Mitigating oscillatory systematics in MeerKLASS L-band H i intensity maps

arXiv:2611.13033 · astro-ph.IM, astro-ph.CO · Submitted 2026-11-09 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics. The day in astrophysics: observations, models and what the data forces.

Vera: We're looking at the paper: "Mitigating oscillatory systematics in MeerKLASS L-band H i intensity maps".

Jocelyn: Here's the gist of the paper. H i intensity mapping is a promising probe of the large-scale structure of the Universe, but is limited by foregrounds and instrumental systematics The chromaticity and sidelobes of the MeerKAT beam modulate the frequency-smooth foreground emission and introduce oscillating structures that persist after foreground removal In MeerKLASS L-band data, processing these two types of scans separately significantly suppresses the oscillating residuals while preserving the stacked H i signal, increasing the ratio of the H i signal amplitude to the systematic fluctuation amplitude by 81% The strategy helps mitigate spectral ripples induced by the beam in future MeerKAT and SKA-Mid surveys.

Vera: Origin of Oscillations The authors find that the oscillation originates from frequency-smooth foreground emission modulated by the chromatic MeerKAT beam This oscillation can be seen in the correlation matrix even after removing 10 PCA modes in the analysis in Chen et al.

Jocelyn: (2025), and in this work we find a similar structure persists in mode 11.

Vera: The authors simulate the sky with a full-sky electromagnetic model of the MeerKAT beam, including complex frequency dependence and far sidelobes The far sidelobes introduce additional spectral structures into the frequency spectra.

Paper summary: Vera: So, to recap this paper, "Mitigating oscillatory systematics in MeerKLASS L-band H i intensity maps," the authors are addressing how instrumental beam effects create oscillations in frequency-smooth foreground emission that stick around even after standard foreground removal techniques.

Jocelyn: The main claim is that the chromaticity and sidelobes of the MeerKAT beam modulate this background emission, introducing these persistent structures in their data.

Subrahmanyan: This matters because these residuals, which they see in mode eleven and in a correlation matrix ripple pattern after removing ten PCA modes, are not just noise; they are systematic artifacts that can bias the measurement of the large-scale structure of matter <ref:2611.13033#pg1>.

Vera: The authors show that processing two types of scans separately and then combining them before foreground cleaning significantly suppresses these oscillating residuals while keeping the stacked HI signal intact.

Jocelyn: They also found that this strategy boosts the ratio between their actual HI signal amplitude and the systematic fluctuation amplitude by eighty-one percent, which is a big improvement in how clean their final spectrum is.

Subrahmanyan: This suggests that accounting for the observing strategy and beam orientation can improve foreground cleaning without needing more aggressive PCA subtraction, which addresses a real issue in large-scale structure surveys.

Conclusion: Vera: Looking at the title, "Mitigating oscillatory systematics in MeerKLASS L-band H i intensity maps," it clearly points to fixing a specific kind of noise that pops up when you map neutral hydrogen intensity using these instruments and surveys.

Jocelyn: The authors, including Boyan Zhao and Eloy de Lera Acedo, are essentially telling us that the way we observe the sky matters just as much as what we observe.

Subrahmanyan: They are showing that by simply changing how you combine your data based on the beam orientation—like rising versus setting scans—you can get a much cleaner measurement of how matter is distributed in the Universe.

Vera: It means for future surveys, whether it's MeerKAT or SKA-Mid, we shouldn't just focus on cleaning the data with PCA modes; we need to be mindful of the observing strategy and beam orientation itself.

Jocelyn: The implication is that accounting for these beam effects could provide an additional way to improve foreground cleaning, perhaps suggesting things like multi-scale PCA or GMCA as a potential next step.

Subrahmanyan: It’s about finding ways to mitigate these systematic residuals so we can get a more accurate picture of the large-scale distribution of matter in the Universe, which is what this work is fundamentally about.

