School and Workshop on 21-cm Cosmology

3 Mar 2026, 08:00 → 11 Mar 2026, 17:35 Asia/Kolkata

Tata Institute of Fundamental Research (Mumbai)

School: 3–6 March 2026

Workshop: 9–11 March 2026

Registration deadline: 30 January 2026 (Registration closed)

The Cosmic Dawn and the Epoch of Reionization mark the Universe’s last major phase transition: the formation of the first stars and black holes, and the transformation of the intergalactic medium from cold, neutral hydrogen into a hot, ionised plasma. Despite rapid progress, central questions remain open—when reionisation began and ended, how extended and patchy it was, and which sources dominated the ionising photon budget. We are now entering a decisive observational era: pathfinder low-frequency arrays (LOFAR, HERA, MWA, NenuFAR, uGMRT) are delivering increasingly stringent constraints on the redshifted H I 21-cm signal, while JWST is rapidly sharpening our empirical view of galaxies and black holes at –12. In this landscape, SKA-Low is the step-change facility expected to deliver precision 21-cm statistics—making this an especially timely moment to consolidate training and community readiness.

Organised under the Epoch of Reionization & Cosmology Working Group of the SKA India Consortium (SKAIC), this event combines a four-day School with a three-day Workshop to strengthen national capability and accelerate SKA-Low preparedness. The School (aimed at ~50 graduate students and early-stage researchers) will provide structured training in both foundations and modern practice, combining lectures with guided tutorials and hands-on sessions. The Workshop will bring together the broader community for project updates, focused topical discussions, and dedicated time for collaboration—culminating in shared priorities and concrete plans for the coming year.

School focus areas (lectures + hands-on tutorials):

• Instrumentation basics and measurement systematics
• Calibration concepts and practical workflows
• Foreground modelling and mitigation strategies
• Signal modelling and astrophysical interpretation
• Power-spectrum estimation and validation
• End-to-end simulations and parameter inference
• Multi-wavelength cross-correlation and synergies with JWST-era constraints

Workshop format and goals:

• Sessions on methods, data challenges, and lessons from current pathfinder datasets
• Progress reports from ongoing Indian projects
• Collaborative planning for SKA-Low readiness and high-impact science

Participation and selection

• Mode: In-person (TIFR Colaba campus)

School participants will be selected via a short application to maximise scientific fit and national representation. Travel support will be prioritised for early-career researchers without other funding.

Tuesday, 3 March

08:30 → 09:00 Breakfast

09:00 → 09:20 Registration

09:20 → 09:30 Welcome

Convener: Girish Kulkarni (TIFR)

09:30 → 10:45 Lecture 1: The 21-cm signal and its physics

Convener: Tirthankar Roy Choudhury (NCRA-TIFR, Pune)

10:45 → 11:15 Tea/Coffee

11:15 → 12:30 Lecture 2: Statistical measures of the 21-cm Signal

Convener: Samir Choudhuri (IIT Madras)

12:30 → 13:30 Lunch

13:30 → 14:30 Guided discussion: Current constraints on reionization and cosmic dawn

Convener: Tirthankar Roy Choudhury (NCRA-TIFR, Pune)

14:30 → 15:30 Free time

15:30 → 16:00 Tea/Coffee

16:00 → 18:00 Lab/tutorial: Model signal using a semi-analytical code and compute power spectra

Conveners: Samir Choudhuri (IIT Madras), Janakee Raste (NCRA-TIFR, Pune)

Wednesday, 4 March

08:30 → 09:30 Breakfast

09:30 → 10:45 Lecture 3: Basic radio astronomy and introduction to SKA-Low and GMRT

Convener: Narendranath Patra (IIT Indore)

10:45 → 11:15 Tea/Coffee

11:15 → 12:30 Lecture 4: Foregrounds, systematics, calibration

Convener: Prasun Dutta (IIT BHU)

12:30 → 13:30 Lunch

13:30 → 14:30 Free time

14:30 → 15:30 Guided discussion: Recent upper-limit papers from LOFAR and HERA

Convener: Prasun Dutta (IIT BHU)

15:30 → 16:00 Tea/Coffee

16:00 → 18:00 Lab/tutorial: Simulate visibilities for a small array layout

Conveners: Narendranath Patra (IIT Indore), Atharva Mirashi (IIT Indore)

Thursday, 5 March

08:30 → 09:30 Breakfast

09:30 → 10:45 Lecture 5: Parameter inference

Convener: Raghunath Ghara (IISER Kolkata)

10:45 → 11:15 Tea/Coffee

11:15 → 12:30 Lecture 6: 21-cm synergies with other probes of the high-redshift universe

Convener: Girish Kulkarni (TIFR, Mumbai)

12:30 → 13:30 Lunch

13:30 → 14:30 Free time

14:30 → 15:30 Guided discussion: Recent inference papers

Convener: Raghunath Ghara (IISER Kolkata)

15:30 → 16:00 Tea/Coffee

16:00 → 18:00 Lab/tutorial: Bayesian inference from mock data

Conveners: Raghunath Ghara (IISER Kolkata), Anirban Chakraborty (NCRA-TIFR, Pune)

Friday, 6 March

08:30 → 09:30 Breakfast

09:30 → 10:45 Lecture 7: Global signal

Convener: Saurabh Singh (RRI)

