Statistical 21-cm signal separation via Gaussian Process Regression analysis

Detecting and characterizing the Epoch of Reionization (EoR) and Cosmic Dawn via the redshifted 21-cm hyperfine line of neutral hydrogen will revolutionize the study of the formation of the first stars, galaxies, black holes, and intergalactic gas in the infant Universe. The wealth of information encoded in this signal is, however, buried under foregrounds that are many orders of magnitude brighter. These must be removed accurately and precisely in order to reveal the feeble 21-cm signal. This requires not only the modelling of the Galactic and extragalactic emission, but also of the often stochastic residuals due to imperfect calibration of the data caused by ionospheric and instrumental distortions. To stochastically model these effects, we introduce a new method based on ` Gaussian Process Regression' (GPR) which is able to statistically separate the 21-cm signal from most of the foregrounds and other contaminants. Using simulated LOFAR-EoR data that include strong instrumental mode mixing, we show that this method is capable of recovering the 21-cm signal power spectrum across the entire range k = 0.07 - 0.3 h cMpc (1). The GPR method is most optimal, having minimal and controllable impact on the 21-cm signal, when the foregrounds are correlated on frequency scales greater than or similar to 3 MHz and the rms of the signal has sigma(21cm) greater than or similar to 0.1 sigma(noise). This signal separation improves the 21-cm power-spectrum sensitivity by a factor greater than or similar to 3 compared to foreground avoidance strategies and enables the sensitivity of current and future 21-cm instruments such as the Square Kilometre Array to be fully exploited.

[1]  L. Koopmans,et al.  Scintillation noise power spectrum and its impact on high redshift 21-cm observations , 2015, 1512.00159.

[2]  S. Zaroubi,et al.  Foreground simulations for the LOFAR-epoch of reionization experiment , 2008, 0804.1130.

[3]  Saleem Zaroubi,et al.  Foreground removal using fastica: A showcase of LOFAR-EoR , 2012, 1201.2190.

[4]  Alex G. Kim,et al.  Robust Strong Lensing Time Delay Estimation , 2013, 1304.0309.

[5]  J. Schaye,et al.  Upper Limits on the 21 cm Epoch of Reionization Power Spectrum from One Night with LOFAR , 2017, 1702.08679.

[6]  L. Koopmans,et al.  Deconvolving the wedge: maximum-likelihood power spectra via spherical-wave visibility modelling , 2017, 1709.06752.

[7]  David B. Dunson,et al.  Bayesian Data Analysis , 2010 .

[8]  Nico S. Gorbach,et al.  Model Selection for Gaussian Process Regression by Approximation Set Coding , 2016, 1610.00907.

[9]  H. Jeffreys,et al.  Theory of probability , 1896 .

[10]  N. Kern,et al.  Emulating Simulations of Cosmic Dawn for 21 cm Power Spectrum Constraints on Cosmology, Reionization, and X-Ray Heating , 2017, 1705.04688.

[11]  K. Grainge,et al.  Analysing the impact of far-out sidelobes on the imaging performance of the SKA-LOW telescope , 2016, 1602.01805.

[12]  Cathryn M. Trott,et al.  THE IMPACT OF POINT-SOURCE SUBTRACTION RESIDUALS ON 21 cm EPOCH OF REIONIZATION ESTIMATION , 2012, 1208.0646.

[13]  Jonathan C. Pober,et al.  The impact of foregrounds on redshift space distortion measurements with the highly redshifted 21-cm line , 2014, 1411.2050.

[14]  Abhirup Datta,et al.  BRIGHT SOURCE SUBTRACTION REQUIREMENTS FOR REDSHIFTED 21 cm MEASUREMENTS , 2010 .

[15]  Judd D. Bowman,et al.  IMPROVING FOREGROUND SUBTRACTION IN STATISTICAL OBSERVATIONS OF 21 cm EMISSION FROM THE EPOCH OF REIONIZATION , 2006 .

[16]  Daniel A. Mitchell,et al.  CHIPS: THE COSMOLOGICAL H i POWER SPECTRUM ESTIMATOR , 2016, 1601.02073.

[17]  Bryna Hazelton,et al.  FOUR FUNDAMENTAL FOREGROUND POWER SPECTRUM SHAPES FOR 21 cm COSMOLOGY OBSERVATIONS , 2012, 1202.3830.

[18]  M. Morales,et al.  THE FUNDAMENTAL MULTI-BASELINE MODE-MIXING FOREGROUND IN 21 cm EPOCH OF REIONIZATION OBSERVATIONS , 2013, 1301.3126.

[19]  Saleem Zaroubi,et al.  Non-parametric foreground subtraction for 21-cm epoch of reionization experiments , 2009 .