Boyan Zhao, Eloy de Lera Acedo, Zhaoting Chen, John Cumner, Vijay Mahatma, Marta Spinelli, Dirk I. L. de Villiers, § Camera12Acedo Isabella Paola Carucci Xuelei Chen Steven Cunnington Furen Deng Brandon N Engelbrecht José Fonseca Keith Grainge Wenkai Hu Melis O Irfan Yichao Li Mario G Santos Jingying Wang Yougang Wang Laura Wolz

State Key Laboratory of Radio Astronomy and Technology, National Astronomical Observatories, Chinese Academy of Sciences · Cavendish Laboratory, University of Cambridge · School of Astronomy and Space Science, University of Chinese Academy of Sciences · Kavli Institute for Cosmology in Cambridge, University of Cambridge · Institute for Astronomy, The University of Edinburgh · Observatoire de la Côte d’Azur Laboratoire Lagrange · Department of Physics & Astronomy, University of the Western Cape · Department of Electrical and Electronic Engineering, Stellenbosch University · Dipartimento di Fisica, Universita degli Studi di Milano INFN – Istituto Nazionale di Fisica Nucleare Sezione di Milano INAF – Istituto Nazionale di Astrofisica Osservatorio Astrofisico di Brera-Merate INAF – Istituto Nazionale di Astrofisica Osservatorio Astronomico di Trieste IFPU Institute for Fundamental Physics of the Universe Key Laboratory of Cosmology and Astrophysics Liaoning College of Sciences Institute of Cosmology & Gravitation University of Portsmouth Instituto de Astrofísica e Ciçes do Espaía, Universidade do Porto Departamento de Fnica e Astronomia Faculdade de Ci➚ncia Universidade do Porto

astro-ph.IM, astro-ph.CO

Submitted: 2026-11-09

Updated: 2026-11-09

Comments: 16 pages, 16 figures, submitted to MNRAS

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 79/100

The gist: H i intensity mapping is a promising probe of the large-scale structure of the Universe, but is limited by foregrounds and instrumental systematics The chromaticity and sidelobes of the MeerKAT beam…

Key concepts

Beam Chromaticity and Sidelobes
The MeerKAT telescope's beam has properties that change depending on the frequency being observed (chromaticity) and produces unwanted signals outside the main beam (sidelobes). These features modulate the smooth foreground emission, creating oscillating structures in the data that are hard to remove with standard methods.
Oscillating Residuals
These are specific, repeating patterns or ripples left in the data after attempting to clean out foreground emission. The authors found these oscillations were caused by the beam's effects and persisted even after removing 10 PCA modes, indicating a systematic error that standard cleaning methods miss.
Data Splitting for Validation
Splitting the large dataset into subsets based on observing strategy (e.g., 'super-cross') and cleaning them individually confirmed the effectiveness of the strategy. This splitting showed that combining data from different scanning strategies enhances these frequency-correlated systematics, proving that accounting for beam orientation is crucial.
Signal-to-Systematic Ratio Improvement
The core finding is that by combining datasets observed with different beam orientations after PCA cleaning, the ratio of the actual HI signal amplitude to the systematic fluctuation amplitude increased by 81%. This means the true astronomical signal became much clearer and less contaminated by instrumental noise.

Summary

H i intensity mapping is a promising probe of the large-scale structure of the Universe, but is limited by foregrounds and instrumental systematics The chromaticity and sidelobes of the MeerKAT beam modulate the frequency-smooth foreground emission and introduce oscillating structures that persist after foreground removal In MeerKLASS L-band data, processing these two types of scans separately significantly suppresses the oscillating residuals while preserving the stacked H i signal, increasing the ratio of the H i signal amplitude to the systematic fluctuation amplitude by 81% The strategy helps mitigate spectral ripples induced by the beam in future MeerKAT and SKA-Mid surveys. p. 0

Where the paper says it

In MeerKLASS L-band data, processing these two types of scans separately significantly suppresses the oscillating residuals while preserving the stacked H i signal. The ratio of the H i signal amplitude to the systematic fluctuation amplitude increases by 81%.

Origin of Oscillations

The authors find that the oscillation originates from frequency-smooth foreground emission modulated by the chromatic MeerKAT beam This oscillation can be seen in the correlation matrix even after removing 10 PCA modes in the analysis in Chen et al. (2025), and in this work we find a similar structure persists in mode 11. p. 0

Where the paper says it

The ratio of the H i signal amplitude to the systematic fluctuation amplitude increases by 81%. We find an oscillating structure in PCA mode 11 in our previous data processing, but simply removing this mode does not improve the stacked spectrum due to the signal loss.

The authors simulate the sky with a full-sky electromagnetic model of the MeerKAT beam, including complex frequency dependence and far sidelobes The far sidelobes introduce additional spectral structures into the frequency spectra. p. 7
Where the paper says it

First, the far sidelobes introduce additional spectral structures into the frequency spectra. Second, what PCA captures is the foreground emission modulated by the chromatic beam response.