10:45 → 11:15 Tea/Coffee

11:15 → 12:30 Lecture 8: Introduction to AIML methods

Convener: Suman Majumdar (IIT Indore)

12:30 → 13:30 Lunch

13:30 → 15:30 Lab/tutorial: Inference using AIML

Conveners: Suman Majumdar (IIT Indore), Yashrajsinh Mahida (IIT Indore)

15:30 → 16:00 Tea/Coffee

16:00 → 17:00 Guided discussion: Recent global signal papers

Convener: Saurabh Singh (RRI)

17:00 → 18:00 Summary and reflection

Monday, 9 March

08:30 → 09:00 Breakfast

09:00 → 09:20 Registration

09:20 → 09:30 Welcome

09:30 → 10:05 Probing high redshifts with HERA

Convener: Saurabh Singh (RRI)

10:05 → 10:45 Session 1

Convener: Chair: Mayuri Rao (RRI)

10:05 Distinguishing between Reionisation Source Models in SKA-Low 21-cm Intensity Maps using the Largest Cluster Statistic

The redshifted 21-cm signal emitted by neutral Hydrogen (HI) is a promising probe to understand the evolution of the topology and morphology of ionised regions during the Epoch of Reionisation (EoR). The morphology of ionised regions allows us to infer the nature and properties of ionising sources, i.e., early galaxies and AGNs. Traditional Fourier statistics, such as the power spectrum, help us quantify the strength of fluctuations in this field at different length scales but do not preserve its phase information. Analysing the 21-cm brightness temperature field in the image domain retains its non-Gaussian characteristics and morphological information. The coalescence of multiple ionised regions to form one contiguous ionised region spanning the universe is known as percolation, and its onset is quantified by the value of the Largest Cluster Statistic (LCS) approaching 1. In this work, we perform a percolation analysis of interferometric 21-cm brightness temperature maps to be observed by SKA-Low by studying the redshift evolution of the LCS along a lightcone to distinguish between several simulated reionisation source models. We have extended previous results on reionisation model comparison from the analysis of coeval 21-cm maps to understand how the lightcone effect, which is inherent to observations where multiple spectral channels are used to understand the time evolution of the signal, biases the observed percolation behaviour and affects the distinguishability of the source models. We estimate the LCS of subvolumes of varying sizes in the 21-cm lightcone maps and study their redshift evolution across different reionisation scenarios using a moving-volume approach. We comment on the spectral bandwidth corresponding to the size of the optimal subvolume that allows the least biased recovery of the percolation transition under different reionisation scenarios.

Speaker: Manas Mohit Dosibhatla (IIT Indore)

10:25 Propagation, Refraction and Scattering Effects in Low-Frequency Radio Observations: Implications for Precision 21-cm Cosmology

"Detecting the redshifted H I 21-cm signal from the Cosmic Dawn and Epoch of Reionization requires exquisite control over propagation effects that distort low-frequency radio measurements before they reach the interferometer. In this talk, I will discuss how plasma-induced refraction, scattering and frequency-dependent path distortions, arising from both the terrestrial ionosphere and astrophysical plasma media, can bias precision 21-cm observables if not carefully modelled and mitigated.
Drawing on my ongoing research in radiative transfer through magnetised plasma including refraction and scattering of radio waves in the solar corona, I will outline a unified physical framework to understand ray bending, angular broadening and spectral mode-mixing in low-frequency radio propagation. I will highlight conceptual parallels between solar-coronal propagation and line-of-sight plasma effects relevant for EoR experiments particularly their impact on foreground chromaticity, calibration stability and power-spectrum contamination.
The talk will emphasise how propagation effects couple to instrumental response, potentially leaking foreground power into the EoR window, and how forward modelling via ray-based or wave-optics approaches can complement existing calibration and foreground-mitigation pipelines. I will also discuss prospects for incorporating physically motivated plasma models into end-to-end simulations and parameter-inference frameworks for SKA-Low.
Finally, I will outline how multi-frequency diagnostics and cross-correlations with external datasets can help disentangle propagation-induced systematics from cosmological signals. This work aims to contribute toward improving robustness in 21-cm measurements and strengthening SKA-Low readiness by bridging plasma physics, radiative transfer and observational cosmology."

Speaker: Ganesh Jaiswal (IIT Kanpur)

10:45 → 11:35 Tea/Coffee

11:35 → 12:10 Latest results from LOFAR

Convener: Raghunath Ghara (IISER Kolkata)

12:10 → 12:50 Session 2

Convener: Chair: Rashmi Sagar (IIT Indore)

12:10 Towards a robust ML framework in 21cm data analysis

The 21-cm signal from the Epoch of Reionization (EoR) is a powerful probe of the evolution of the Universe. However, accurate measurements of the EoR signal from radio interferometric observations are sensitive to efficient foreground removal, mitigating radio-frequency interference and accounting for instrumental systematics. We investigate the impact of the variations in the point spread function (PSF) on parameter estimation by simulating visibilities corresponding to input 21-cm maps as observed by the 128-antenna configuration of the Murchison Widefield Array (MWA) Phase II. These visibilities are imaged to obtain dirty images, which are then used to train a 2D convolutional neural network (CNN). To systematically assess the effect of PSF mis-modelling, we generate multiple test sets by varying the MWA's antenna layout, thereby introducing controlled variations in the PSF; we then feed these alternative PSF dirty images to our CNN trained using only dirty images with the PSF of the true antenna layout. Our results demonstrate that PSF variations introduce biases in the CNN's predictions of , with errors depending on the extent of PSF distortion. We quantify these biases and discuss their implications for the reliability of machine-learning-based parameter inference in 21-cm cosmology and how they can be utilized to improve the robustness of estimation against PSF-related systematics in future 21-cm surveys.

