic regression on the injected events builds an explicit detection probability with interaction terms. Simulation-based inference folds that selection function back in. The scattering distribution shows a slight downturn at long times, though flat or gently rising shapes survive.

Catalog record

CREATORKyle McGregor, Jason W. T. Hessels, Victoria M. Kaspi, Kaitlyn Shin, Fengqiu Adam Dong, Naman Jain, Robert Main, Mawson Sammons, Michele Woodland, Daniel Amouyal, Derek Bingham, Charanjot Brar, Amanda M. Cook, Radu V. Craiu, Alice Curtin, Gwendolyn Eadie, B. M. Gaensler, Jeff Huang, Afrokk Khan, Calvin Leung, Kiyoshi W. Masui, Ayush Pandhi, Swarali S. Patil, Aaron B. Pearlman, Sachin Pradeep E. T, Paul Scholz, Seth Siegel, Kendrick M. Smith, David Stenning
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE86 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2606.26334

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Debiasing the Observed Fast Radio Burst Population with the CHIME/FRB Selection Function Kyle McGregor, Jason W. T. Hessels, Victoria M. Kaspi, Kaitlyn Shin, Fengqiu Adam Dong, Naman Jain, Robert Main, Mawson Sammons, Michele Woodland, Daniel Amouyal, Derek Bingham, Charanjot Brar, Amanda M. Cook, Radu V. Craiu, Alice Curtin, Gwendolyn Eadie, B. M. Gaensler, Jeff Huang, Afrokk Khan, Calvin Leung, Kiyoshi W. Masui, Ayush Pandhi, Swarali S. Patil, Aaron B. Pearlman, Sachin Pradeep E. T, Paul Scholz, Seth Siegel, Kendrick M. Smith, David Stenning Abstract The recent release of CHIME/FRB Catalog 2 provides the largest sample to date with which to investigate the intrinsic distributions of fast radio bursts (FRBs). Leveraging an expanded campaign of 587,367 synethetic bursts injected into the live CHIME/FRB search pipeline, we perform a population analysis of the fluence, scattering timescale, pulse width, and dispersion measure distributions of Catalog 2 FRBs. We first infer the intrinsic population using a resampling-based framework that accounts for instrumental selection effects following previous CHIME/FRB population studies. A central goal of this work is to constrain the intrinsic distribution of scattering timescales, that remained weakly constrained in Catalog 1 owing to limited statistics at moderate and large scattering times ( τ ≳ 10 ​ ms \tau\gtrsim 10\,\mathrm{ms} at 600 MHz) and sparse injection coverage in this regime. Second, we construct an explicit multidimensional selection function by training a logistic regression model on the injected events. This model estimates the detection probability as a function of FRB observable properties, including higher-order interaction terms. We incorporate this selection function into a simulation-based inference framework to refine the inferred intrinsic scattering-timescale distribution. We find evidence for a slight downturn in the intrinsic FRB scattering timescale distribution, though a flat or slightly rising distribution cannot be ruled out, that is further supported through a comparison with the higher-frequency scattering timescale distribution observed by Commensal Real-time ASKAP Fast Transients (CRAFT) survey. I Introduction Fast radio bursts (FRBs) are bright, millisecond-scale radio transients that have emerged as an intriguing frontier of time-domain survey astrophysics. First discovered nearly 20 years ago ( 31 ) , the study of FRBs has matured from a handful of enigmatic discoveries into a statistically rich population with the advent of wide-field radio surveys. In particular, with the release of more than 4500 FRBs observed by the Canadian Hydrogen Intensity Mapping Experiment ( 7 , hereafter referred to as CHIME/FRB Catalog 2;) the cumulative number of known FRB sources has greatly increased, enabling population-level analyses with improved statistical precision relative to earlier studies ( 58 ; 42 ; 7 ; 53 , Jain et al. in prep.) . Selection effects are inherent to every astrophysical survey and are implicitly encoded in the sample detected by CHIME/FRB. Correcting for these effects is therefore imperative for making unbiased inferences about the underlying FRB population. Formally, these corrections are encoded through a survey’s selection function, which specifies the probability that an instrument will detect a given source with certain incident properties. With the emergence of large-scale, real-time transient search pipelines ( 9 ; 23 ; 28 ; 55 ; 24 , e.g.) , the characterization of selection effects has become increasingly complex, as detection is mediated by multi-stage