Systematic Mitigation Strategy

The authors propose a strategy to mitigate the systematics by combining datasets observed with different beam orientations before PCA cleaning This combination produces a more complex effective beam and strengthens the residual frequency correlations, suggesting that more frequency-correlated systematics remain when datasets are combined. The authors also test two approaches for combining data after PCA cleaning, finding that applying PCA cleaning to the rising and setting cubes separately and then combining them yields a ratio of H i signal amplitude to systematic amplitude of approximately 81% compared to the original analysis. p. 9

Where the paper says it

We construct the rising and setting cubes by separately combining the data from all rising and setting scans, respectively. PCA foreground cleaning is then applied independently to the two cubes.

Role of Spatial Information

The stacking procedure enhances the visibility of these spatially correlated structures because it integrates over pixels around the positions of stacked galaxies, where neighboring pixels have relatively similar oscillatory features The authors test this by removing the summation over neighboring pixels and using only the central pixel at each galaxy position, which substantially reduces the ripple pattern in the correlation matrix. This spatial information is not explicitly exploited by standard PCA, suggesting that methods like multi-scale PCA or GMCA may provide additional routes for improving foreground cleaning.

Implications for Future Surveys

The results demonstrate that accounting for the observing strategy and beam orientation can improve foreground cleaning without requiring more aggressive PCA subtraction The findings suggest that future H i intensity-mapping analyses with MeerKAT and SKA-Mid should account not only for beam chromaticity and sidelobes but also for changes in beam orientation during the observations. The combination of observations after PCA cleaning substantially reduces the oscillating foreground residuals while only 9 PCA modes are removed, achieving both goals simultaneously. p. 9

Where the paper says it

Although using only one type of scan can effectively mitigate the oscillating foreground residuals, only half of the observation time is used. To make full use of the observations, we still need to combine the rising and setting scans, but after PCA cleaning.

Conclusion and Trade-off Alleviation

The study demonstrates that the systematics are an additive foreground residual component of the data vector, which does not correlate with optical galaxy samples The authors show that combining the observations after PCA cleaning enhances the frequency-correlated systematics, and this strategy alleviates the traditional trade-off in PCA foreground cleaning between foreground removal and signal preservation. The systematic component is reduced so that the ratio increases to 2A0/A5k ≈ 2. p. 0

Where the paper says it

In MeerKLASS L-band data, processing these two types of scans separately significantly suppresses the oscillating residuals while preserving the stacked H i signal. The ratio of the H i signal amplitude to the systematic fluctuation amplitude increases by 81%.

p. 0
Where the paper says it

In MeerKLASS L-band data, processing these two types of scans separately significantly suppresses the oscillating residuals while preserving the stacked H i signal. The ratio of the H i signal amplitude to the systematic fluctuation amplitude increases by 81%.

0 in both cases, an increase of 81% compared to the original result. p. 9
Where the paper says it

In the right half of Figure 14, the stacked spectra of the rising and setting cubes are presented. In the left half, we present the original result from Chen et al.

This demonstrates that accounting for the observing strategy and beam orientation can improve foreground cleaning without requiring more aggressive PCA subtraction. p. 9
Where the paper says it

The trough near |𝑣| = 2500 km s−1 is partly caused by the mean subtraction in the PCA cleaning, which forces the cleaned spectrum to have approximately zero mean. Therefore, the positive excess around 𝑣 = 0 km s−1 must be accompanied by a negative excess elsewhere, making the trough unsuitable for independently measuring the foreground residuals.

The combination of datasets observed with different beam orientations makes a fixed number of PCA modes less effective, leaving stronger correlated residuals. p. 8
Where the paper says it

Consequently, a fixed number of PCA modes may no longer capture these systematics as efficiently as when each dataset is cleaned separately. The remaining frequency-correlated residuals then give rise to the stronger ripple pattern seen in the stacked spectrum and its estimated correlation matrix.

Data Splitting for Validation

The authors show that splitting the datasets to form the “super-cross” and cleaning the subsets separately allows measurements to be free of significant systematics as shown in the comprehensive validation tests therein. This approach confirms that combining data from different scanning strategies enhances the frequency-correlated systematics, and that splitting the datasets to form the “super-cross” leads to measurements free of significant systematics as shown in Figure 2 of MeerKLASS Collaboration et al. p. 1

Where the paper says it

One of the major challenges in H i intensity mapping is the removal of foreground contamination, which is dominated by Galactic synchrotron and free–free emission, together with their extragalactic counterparts. The foreground remains spectrally smooth and is predominantly described by a power law, but its amplitude varies at the Kelvin level across different lines of sight, as shown in MeerKLASS Collaboration et al.

p. 1
Where the paper says it

This excess, which is later mitigated by performing internal cross-correlation between different H i datasets (MeerKLASS Collaboration et al. 2026), indicates that residual systematics may still be present in the data after foreground cleaning.