Speaker: Madhurima Choudhury (BITS Pilani, Hyderabad)

12:30 RRIvis: A GPU-Accelerated, Full-Polarization Visibility Simulator for Radio Interferometry

Precision modelling of instrumental effects is critical for radio interferometry, particularly for 21cm cosmology and Epoch of Reionisation (EoR) experiments, where detecting faint signals requires rigorous foreground subtraction. We present RRIvis, a new open-source Python package designed to address the need for accuracy and flexibility in simulating interferometric visibilities. Built on the RIME formalism, RRIvis provides a framework for modelling the full signal propagation chain. It implements a chain of Jones matrices, covering geometric delays, direction-dependent primary beams, ionospheric effects, polarization leakage, and complex gains. This allows researchers to simulate realistic instrumental systematics with the highest fidelity. A defining feature of RRIvis is its hardware-agnostic acceleration powered by JAX. This enables the code to run transparently on CPUs, NVIDIA/AMD GPUs, and Apple Silicon. The package prioritises usability, type-safe config, compatibility with standard data formats (such as CASA MS), and a modular design that allows users to implement custom Jones terms. We present validation results confirming its accuracy against established simulators and demonstrate its performance benchmarks, offering the community a tool for pipeline verification and experimental planning.

Speaker: Kartik Mandar (RRI)

12:50 → 13:50 Lunch

13:50 → 14:25 Foregrounds, Systematics, and Constraints on the 21-cm Signal: Lessons learnt from uGMRT Band-2

Convener: Rashmi Sagar (IIT Indore)

14:25 → 15:25 Session 3

Convener: Chair: Madhurima Choudhury (BITS Pilani, Hyderabad)

14:25 The Tracking Tapered Gridded Estimator for the 21-cm power spectrum from MWA drift scan observations – III. Multiple pointings

We present an analysis of 55 hours of drift-scan data from the Murchison Widefield Array (MWA), a precursor to the SKA. The observation spans right ascensions from $\alpha = 349.0^{\circ}$ to $70^{\circ}$, covering two well-studied MWA fields, EoR0 and EoR1, divided into 163 Pointing Centres (PCs), each corresponding to approximately 20 minutes of observation time. For each PC, we estimate both the angular power spectrum ($D_{\ell}$) and the power spectrum $P(k_{\perp}; k_{\parallel})$. Due to the periodic missing channels in MWA data and the strong spectral dependence of foregrounds, the power spectra exhibit horizontal stripes in the $(k_{\perp}; k_{\parallel})$ plane. Following Elahi et al. (2025), we apply the Smooth Component Filtering (SCF) technique to mitigate this effect. Our analysis reveals a gradual increase in power from $\alpha = 37.0^{\circ}$ to $70^{\circ}$, attributed to Fornax A entering the main lobe of the primary beam. After examining the $(k_{\perp}; k_{\parallel})$ space, we implement a masking procedure prior to computing $\Delta^{2}(k)$ for all PCs. I will present the upper limits of the 21cm signal for each pointing separately, as well as the incoherently combined analysis, which aims to achieve improved sensitivity. This sky region overlaps with EoR0, EoR1, and the HERA stripe, and identifying relatively uncontaminated regions will aid future SKA-EoR observations.

Speaker: Shouvik Sarkar (IIT Madras)

14:45 EoR 21 cm bispectrum measurements from MWA observations at z=8.2

While the detection of the power spectrum has been the primary goal of 21 cm observations, it does not tell us the whole story of the Epoch of Reionization (EoR). Bispectrum, the next higher-order statistic, can access the additional information on non-Gaussianity, mode coupling, and topology of HI distribution during the EoR. Observationally, the bispectrum can be computed from radio-interferometric data by correlating the 3-visibilities (https://arxiv.org/abs/2506.10526). I will talk about our recent efforts to measure the bispectrum using the drift scan MWA observations at z = 8.2 (https://arxiv.org/abs/2507.04964). A significant contribution of this work is to characterize the behavior of foregrounds and the identification of the 'foreground wedge' feature in the cylindrical bispectrum, analogous to the case of the power spectrum. We have performed the comprehensive study considering all possible triangle configurations and placed the preliminary upper limits on the EoR 21 cm bispectrum signal. I will talk on (i) why we need to go beyond 2-point statistics (ii) how to compute the bispectrum directly from radio-interferometric visibility data in an efficient way (iii) how the bispectrum of astrophysical foregrounds looks like (iv) current upper limits on the EoR bispectrum (v) lessons learned for the EoR bispectrum detection - the goal which aligns with the capabilities of the SKA.

