🔭 astrophysics

The Tracking Tapered Gridded Estimator for the 21-cm power spectrum from the Murchison Widefield Array (MWA) drift scan observations -- III. Improved upper limits at $z = 8.2$ from multiple pointings

By applying a tracking tapered gridded estimator to MWA drift scan observations, this study achieves the tightest 21-cm power spectrum upper limit from the MWA to date, though it remains significantly higher than limits from other telescopes and theoretical predictions.

Original authors: Shouvik Sarkar, Khandakar Md Asif Elahi, Samir Choudhuri, Somnath Bharadwaj, Suman Chatterjee, Baijayanta Bhattacharyya, Shiv Sethi, Akash Kumar Patwa

Published 2026-04-28

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Shouvik Sarkar, Khandakar Md Asif Elahi, Samir Choudhuri, Somnath Bharadwaj, Suman Chatterjee, Baijayanta Bhattacharyya, Shiv Sethi, Akash Kumar Patwa

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 Cosmic Whisper: Hunting for the Universe’s First Light

Imagine you are standing in a massive, crowded stadium during a championship game. The roar of the crowd is deafening—thousands of people cheering, shouting, and whistling. Now, imagine that amidst this thunderous noise, someone in the very back row is trying to whisper a secret to a friend.

Your job is to hear that whisper. But there’s a catch: you can’t just plug your ears. You have to use math and technology to figure out exactly which sounds are the "crowd" (the noise) and which tiny, specific sound is the "whisper" (the signal).

This is exactly what the scientists in this paper are doing, but instead of a stadium, they are looking at the Universe.

1. The "Whisper": The 21-cm Signal

The "whisper" these scientists are hunting for is a very faint radio signal from the Epoch of Reionization (EoR). This was a pivotal moment in cosmic history when the very first stars and galaxies turned on, lighting up the dark, foggy universe.

We look for this using a specific "note" played by hydrogen gas, known as the 21-cm line. If we can hear this note, we can map out how the first stars changed the universe.

2. The "Crowd": The Foreground Problem

The problem is that the "crowd" is incredibly loud. Our own Milky Way galaxy and other bright, distant objects (like a massive radio source called Fornax A) are screaming at us with radio noise that is 100,000 times louder than the signal we want to hear.

It’s like trying to hear a mosquito buzz in the middle of a heavy metal concert.

3. The "Noise-Canceling Headphones": TTGE and SCF

To solve this, the researchers used a sophisticated mathematical toolkit. Think of it like high-tech, intelligent noise-canceling headphones:

The TGE (The Filter): This is a method that helps organize the messy radio data into a clean grid, making it easier to process.
The SCF (The Smoothness Filter): Most of the "crowd noise" (the galaxies) is "smooth"—it sounds like a constant, low hum. The 21-cm signal, however, is "jittery" and changes quickly. The scientists used a technique called Smooth Component Filtering to "subtract" the smooth hum, leaving behind only the jittery bits where the signal might be hiding.

4. The "Search Party": Multiple Pointings

Instead of looking through just one telescope lens, the team used the Murchison Widefield Array (MWA) to scan the sky in 163 different spots (called "pointings").

They realized that some spots in the sky were "louder" than others. For example, when the bright source Fornax A passed through their view, it was like a giant speaker being wheeled right in front of them—it made it impossible to hear anything.

By scanning many different areas, they were able to find the "quietest" corners of the sky—the best places to listen for the cosmic whisper.

5. The Result: A New Record

The scientists didn't actually "hear" the whisper yet (the signal is still too faint for current technology), but they did something incredibly important: they set a new "Upper Limit."

In science, an "upper limit" is like saying, "I couldn't hear the whisper, but I can tell you for certain that it isn't louder than THIS."

By combining the data from 23 of their quietest sky spots, they created the tightest (most precise) limit ever recorded by the MWA telescope. They have effectively lowered the "volume" of the background noise in their models, making it much easier for the next generation of telescopes (like the upcoming SKA-Low) to finally hear that first cosmic light.

Summary in a Nutshell

The researchers used clever math to "mute" the screaming galaxies of the universe, scanned 163 different parts of the sky to find the quietest spots, and proved that the "first light" of the universe is even fainter than we previously thought. They have cleared the path for future astronomers to finally hear the story of how the universe began to glow.