, 2026. The systematic component is reduced so that the ratio increases to 2A0/A5k ≈ 2.0 in both cases, an increase of 81% compared to the original analysis. p. 1
Where the paper says it

In this work, we focus on the latter. The single-dish observations retain large-scale modes inaccessible to the interferometric measurements, while the long integration time of approximately 62 hours per dish provides improved sensitivity compared with previous MeerKLASS singledish observations (MeerKLASS Collaboration et al.

The oscillation is significantly stronger in the final cube than in either the rising or setting cubes, indicating that combining the two scanning strategies enhances the frequency-correlated systematics. p. 9
Where the paper says it

72, respectively. These results indicate that combining data from the rising and setting scans significantly enhances the foreground residual.

p. 9
Where the paper says it

72, respectively. These results indicate that combining data from the rising and setting scans significantly enhances the foreground residual.

p. 9
Where the paper says it

72, respectively. These results indicate that combining data from the rising and setting scans significantly enhances the foreground residual.

Beam Effects and Chromaticity

The simulation results show that a wide range of sidelobes at theta ≲ 20◦, 40◦, 90◦ can introduce additional oscillatory spectral structures into the foreground emission These beam-induced structures can dominate some of the leading PCA modes and therefore affect foreground cleaning. The oscillation is significantly stronger in the final cube than in either the rising or setting cubes, indicating that combining the two scanning strategies enhances the frequency-correlated systematics. The authors find that some of the modes show surprising similarities to those from the real data, such as mode 3 having a bowl-shaped profile. p. 8

Where the paper says it

It is noteworthy that some of the modes show surprising similarities to those from the real data. For example, in Figure 8, mode 3 has a bowl-shaped profile, and mode 6 resembles its counterpart in the simulation with 𝜃max = 90◦ , except for a sharp trough near 986 MHz.

The sinusoidal mode 11 in the real data resembles mode 5 for thetamax ≥ 40◦ and mode 4 for thetamax = 20◦ in Figure 10. p. 8
Where the paper says it

For example, in Figure 8, mode 3 has a bowl-shaped profile, and mode 6 resembles its counterpart in the simulation with 𝜃max = 90◦ , except for a sharp trough near 986 MHz. The sinusoidal mode 11 in the real data, which contributes strongly to the ripple pattern, resembles mode 5 for 𝜃max ≥ 40◦ and mode 4 for 𝜃max = 20◦ in Figure 10.

Beam Rotation Impact

The authors model the second scenario where beam orientations remain close to one value during rising scans and another during setting scans, leading to a more complex effective beam This suggests that combining scans with different beam orientations makes the effective beam more complex and consequently makes PCA less effective. The ripple patterns are substantially stronger when beam rotation is included than when no rotation is applied, suggesting that more frequency-correlated systematics remain in the cleaned data when datasets observed with different beam orientations are combined. p. 8

Where the paper says it

The ripple patterns are substantially stronger when beam rotation is included than when no rotation is applied. This suggests that more frequency-correlated systematics remain in the cleaned data when datasets observed with different beam orientations (i.

This indicates that different beam orientations imprint different systematic structures on the data, which can appear as well-defined higher-order PCA modes when datasets are considered separately. p. 8
Where the paper says it

A possible explanation is that different beam orientations imprint different systematic structures on the data. When each dataset is considered separately, these structures can appear as well-defined higher-order PCA modes, as shown in Figure 11.

Data Splitting for Power Spectrum Analysis

The cross-correlation with optical galaxies will remove the systematic contribution from the power spectrum measurements, while still having an impact on the covariance The authors find that their results on the split data provide additional support for the recent auto-power spectrum measurements presented in MeerKLASS Collaboration et al., 2026. By splitting the datasets to form the “super-cross” (see Fig. p. 1

Where the paper says it

(2025), the measured autopower spectrum of the MeerKLASS L-band deep field was found to contain substantially more power than expected. This excess, which is later mitigated by performing internal cross-correlation between different H i datasets (MeerKLASS Collaboration et al.