Speaker: Sukhdeep Singh (IIT Kharagpur)

15:05 Probing ionized bubbles around luminous sources during reionization with SKA 21-cm observations

During the Epoch of Reionization, luminous galaxies and quasars are expected to produce large ionized regions embedded within the neutral intergalactic medium. Detecting these individual ionized bubbles in the redshifted HI 21-cm signal offers a direct probe of the ionizing sources and the evolving IGM. However, such detections remain challenging for current radio interferometers due to sensitivity limitations. We propose a visibility-based matched-filter method to detect and characterize individual ionized bubbles using SKA1-Low AA* and AA4 observations, significantly reducing the required observing time. We develop and validate a computationally efficient estimator that operates on gridded visibilities to detect ionized bubbles. This serves as an accurate alternative to the more computationally demanding bare estimator that uses all baselines and frequency channels. Further, we employ a non-parametric foreground-subtraction method based on Gaussian process regression, which minimizes loss of the HI 21-cm signal and yields improved signal-to-noise ratios. Our analysis indicates that ionized bubbles at redshifts ~7-8 can be detected with SNR> 10 using 100 hours of SKA1-Low observations. We further derive a scaling relation that connects the SNR to the bubble radius, redshift, total observing time, and the mean neutral hydrogen fraction of the surrounding IGM. This helps to quickly predict the observational outcome for any planned observations and is, therefore, useful for devising observational strategies. Finally, we apply a Bayesian likelihood framework with MCMC sampling to the residual visibilities to recover ionized bubble properties. The resulting posterior distributions demonstrate accurate recovery of the bubble parameters. This confirms the feasibility of robustly characterizing individual ionized regions with the SKA1-Low. In this presentation, we shall discuss our method and simulation pipeline, the matched-filter and Bayesian analysis frameworks, and the implications of our results for probing reionization with upcoming SKA observations.

Speaker: Arnab Mishra (Jadavpur University)

15:25 → 15:55 Tea/Coffee

15:55 → 16:30 Capturing the 21-cm Signal: Observations, Systematics, and Probabilistic Algorithms

Convener: Asif Elahi (IIT Madras)

16:30 → 17:10 Session 4

Convener: Chair: Janakee Raste (NCRA)

16:30 Constraining the neutral hydrogen fraction during reionization: Cross-simulation inference using power spectrum and bispectrum

The redshifted 21-cm signal from neutral hydrogen (HI) is a powerful probe of the early universe, particularly the Epoch of Reionization (EoR). Detecting this signal is highly challenging, as it is both extremely weak and strongly obscured by astrophysical foregrounds. Consequently, one typically relies on the statistical properties of the 21-cm fluctuations to extract astrophysical and cosmological information. While most studies rely on the power spectrum, this two-point statistic fails to fully capture the information present in the highly non-Gaussian 21-cm signal. In this work, we argue that incorporating the next higher-order statistics, specifically the bispectrum, can significantly improve parameter inference estimation. We utilize machine learning methods to build a fast emulator for both statistics, enabling an efficient constraint of the neutral hydrogen fraction with Markov Chain Monte Carlo (MCMC) techniques. We combine the power spectrum and bispectrum to demonstrate that constraints on the neutral hydrogen fraction have significantly improved, leading to a more comprehensive understanding of the reionization process.

Speaker: Anoop Krishna (NIT Calicut)

16:50 Metropolis-within-Gibbs Sampling for Emulator-Accelerated 21-cm Reionization Parameter Inference

The redshifted 21-cm signal from the Epoch of Reionization (EoR) is a powerful probe of the structure formation and ionization history of our Universe. However, extracting the reionization parameters from the observations of this signal requires an efficient and robust inference framework which can perform under the influence of strong parameter correlations and computationally expensive forward models. In this work, we developed and tested a Metropolis-within-Gibbs (MWG) sampling framework to infer three key EoR parameters—minimum halo mass Mmin ,ionizing efficiency Nion, and mean free path Rmf — using the 21-cm power spectrum as the observable. To enable fast likelihood estimation, we used an uncertainty-aware Bayesian Neural Network (BNN) emulator that predicts both the power spectrum and its predictive uncertainty. The likelihood was modelled as a multivariate Gaussian with a diagonal covariance including SKA-Low noise and emulator uncertainty. We compared MWG against a standard Metropolis-Hastings (MH) sampler that proposes all parameters jointly. Across three test cases considering observations at different stages of reionization, MWG recovered posterior constraints consistent with MH while improving sampling efficiency, yielding higher acceptance ratios, fewer sampling steps to converge, and smoother mixing. Building on this inference framework, we are currently developing a foreground-robust extension to perform inference of the 21-cm signal from a contaminated observation cube by masking foreground dominated low-k∥ line-of-sight Fourier modes. The remaining safe modes (which lie in the EoR window) are expected to have only signal, while the masked modes are reconstructed using Gaussian Constrained Realisations (GCR). A prior LoS covariance is expected to be learned from a large ensemble of signal-only simulations, enabling conditional sampling of masked modes for each pixel and generating an ensemble of reconstructed cubes. From these reconstructions, we estimate the power spectrum and its uncertainty, discarding the k-bins dominated by reconstruction variance, and finally infer the EoR parameters using a Bayesian emulator-based likelihood.