[20]  S. Zaroubi,et al.  Wide-field LOFAR-LBA power-spectra analyses: Impact of calibration, polarization leakage and ionosphere , 2017, Proceedings of the International Astronomical Union.

[21]  Cathryn M. Trott,et al.  Epoch of reionization window. I. Mathematical formalism , 2014, 1404.2596.

[22]  Anna Bonaldi,et al.  Foreground removal for Square Kilometre Array observations of the Epoch of Reionization with the Correlated Component Analysis , 2014, 1409.5300.

[23]  Bradley Greig,et al.  21CMMC: an MCMC analysis tool enabling astrophysical parameter studies of the cosmic 21 cm signal , 2015, 1501.06576.

[24]  Cathryn M. Trott,et al.  Epoch of reionization window. II. Statistical methods for foreground wedge reduction , 2014, 1404.4372.

[25]  David F. Moore,et al.  PAPER-64 CONSTRAINTS ON REIONIZATION: THE 21 cm POWER SPECTRUM AT z = 8.4 , 2015, 1502.06016.

[26]  Roger Woodard,et al.  Interpolation of Spatial Data: Some Theory for Kriging , 1999, Technometrics.

[27]  James Aguirre,et al.  A SENSITIVITY AND ARRAY-CONFIGURATION STUDY FOR MEASURING THE POWER SPECTRUM OF 21 cm EMISSION FROM REIONIZATION , 2011, 1103.2135.

[28]  S. Zaroubi,et al.  Realistic simulations of the Galactic polarized foreground: consequences for 21‐cm reionization detection experiments , 2010, 1007.4135.

[29]  A. Mesinger.  Understanding the epoch of cosmic reionization : challenges and progress , 2016 .

[30]  Christopher K. I. Williams,et al.  Gaussian Processes for Machine Learning (Adaptive Computation and Machine Learning) , 2005 .

[31]  David R. DeBoer,et al.  WHAT NEXT-GENERATION 21 cm POWER SPECTRUM MEASUREMENTS CAN TEACH US ABOUT THE EPOCH OF REIONIZATION , 2013, 1310.7031.

[32]  Max Tegmark,et al.  FOREGROUNDS IN WIDE-FIELD REDSHIFTED 21 cm POWER SPECTRA , 2015, 1502.07596.

[33]  A. H. Patil,et al.  Systematic biases in low-frequency radio interferometric data due to calibration: the LOFAR-EoR case , 2016, 1605.07619.

[34]  Hannes Jensen,et al.  The wedge bias in reionization 21-cm power spectrum measurements , 2015, 1509.02277.

[35]  Sarod Yatawatta,et al.  Fine tuning consensus optimization for distributed radio interferometric calibration , 2016, 2016 24th European Signal Processing Conference (EUSIPCO).

[36]  S. Aigrain,et al.  K2SC: flexible systematics correction and detrending of K2 light curves using Gaussian process regression , 2016, 1603.09167.

[37]  A. R. Whitney,et al.  FIRST SEASON MWA EOR POWER SPECTRUM RESULTS AT REDSHIFT 7 , 2016, 1608.06281.

[38]  S. Furlanetto,et al.  Efficient Simulations of Early Structure Formation and Reionization , 2007, 0704.0946.

[39]  J. Hewitt,et al.  The impact of modelling errors on interferometer calibration for 21 cm power spectra , 2016, 1610.02689.

[40]  R. Cen,et al.  21cmfast: a fast, seminumerical simulation of the high‐redshift 21‐cm signal , 2010, 1003.3878.

[41]  A. A. Deshpande,et al.  Empirical covariance modeling for 21 cm power spectrum estimation: A method demonstration and new limits from early Murchison Widefield Array 128-tile data , 2015, 1506.01026.

[42]  L. Koopmans.  IONOSPHERIC POWER-SPECTRUM TOMOGRAPHY IN RADIO INTERFEROMETRY , 2010, 1005.1898.

[43]  J. Starck,et al.  The scale of the problem: Recovering images of reionization with Generalized Morphological Component Analysis , 2012, 1209.4769.

[44]  Abhik Ghosh,et al.  A Bayesian analysis of redshifted 21-cm H I signal and foregrounds: simulations for LOFAR , 2015, 1506.04982.

[45]  M. Morales,et al.  Calibration requirements for detecting the 21 cm epoch of reionization power spectrum and implications for the SKA , 2016, 1603.00607.

[46]  A. Ghosh,et al.  Improved foreground removal in GMRT 610 MHz observations towards redshifted 21-cm tomography , 2011, 1108.3707.