2 of MeerKLASS Collaboration et al., 2026) and cleaning the subsets separately, the measurements are free of significant systematics as shown in the comprehensive validation tests therein. The systematic component is reduced so that the ratio increases to 2A0/A5k ≈ 1. p. 0
Where the paper says it

In MeerKLASS L-band data, processing these two types of scans separately significantly suppresses the oscillating residuals while preserving the stacked H i signal. The ratio of the H i signal amplitude to the systematic fluctuation amplitude increases by 81%.

72 and 2.72, respectively, for rising and setting cubes. p. 9
Where the paper says it

Here, we adopt a different approach. We construct the rising and setting cubes by separately combining the data from all rising and setting scans, respectively.

The oscillation is significantly stronger in the final cube than in either the rising or setting cubes. p. 9
Where the paper says it

In Figure 13 and Figure 14, we remove 10 PCA modes from each cube to allow a direct comparison with the results obtained from the final cube. In Figure 15, we instead remove 9 PCA modes from the rising and setting cubes, as this is the optimised case in which the ratio of the H i signal to the foreground residual reaches its maximum.

Final Recommendations

The authors conclude that accounting for the observing strategy and beam orientation can improve foreground cleaning without requiring more aggressive PCA subtraction Future simulations and forecasts for H i intensity-mapping surveys should therefore account not only for beam chromaticity and sidelobes but also for changes in beam orientation during the observations. The authors recommend that combining the rising and setting scans after PCA cleaning is a viable strategy to suppress the oscillating foreground residuals while preserving the stacked H i signal. p. 0

Where the paper says it

The beam rotates with respect to the sky between two types of scans, and combining datasets with different beam orientations before foreground cleaning makes the residuals stronger. In MeerKLASS L-band data, processing these two types of scans separately significantly suppresses the oscillating residuals while preserving the stacked H i signal.

This study provides guidance for developing strategies to mitigate these systematics in future MeerKAT and SKA-Mid surveys. p. 0
Where the paper says it

We find an oscillating structure in PCA mode 11 in our previous data processing, but simply removing this mode does not improve the stacked spectrum due to the signal loss. Our strategy can help mitigate spectral ripples that are induced by the beam in future MeerKAT and SKA-Mid surveys.

The authors suggest that including such information may provide an additional route for improving foreground cleaning, such as multi-scale PCA (Carucci et al. p. 1
Where the paper says it

A variety of blind component separation methods have therefore been developed (Spinelli et al. 2022), such as Principal Component Analysis (PCA) and its multi-scale extension (Carucci et al.

2025) or GMCA (Carucci et al. p. 1
Where the paper says it

2025); Independent Component Analysis (ICA; Chapman et al. 2012); Generalized Morphological Component Analysis (GMCA; Carucci et al.

2020). The authors suggest that combining the rising and setting scans after PCA cleaning is a viable strategy to suppress the oscillating foreground residuals while preserving the stacked H i signal. p. 0
Where the paper says it

The beam rotates with respect to the sky between two types of scans, and combining datasets with different beam orientations before foreground cleaning makes the residuals stronger. In MeerKLASS L-band data, processing these two types of scans separately significantly suppresses the oscillating residuals while preserving the stacked H i signal.

The oscillation is significantly stronger in the final cube than in either the rising or setting cubes, indicating that combining the two scanning strategies enhances the frequency-correlated systematics.

Acknowledgement of Support

This work is supported by the National SKA Program of China (Nos. 2022SKA0110100, 2022SKA0110101, 2022SKA0114) and the NSFC International (Regional) Cooperation and Exchange Project (No.. The authors acknowledge the financial support from the China Scholarship Council (CSC; No. p. 0

Where the paper says it

Our strategy can help mitigate spectral ripples that are induced by the beam in future MeerKAT and SKA-Mid surveys. Key words: cosmology: observations – large-scale structure of Universe – radio lines: general – methods: observational – methods: data analysis © 2026 The Authors ★ E-mail: zhaoby@nao.

. ZC is supported by the Poincaré fellowship, and was funded by a UKRI Future Leaders Fellowship [grant MR/X005399/1; PI: Alkistis Pourtsidou]. p. 0
Where the paper says it

We find an oscillating structure in PCA mode 11 in our previous data processing, but simply removing this mode does not improve the stacked spectrum due to the signal loss. Our strategy can help mitigate spectral ripples that are induced by the beam in future MeerKAT and SKA-Mid surveys.