Speaker: Mayukh Mandal (IIT Indore)

Tuesday, 10 March

08:30 → 09:30 Breakfast

09:30 → 10:05 Cosmic Dawn 21-cm efforts

Convener: Mayuri Rao (RRI)

10:05 → 11:05 Session 5

Convener: Chair: Suman Majumdar (IIT Indore)

10:05 Non-Gaussian Signatures of High HI 21-cm Optical Depth in Real and Redshift Space during Cosmic Dawn

The HI 21-cm optical depth can be significantly high in regions with dense neutral hydrogen and low spin temperature during CD. Therefore, during this epoch the commonly used equation for calculating HI 21-cm differential brightness temperature considering linear approximation of optical depth tends to overestimate the signal compared to the signal using exact equation. These corrections will change the statistics of the signal considerably in particular for skewness, power spectrum and bi-spectrum. We study the numerical simulations we have used to accurately model the signal as well as the optical depth distribution during CD. There would be a suppression in the signal while using the accurate model of the signal. We see from our simulation high optical depth points will reside in the tail of the differential brightness temperature distribution which will lead to non-Gaussianity. Therefore, we discuss how the statistics will change using accurate model of the signal in comparison to linearly approximated one.

Speaker: Iffat Nasreen (Jadavpur University)

10:25 Global 21 cm Cosmology from Space

The Cosmic Dawn (CD) and the Epoch of Reionisation (EoR) mark two pivotal phases in the thermal and ionisation history of the Universe, driven by the formation of the first stars and galaxies. The redshifted 21-cm hyperfine transition of neutral hydrogen provides a powerful probe of these epochs. The global sky-averaged 21-cm monopole signal traces the mean evolution of the intergalactic medium and sets the baseline for power spectrum measurements from interferometers like SKA. This monopole signal directly constrains the timing, duration, and depth of key transitions during CD and EoR, thereby anchoring the mean brightness temperature evolution and providing physically motivated priors that help break degeneracies in SKA-Low power spectrum analyses. Together, global and fluctuation measurements enable a more complete reconstruction of the astrophysics governing early structure formation. Probing ReionizATion of the Universe using Signal from Hydrogen (PRATUSH) is a proposed space-based radiometer designed to measure the sky-averaged 21-cm signal from CD, operating over 55–110 MHz. Ground-based experiments are confronted with limitations caused by terrestrial radio frequency interference (RFI), ionosphere-induced chromaticity, and modified antenna response in the presence of objects on the horizon and associated terrains. PRATUSH seeks to operate in an orbit around the moon, making scientific observations when in the lunar farside shielded from both Earth and the Sun, alleviating the significant challenges faced by ground-based experiments. In this talk, we discuss the foreground removal and instrument design philosophy of PRATUSH, rooted in spectral smoothness. We also present the latest results from integrated laboratory concept model tests for receiver qualification, achieving a sensitivity of 3.4 mK at a spectral resolution of 610 kHz.

Speaker: Narendra S (RRI)

10:45 Machine Learning Framework for Reliable 21-cm Extraction from Single-Radiometer Observations

The Cosmic Dawn (CD) and the Epoch of Reionization (EoR) mark pivotal stages in the early evolution of the Universe, occurring within the first billion years after the Big Bang. Despite their importance, the physical properties of the intergalactic medium (IGM) during these epochs remain poorly constrained by observations. Current and forthcoming low-frequency radio experiments, such as EDGES, SARAS, MWA, and the SKA, aim to detect the redshifted 21-cm signal from neutral hydrogen. However, these efforts are challenged by severe foreground contamination, instrumental systematics, and the complexity of accurate foreground removal. Further complications arise from the Earth’s ionosphere, which introduces frequency-dependent distortions and beam chromatic effects that significantly impede the detection of the global 21-cm signal. Accurate modelling of these ionospheric effects requires accounting for processes such as refraction, absorption, and thermal emission. To overcome these challenges, it is crucial to assess the impact of each source of corruption when applying non-parametric signal-recovery methods. In this work, we are developing a robust machine-learning–based regression framework to recover global 21-cm signal parameters from sky-averaged observations that include contributions from astrophysical foregrounds, ionospheric distortions, and instrumental beam chromaticity. The framework also emphasizes the identification of optimal machine-learning models based on a balance between computational efficiency and predictive performance. Overall, this approach shows strong potential to improve ground-based global 21-cm experiments by enabling reliable signal recovery across both the Cosmic Dawn and the Epoch of Reionization, thereby offering new insights into the early Universe.

Speaker: Anshuman Tripathi (IIT Indore)

11:05 → 11:35 Tea/Coffee

11:35 → 12:10 Updates from the REACH project

Convener: Girish Kulkarni (TIFR)

12:10 → 12:50 Session 6

Convener: Chair: Yogesh Wadadekar (NCRA)

12:10 Convergence Tests for Lyman-alpha Radiative Transfer Simulations at Cosmic Dawn and its impact on 21cm Signal.

Interpreting the 21-cm signal from Cosmic Dawn requires accurate modeling of the Lyman-α background, yet most theoretical studies rely on ideal-gas assumptions and simplified treatments of multiple scattering. We present convergence tests of a Monte Carlo radiative transfer framework that directly follows Lyman-α photons in cosmological simulations, avoiding these approximations. By systematically varying numerical and physical parameters—resolution, photon statistics, propagation step length, stopping criteria, and box size—we identify the regimes where the simulated Lyman-α background is numerically stable. These results strengthen confidence in 21-cm signal predictions and provide a robust basis for extracting astrophysical and cosmological constraints from upcoming observations.