This paper, titled "The Tracking Tapered Gridded Estimator for the 21-cm power spectrum from the Murchison Widefield Array (MWA) drift scan observations – III," presents a comprehensive analysis of MWA drift scan data to constrain the 21-cm power spectrum (PS) from the Epoch of Reionization (EoR) at redshift $z=8.2$ .

1. The Problem

The primary challenge in detecting the 21-cm signal from the EoR is that the cosmological signal (expected to be $\sim$ tens of mK) is buried under astrophysical foregrounds (Galactic synchrotron and extra-galactic point sources) that are 4 to 5 orders of magnitude stronger. Additionally, instrumental effects, such as radio frequency interference (RFI) and "missing frequency channels" (flagged data), introduce spectral artifacts—specifically periodic spikes in the delay-space power spectrum—that can mimic or contaminate the cosmological signal.

2. Methodology

The authors employ an advanced statistical framework to mitigate these issues:

Tracking Tapered Gridded Estimator (TTGE): A visibility-based estimator that uses a Gaussian window function to suppress side-lobe contributions from bright sources and correlates gridded visibilities for computational efficiency and noise-bias subtraction.
Multi-frequency Angular Power Spectrum (MAPS): To handle missing frequency channels, the authors estimate the MAPS first and then Fourier transform it along the frequency separation ( $\Delta\nu$ ). This ensures the power spectrum remains free from artifacts caused by frequency flagging.
Smooth Component Filtering (SCF): To combat the "periodic spikes" caused by the interaction of foregrounds and missing channels, the authors apply a Hann window to filter out the spectrally smooth foreground components from the gridded visibilities before estimating the PS.
Data Selection & Masking: They analyze 163 Pointing Centers (PCs) across an 81° span. They identify a "clean" region in the $(k_\perp, k_\parallel)$ plane, specifically $0.007 \leq k_\perp \leq 0.045 \text{ Mpc}^{-1}$ and $0.135 \leq k_\parallel \leq 1.399 \text{ Mpc}^{-1}$ , and apply a mask to remove residual periodic spikes.
Incoherent Combination: To improve sensitivity, they combine the results from multiple "good" (noise-dominated) PCs using inverse-variance weighting.

3. Key Contributions

Large-scale Sky Survey: A systematic measurement of the 2D angular power spectrum ( $C_\ell$ ) and 3D power spectrum ( $P(k)$ ) across 163 different pointing centers.
Foreground Mapping: Identification of specific sky regions (e.g., near Fornax A) that are unsuitable for EoR science due to extreme contamination.
Methodological Refinement: Demonstration that the combination of TTGE, MAPS, and SCF effectively mitigates foreground leakage and spectral artifacts.

4. Results

Angular Fluctuations ( $D_\ell$ ): The study finds that $D_\ell \propto \ell^2$ for $\ell \geq 200$ in most regions, indicating dominance by Poisson fluctuations from point sources. A massive peak in $D_\ell$ occurs at $\alpha \approx 50^\circ$ due to the bright source Fornax A entering the primary beam.
Best Performing Field: The PC at $\alpha = 11.0^\circ$ was identified as one of the cleanest, yielding an individual $2\sigma$ upper limit of $\Delta^2_{UL}(k) = (173.13)^2 \text{ mK}^2$ at $k = 0.161 \text{ Mpc}^{-1}$ .
Tightest Upper Limit: By incoherently combining 23 optimal PCs (Case I), the authors achieved a $2\sigma$ upper limit of $\Delta^2_{UL}(k) = (98.67)^2 \text{ mK}^2$ at $k = 0.156 \text{ Mpc}^{-1}$ .
Comparison: This is the tightest upper limit from the MWA to date, being $\approx 3$ times lower than previous MWA limits and $\approx 15$ times lower than specific previous results at similar scales. However, it remains $\approx 21$ times higher than the limits set by HERA.

5. Significance

The work provides a vital roadmap for future EoR observations. By mapping the foreground contamination across a wide area, it identifies the most "foreground-free" windows (specifically $358.5^\circ \leq \alpha \leq 11.5^\circ$ ) for deep observations. Furthermore, since the MWA is located at the same site as the upcoming SKA-Low, these findings serve as a direct guide for selecting target fields to maximize the chances of a successful cosmological detection with the Square Kilometre Array.