triggering, hierarchical data products, and nonlinear interactions between burst properties and search algorithms. Complex radio-frequency interference environments may also cause false positive and false negative triggers ( 44 ; 25 ; 8 , e.g.) . This motivates the development of empirical selection function modeling, which in FRB and pulsar surveys is routinely achieved through injection-based tests of search pipelines ( 29 ; 13 ; 46 , e.g.,) . For these studies, synthetic signals are injected directly into the data streams and processed through the search pipeline to empirically measure detection efficiency. To correct for selection biases in the first large-scale catalog of CHIME/FRB sources ( 13 ) ( \sim\!500 FRBs, hereafter referred to as Catalog 1), a population synthesis was performed using a large dataset of synthetic bursts injected into the instrument’s real-time live search pipeline (hereafter, injections ). This dataset, along with the observed catalog, was applied in Section 6 of 13 to infer a fiducial, selection-corrected model for the marginalized distributions of FRB observables; the injections were subsequently leveraged by 56 to constrain the redshift and energy distributions of FRBs. The design and implementation of CHIME/FRB’s injection pipeline, as well as a detailed description of the resulting injection dataset and its intended use, were presented by 35 . In the work here, we build on the 13 analysis by fitting a parametric selection function model that explicitly accounts for correlations between observables, employing a logistic regression framework to more accurately capture the structure of the CHIME/FRB detection process and extend the range of the parameter space to better quantify CHIME/FRB’s sensitivity to highly-scattered FRBs. This is in accompaniment to fitting an analogous single-variate model to 13 using Catalog 2 and a newly-expanded set of injections. We structure this paper as follows: In Section II we describe the CHIME/FRB injection pipeline and the generation of a large population of synthetic bursts. In Section III we present a procedure for fitting a fiducial model for the intrinsic FRB observable distributions to the marginalized distributions of observed quantities, using an importance-sampling procedure similarly applied by 13 . In Section IV we use the injections to fit a logistic regression model for the selection function. In Section V we refine the estimate of the intrinsic scattering timescale distribution using a simulation-based inference procedure to better constrain the behavior of the distribution at large scattering times. In Section VI we compare the resulting debiased distributions to previous work and to results from other FRB surveys, and in Section VII we present our conclusions. II CHIME/FRB Injections This analysis requires a large catalog of synthetic FRB events injected into the CHIME/FRB real-time intensity data stream. These injections are processed by the same real-time detection, classification, and triggering pipeline used to identify astrophysical bursts, enabling an empirical characterization of the CHIME/FRB selection function. The injection framework used in this analysis was originally introduced by 35 , who included an overview of the synthetic pulse generation pipeline and database networking structure that remain unchanged since the 13 deployment. Here we summarize the architectural and population-level details most relevant to the present analysis. II.1 Injection Pipeline Architecture The injection system described in 35 is designed to forward-model synthetic FRB signals through the CHIME/FRB real-time search pipeline, enabling direct measurement of survey completeness as a function of intrinsic burst properties. The CHIME/FRB search pipeline itself is implemented as a hierarchical, multi-stage software stack operating on beamformed intensity data generated by the CHIME F/X-engine (correlator and beamformer), as described in detail in the CHIME/FRB system overview ( 9 ) . In this architecture, signal processing proceeds through four principal stages, denoted L1 through L4, following an initial beamforming and upchannelization stage (L0). At L0, voltage data from the CHIME antennas are beamformed into 1024 static, FFT-formed intensity beams spanning the primary field of view, with 16,384 frequency channels at 0.983 ms cadence ( 9 ) . These intensity streams are distributed across a dedicated L1 compute cluster, with each L1 node processing data from eight static beams. The

Read the complete text