Speaker: Pooja Rani (TIFR)

12:30 Constraining Primordial Black Holes in light of CMB Spectral Distortion and 21-cm global signal

Primordial black holes (PBHs) can form spike density halos through the accretion of weakly interacting massive particles (WIMPs). In these halos, the enhanced density significantly boosts the annihilation rate of WIMPs. For Majorana dark matter annihilation into light fermions, the s-wave part of the annihilation cross section is helicity-suppressed, making the p-wave contribution dominant. We study the velocity-dependent p-wave annihilation case, whose resulting energy injection can modify the thermal and ionization history of the Universe, leaving observable imprints on the cosmic microwave background (CMB) spectrum and the global 21-cm signal. From the predicted energy injection into the plasma, we derive stringent upper limits on the fraction of dark matter in form of PBHs for p-wave annihilation models, based on the observational constraints of the CMB spectral distortions (y-type), and from the measurement of the 21-cm absorption signal at cosmic dawn. Our results highlight that accounting for the p-wave nature of annihilation is crucial for
deriving robust constraints on the PBH abundance.

Speaker: Shibsankar Si (NIT Meghalaya)

12:50 → 13:50 Lunch

13:50 → 14:25 SKA software development

Convener: Yogesh Wadadekar (NCRA)

14:25 → 15:25 Session 7

Convener: Chair: Sukhdeep Singh (IIT Kharagpur)

14:25 CosmoUiT: A Vision Transformer-UNet Hybrid for Fast and Accurate Emulation of 21-cm Maps from the Epoch of Reionization

The observation of the redshifted 21-cm signal from the intergalactic medium will probe the epoch of reionization (EoR) with unprecedented detail. Various simulations are being developed and used to predict and understand the nature and morphology of this signal. However, these simulations are computationally very expensive and time-consuming to produce in large numbers. To overcome this problem, an efficient field-level emulator of this signal is required. However, the EoR 21-cm signal is highly non-Gaussian, making it challenging for neural networks to accurately capture the multi-scale correlations of the field. Here we introduce CosmoUiT, a UNet integrated vision transformer-based architecture, to overcome these difficulties. CosmoUiT emulates the 3D cubes of 21-cm signal from the EoR, for a given input dark matter density field, halo density field, and reionization parameters. CosmoUiT uses the multi-head self-attention mechanism of the transformer to capture the long-range dependencies and convolutional layers in the UNet to capture the small-scale variations in the target 21-cm field. Furthermore, the training of the emulator is conditioned on the input reionization parameters such that it gives a fast and accurate prediction of the 21-cm field for different sets of input reionization parameters. We evaluate the predictions of our emulator by comparing various statistics (e.g., bubble size distribution, power spectra) and morphological features of the emulated and simulated maps. We further demonstrate that this vision transformer-based architecture can emulate the entire 3D 21-cm signal cube with high accuracy at both large and small scales.

Speaker: Yashrajsinh Mahida (IIT Indore)

14:45 Galaxy-Based Tomographic Reconstruction of Reionization: A Foreground-free Complement to SKA-Low 21-cm Observations

Mapping the three-dimensional topology of reionization is one of the defining science goals of the SKA era. While SKA-Low will deliver 21-cm brightness-temperature maps of the neutral intergalactic medium (IGM), foreground contamination and calibration systematics remain central challenges for field-level inference. I will present TORRCH (TOmographic Reconstruction of the Reionization of Cosmic Hydrogen), a deep-learning framework that reconstructs the neutral-hydrogen fraction field during the Epoch of Reionization (EoR) directly from the spatial distributions of Lyman-alpha emitters (LAEs) and non-Lyman-alpha-selected galaxies (NLSGs), providing a foreground-free, galaxy-anchored view of ionization morphology that is complementary to 21-cm observations.
TORRCH employs a deterministic 3D U-Net trained on hydrodynamical simulations post-processed with radiative transfer, spanning diverse reionization scenarios and realistic survey depths comparable to current and forthcoming JWST and Subaru programmes. The framework recovers large-scale ionization topology, the one-point distribution of neutral fractions, the projected power spectrum, and the galaxy-IGM cross-correlation, with good fidelity on scales above ~7 h^-1 cMpc, and remains robust to realistic redshift uncertainties and ionization conditions not seen during training. With JWST delivering ever-deeper galaxy samples in the EoR, such galaxy-based tomography offers an independent, foreground-clean route to field-level reionization constraints and a natural synergistic counterpart to SKA-Low science.

Speaker: Soumak Maitra (TIFR)

15:05 The 21cm-Galaxy Cross-Bispectrum: A Forecast for Synergies between SKA-Mid and Galaxy Surveys

The Square Kilometer Array (SKA-Mid) will map the post-reionization (post-EoR) Universe, via 21cm line intensity mapping (LIM) with accuracy and detail that was never possible. The SKA-Mid will first make the statistical detection of the 21cm LIM via power spectrum measurements, and the inference of astrophysics of the galaxies and Cosmology will follow with these detections. The power spectrum, by definition, can only completely characterize the statistical properties of a signal, which is a Gaussian random field. However, due to the non-linear clustering of the matter, the redshifted 21cm LIM signal from the post-EoR will be highly non-Gaussian. Additionally, the signal will contain the imprints of the primordial non-Gaussianity. Thus, to quantify the non-Gaussianity in the 21cm LIM signal, one has to consider higher-order statistics such as the bispectrum. However, residual foreground and systematics associated with the instrument will hinder a high signal-to-noise ratio detection of the 21cm auto bispectrum with fewer observational hours. One way to boost the signal-to-noise ratio is to do a cross-correlation measurement with various galaxy redshift surveys. Additionally, these cross-correlations have the potential to provide improved constraints on the Cosmological parameters. In this talk, I will discuss the prospects of the 21cm-galaxy cross-bispectrum, primarily focusing on forecasts with SKA-Mid and Euclid. We will show, for the first time, the nature of the 21cm-galaxy cross-bispectrum for all unique shapes and sizes of the bispectrum for various 21cm galaxy cross-combinations. Additionally, we will discuss the forecast on the detectability of the 21cm-galaxy cross bispectrum for a set of theoretical models of the post-EoR 21cm LIM signal with state-of-the-art simulations. These analyses highlight the potential of combining 21cm observations with complementary datasets of the large-scale structure in this era of precision Cosmology.