DILdV was supported by the South African Radio Astronomy Observatory, which is a facility of the National Research Foundation, an agency of the Department of Science and Technology (Grant Number: 75322). p. 0
Where the paper says it

Santos7 , Jingying Wang24,7 , Yougang Wang1,3,18, and Laura Wolz13 1State Key Laboratory of Radio Astronomy and Technology, National Astronomical Observatories, CAS, A20 Datun Road, Chaoyang District, Beijing, 100101, P. China; 2Cavendish Laboratory, University of Cambridge, Cambridge, CB3 0HE, UK 3School of Astronomy and Space Science, University of Chinese Academy of Sciences, Beijing 100049, China 4Kavli Institute for Cosmology in Cambridge, University of Cambridge, Cambridge, CB3 0HE, UK 5 Institute for Astronomy, The University of Edinburgh, Royal Observatory, Edinburgh EH9 3HJ, UK 6Observatoire de la Côte d’Azur, Laboratoire Lagrange, Bd de l’Observatoire, CS 34229, 06304 Nice cedex 4, France 7Department of Physics & Astronomy, University of the Western Cape, Robert Sobukwe Road, Cape Town 7535, South Africa 8Department of Electrical and Electronic Engineering, Stellenbosch University, Stellenbosch, South Africa 9Dipartimento di Fisica, Università degli Studi di Milano, via G.

JF acknowledges support of Fundação para a Ciência e a Tecnologia through the Scientific Employment Incentive program reference No.. The authors have provided direct contributions to the results of this work The authors listed alphabetically from EdLA to DILdV make direct contributions to the results of this work. p. 1
Where the paper says it

Unlike the power spectrum, the stacking results can be directly examined as angular images and frequency spectra. This allows us to see how residual systematics vary across the sky and with frequency, providing a more direct way to investigate their properties and possible origins.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improve foreground cleaning by incorporating information about observing strategy into PCA mode selection, as accounting for the observing strategy and beam orientation can improve foreground cleaning without requiring more aggressive PCA subtraction. This allows AI systems to select optimal PCA modes based on whether data is from a single scan or combined scans with different beam orientations.

  2. Develop a methodology to exploit spatial information during stacking by moving beyond simple averaging, as the stacking procedure of Chen et al. (2025), which uses clustered galaxy positions and sums the signal within 1.2 deg around each galaxy, therefore enhances the visibility of these spatially correlated structures. This enables AI to preferentially combine pixels with similar oscillatory features based on local galaxy clustering.

  3. Implement a strategy for combining datasets from different beam orientations before PCA cleaning to mitigate frequency-correlated residuals, as combining scans with different beam orientations before PCA cleaning produces a more complex effective beam and strengthens the residual frequency correlations. This suggests that combining data sets observed under different scanning conditions can be beneficial for reducing systematic effects.

  4. Integrate a model of the full-sky electromagnetic beam, including chromatic sidelobes, into simulation frameworks to better understand systematics, as we find that a wide range of sidelobes at theta less than 20 deg, 40 deg, 90 deg can introduce additional oscillatory spectral structures into the foreground emission. This helps AI systems to predict and account for complex frequency-dependent structures arising from beam effects.

  5. Optimize PCA mode selection by comparing the impact on both foreground residuals and H i signal loss, as removing mode 11 causes greater H i signal loss and does not improve the stacked spectrum, suggesting that the systematics are more broadly spread among the eigenmodes. This guides AI to choose a balance where removing modes preserves the stacked H i signal while effectively suppressing correlated noise.

Abstract

HI intensity mapping is a promising probe of the large-scale structure of the Universe, but is limited by foregrounds and instrumental systematics. For the MeerKAT Large Area Synoptic Survey (MeerKLASS) L-band data, it is found that emission line stacking measurement reveals a type of systematics related to the instrument beam. Using simulations with a full-sky electromagnetic beam model, we show that the chromaticity and sidelobes of the MeerKAT beam modulate the frequency-smooth foreground emission and introduce oscillating structures that persist after foreground removal. The beam rotates with respect to the sky between two types of scans, and combining datasets with different beam orientations before foreground cleaning makes the residuals stronger. In MeerKLASS L-band data, processing these two types of scans separately significantly suppresses the oscillating residuals while preserving the stacked HI signal. The ratio of the HI signal amplitude to the systematic fluctuation amplitude increases by 81%. We find an oscillating structure in PCA mode 11 in our previous data processing, but simply removing this mode does not improve the stacked spectrum due to the signal loss. Our strategy can help mitigate spectral ripples that are induced by the beam in future MeerKAT and SKA-Mid surveys.

Sources

Related papers