Here is a curated list of relevant astro-ph submissions from February 2026 that align with your research interests in 21-cm cosmology, the Epoch of Reionization (EoR), foreground mitigation, and machine learning. Each paper is mapped directly to the relevant CASSA projects.

Authors: Anirban Chakraborty, Kwanit Gangopadhyay, Arka Banerjee, Tirthankar Roy Choudhury

arXiv Link: arXiv:2602.15803

Summary: To overcome severe contamination from astrophysical foregrounds and instrumental noise, this paper investigates k-nearest-neighbour cumulative distribution functions as an alternative to two-point statistics. The method shows promise in extracting maximal information from 21cm-galaxy cross-correlations even under aggressive foreground filtering.

Relevant CASSA Projects: RiFInE, PrimaBERA, MiMiC

Authors: Sukhdeep Singh Gill, Khandakar Md Asif Elahi, Somnath Bharadwaj, Shiv K. Sethi, Akash Kumar Patwa

arXiv Link: arXiv:2602.17304

Summary: This paper demonstrates the mitigation of dominant astrophysical foregrounds using Smooth Component Filtering (SCF) to estimate the 21 cm bispectrum from Murchison Widefield Array (MWA) observations, establishing new limits on the EoR signal.

Relevant CASSA Projects: RiFInE, PrimaBERA, START

Authors: Pascal M. Keller, Bojan Nikolic, Nithyanandan Thyagarajan

arXiv Link: arXiv:2602.04965

Summary: Aims to establish a mathematical relationship between calibration-independent closure phases and the cosmological 21 cm power spectrum. It provides a robust technique to sidestep calibration inaccuracies that typically compromise the spectral fidelity of the smooth foreground continuum.

Relevant CASSA Projects: CHronOS, START

Authors: Hemanth Potluri, Manas Mohit Dosibhatla, Leon Noble, Chandra Shekhar Murmu, Suman Majumdar, Samit Kumar Pal, Saswata Dasgupta, Satadru Bag, Abhirup Datta

arXiv Link: arXiv:2602.16651

Summary: By analyzing 21-cm brightness temperature fields in the image domain to retain non-Gaussian characteristics, this work tracks the coalescence of ionized regions using the Largest Cluster Statistic (LCS) to infer the topology of the EoR along a lightcone.

Relevant CASSA Projects: PrimaBERA

Authors: Kaifeng Yu, Xin Wang

arXiv Link: arXiv:2602.03313

Summary: Employs a deep learning approach to recover large-scale baryon acoustic oscillations (BAO) lost to foreground avoidance strategies. It attempts to restore missing large-scale modes by exploiting physical non-linear mode coupling from short-wavelength data.

Relevant CASSA Projects: MiMiC, RiFInE

Authors: Sudipta Sikder, Rajesh Mondal, Rennan Barkana, Anastasia Fialkov

arXiv Link: arXiv:2602.13445

Summary: Investigates how an inhomogeneous excess galactic radio background and line-of-sight fluctuations from early galaxies introduce significant non-Gaussianity into the Cosmic Dawn 21-cm signal, which can be measured using the squeezed-limit bispectrum.

Relevant CASSA Projects: PrimaBERA, CHronOS

Authors: Emilie Thélie, Sarah Libanore, Yonatan Sklansky, Julian B. Muñoz, Ely D. Kovetz

arXiv Link: arXiv:2602.12277

Summary: Because directly mapping ionized bubbles is challenging due to high redshifts and intense foregrounds, this paper reconstructs the ionized-bubble size distribution during the EoR through the real-space cross-correlation of 21-cm and star-forming line-intensity maps.

Relevant CASSA Projects: PrimaBERA