Speaker: Leon Noble (IIT Indore)

15:25 → 15:55 Tea/Coffee

15:55 → 16:30 Results from the MIGHTEE survey

Convener: Sourabh Paul (University of Manchester)

16:30 → 17:10 Session 8

Convener: Chair: Soumak Maitra (TIFR)

16:30 ANN-Based Emulators for Efficient Inference of Reionization Parameters

Constraining the Epoch of Reionization with physically motivated simulations is hampered by the high cost of conventional parameter inference. We present an efficient emulator-based framework that dramatically reduces this bottleneck for the photon-conserving semi-numerical code SCRIPT. Our approach combines (i) a reliable coarse resolution MCMC to locate the high-likelihood region (exploiting the large-scale convergence of SCRIPT) with (ii) an adaptive, targeted sampling strategy to build a compact high resolution training set for an artificial neural network based emulator of the model likelihood. With only ≈ 10^3 high-resolution simulations, the trained emulators achieve excellent predictive accuracy (R^2 ≈ 0.97-0.99) and, when embedded within an MCMC framework, reproduce posterior distributions from full high-resolution runs. Compared to conventional MCMC, our pipeline reduces the number of expensive simulations by a factor of ~100 and lowers total CPU cost by up to a factor of ~70, while retaining statistical fidelity. This computational speedup makes inference in much higher-dimensional models tractable (e.g., those needed to incorporate JWST and upcoming 21 cm datasets) and provides a general strategy for building efficient emulators for next generation of EoR constraints.

Speaker: Saptarshi Sarkar (NCRA-TIFR, Pune)

16:50 Probing Reionization upto the 'mark': The Marked Power spectrum to unviel the HI 21-cm signal from the Epoch of Reionization

The 21-cm signal arising from the spin-flip transition from neutral hydrogen (HI) in the intergalactic medium (IGM) serves as a powerful probe of the Epoch of Reionization (EoR), the period during which the first luminous sources were formed and Universe went through a phase transition from being neutral to ionized. This signal traces the distribution and properties of the IGM in this critical phase change of the cosmic history. More importantly this field exhibits a significantly strong non-Gaussianity due to the complex distribution and growth of ionized regions, which is tied to the various astrophysical processes in the IGM and within the sources. As a result, the standard 21-cm power spectrum fails to capture all the relevant statistical information hidden in this field. While higher-order statistics, such as the bispectrum, can provide additional insights, they are computationally intensive and often difficult to interpret. This work investigates the application of marked statistics as an alternative approach to characterize the EoR 21-cm field. By employing non-linear transformations and assigning weights based on the local property in the map, the contributions from distinct types of fluctuations in that environment can be modified, allowing for targeted analysis of specific features. In particular, up-weighting specific field properties amplifies their corresponding fluctuations, improving the ability of the marked power spectrum to capture the relevant information content. Through this project, we demonstrated that it is possible to construct meaningful transformations that render the marked power spectrum a more optimal summary statistic. We show that this approach provides a computationally efficient framework for extracting non-Gaussian information and enables a more detailed statistical description of the 21-cm signal. The project further explores different forms of mark functions and their impact on the resulting statistics, with the broader goal of improving our understanding of the EoR.

Speaker: Vishrut Pandya (IIT Indore)

Wednesday, 11 March

08:30 → 09:30 Breakfast

09:30 → 10:05 Unveiling the Epoch of Reionization: Insights from JWST

Convener: Anirban Chakraborty (NCRA-TIFR, Pune)

10:05 → 11:05 Session 9

Convener: Chair: Girish Kulkarni (TIFR)

10:05 Probing the End of Reionization with 21cm-[CII] Cross-Correlation

The Epoch of Reionization (EoR) marks the last major phase transition of the Universe, during which the Intergalactic Medium (IGM) evolves from being predominantly neutral to fully ionized. The line intensity mapping (LIM) of different spectral lines originating from this era are direct and most promising probes for the EoR. The two spectral lines that stand out for the LIM surveys of the EoR are the redshifted 21-cm line from neutral hydrogen (HI) and singly ionized carbon (CII) line from star forming regions of the early galaxies. The SKA-Low and FYST are the two upcoming instruments that will be capable of performing LIM surveys of the EoR using 21-cm and CII lines respectively. During the EoR, the 21-cm signal predominantly comes from the HI in the IGM, whereas the contribution to this signal from the HI in the ISM is negligible. This changes by the end of reionization when almost all of the IGM is ionized, then the 21-cm signal originates mostly from the galaxies. We propose that by probing the nature of correlation between the 21-cm and CII LIM signals using the surveys to be performed by SKA-Low and FYST one would able to constrain the end of reionization redshift in a robust manner, which may not be possible via other probes of the EoR e.g. quasar absorption spectra, CMBR Thomson scattering optical depth etc. We specifically focus on the cross-power spectrum statistics to perform this analysis and also make forecasts for future SKA-Low and FYST surveys.

Speaker: Shiriny Akthar (IIT Indore)

10:25 Probing Modified Gravity with SKA: Redshift-Space 21-cm Bispectrum Multipoles

The upcoming Square Kilometre Array (SKA) will map the large-scale structure of the Universe with unprecedented sensitivity via 21-cm intensity mapping. While standard analyses focus on the power spectrum, non-linear gravitational evolution encodes rich information in higher-order statistics. In this work, we investigate the redshift-space 21-cm bispectrum and its spherical harmonic multipoles as a probe of gravity in the post-reionization epoch (z∼0.3−3). Using tree-level perturbation theory, we model the bispectrum for both ΛCDM and the Hu-Sawicki f(R) modified gravity model, incorporating local Eulerian bias and redshift-space distortions. We forecast the detectability of modified gravity signatures for an SKA-MID-like survey, accounting for thermal noise and cosmic variance. We find that the signal-to-noise ratio is maximized by the lowest order multipoles (ℓ=0,2) in squeezed and stretched triangle configurations. Our results demonstrate that the 21-cm bispectrum on quasi-linear scales (k∼0.4−0.8hMpc−1) offers a potent, complementary test of General Relativity, capable of constraining scale-dependent growth with high statistical significance in the SKA era. The talk is based on the arXiv:2601.18862

Speaker: Sourav Pal (ISI Kolkata)

10:45 Minkowski Functionals of the 21 cm Signal as a Probe of Primordial Features

The redshifted 21 cm signal from the cosmic dawn and Epoch of Reionization (EoR) encodes important information about both astrophysical processes and primordial physics, such as inflation. In this work, we use morphological statistics to explore the sensitivity of the 21 cm signal to inflationary features and EoR dynamics simultaneously. Focusing on primordial features from particle production during inflation we generate semi-numerical simulations of the 21 cm signal across redshifts 5 < z < 35, incorporating these features. Using Minkowski Functionals (MFs), we analyze the morphology of 21 cm fields: density, neutral hydrogen fraction, spin temperature, and brightness temperature. We demonstrate that MFs are highly sensitive to both the amplitude and scale of primordial features, capturing rich morphological information. In particular, we show that MFs can robustly identify inflationary features and distinguish them from the standard model. We further explore various EoR scenarios, and demonstrate that combining MFs across redshifts can disentangle the signatures of primordial features from EoR effects. This approach opens new avenues for probing inflation with upcoming 21 cm surveys.

Speaker: Kanan Vijay Virkar (IIA)

11:05 → 11:35 Tea/Coffee

11:35 → 12:10 Detection of the Cosmological 21 cm Auto Power Spectrum with CHIME at z > 1

Convener: Arnab Chakraborty (McGill University)

12:10 → 12:50 Session 10

Convener: Chair: Anirban Chakraborty (NCRA)

12:10 21cm Cosmology in the Regime of Scalar Field Theory

We aim to investigate the influence on the 21cm brightness temperature for a non LCDM cosmology.The temperature dip,free electron fraction and the 21cm power spectrum is observed to change in the presence of a scalar field regime which shall be discussed throughout the talk.Alternatively a comparison with the JWST observation will also be made to investigate the contribution of the scalar field.

Speaker: Sohail Shaikh (Jamia Millia Islamia)

12:30 Exploring the signature of enhanced star formation efficiency on the 21 cm signal with SCRIPT

Recent observations with the James Webb Space Telescope (JWST) indicate a higher-than-expected abundance of galaxies at high redshifts (z > 10), potentially suggesting galaxy formation models with an enhanced star formation efficiency during the Cosmic Dawn (CD). These galaxy formation models are expected to leave their imprints on the thermal and ionization history of the intergalactic medium (IGM) at high redshifts and the redshifted 21 cm signal from neutral hydrogen from the CD. In this work, we investigate these imprint using an upgraded version of the semi-numerical framework SCRIPT, which now self-consistently models spatial fluctuations in the hydrogen spin temperature by including the Lyman-α coupling and X-ray heating. With this framework, we explore models with and without enhanced star formation rate and and their impact on the shape of the 21 cm signal. We further assess the extent to which the combined measurements of the sky-averaged 21 cm signal and its power spectra can discriminate between various scenarios. With upcoming observations with the SKA1-LOW and other global and interferometric observations, the 21 cm signal can potentially be an independent probe to study the galaxy formation models at high redshifts.

Speaker: Janakee Raste (NCRA-TIFR, Pune)

12:50 → 13:50 Lunch

13:50 → 14:25 Indian Participation in SKA Data Challenge 3a

Convener: Samir Choudhuri (IIT Madras)

14:25 → 15:00 Indian Participation in SKA Data Challenge 3b

Convener: Suman Majumdar (IIT Indore)

15:00 → 15:50 Tea/Coffee

15:50 → 16:25 SKA science verification

Convener: Samir Choudhuri (IIT Madras)

16:25 → 17:00 Discussion