First upper limits on the 21-cm signal power spectrum from the Cosmic Dawn from one night of observations with NenuFAR 

Posted on by
Reply

The NenuFAR Cosmic Dawn Key-Science project has recently made significant progress in exploring the early universe through the redshifted 21-cm transition line from neutral hydrogen. A recent study [1], using just a single night of observations of the North Celestial Pole deep field with the NenuFAR radio telescope, has set a new upper limits on the 21-cm power spectrum from the Cosmic Dawn at a redshift of z = 20.3. The best 2-σ upper limit of 2.4 × 107 mK2 at k = 0.041 h cMpc−1 at z = 20.3 was observed, the deepest yet in this redshift range.

Spherical and cylindrical power spectra at some key stages of the analysis pipeline. The left-most panel shows the spherical power spectra after pre-processing (“Data”), after sky model subtraction (“Skymodel Sub”), and after GPR (“GPR Residual”), along with the thermal noise power spectrum (“Thermal Noise”). For the cylindrical power spectra (second panels), the ratio with respect to the thermal noise power spectrum is shown. The final spherical power spectra after noise bias subtraction and suppression factor correction are shown in the right-most pane.


NenuFAR, a low-frequency radio interferometer located at the Nançay Radio Observatory in France, stands out for its dense uv-coverage at short baselines, making it exceptionally sensitive to observe the 21-cm signal from the Cosmic Dawn. Observing the 21-cm signal from the Cosmic Dawn is particularly challenging due to the overwhelming presence of foreground emissions from our galaxy and other celestial sources, which significantly overpowers the faint signal from the early universe. This new analysis adopted a sophisticated approach involving several steps of foreground subtraction, with a particular emphasis on accurately removing the brightest radio sources in the sky through direction-dependent calibration. Following the subtraction of compact sources, residual foregrounds were addressed using the recently developed ML-GPR method [2]. The process included extensive and rigorous testing, such as the injection of mock 21-cm signals, to ensure robustness of the result.

This milestone represents a critical stride toward the eventual detection of the signal by NenuFAR. However, the journey ahead is still full of challenges. The team plans to continue investigating the sources of excess power and refining their analysis pipeline. The team also look forward towards upcoming NenuFAR Cosmic Dawn observations, which will focus on newly selected deep fields, carefully chosen to mitigate some systematic effects observed in the North Celestial Pole deep field.

[1] First upper limits on the 21-cm signal power spectrum from the Cosmic Dawn from one night of observations with NenuFAR  (https://arxiv.org/abs/2311.05364)
[2] Retrieving the 21-cm signal from the Epoch of Reionization with learnt Gaussian process kernels (https://arxiv.org/abs/2307.13545)

First NenuFAR Cosmic Dawn paper

Posted on by
Reply

The first Nenufar CD KSP paper, Accurate modelling of the Lyman-α coupling for the 21-cm signal, observability with NenuFAR, and SKA (Semelin et al 2023), has been published in A&A.
The full text can be found here, and here is the abstract :

The measurement of the 21 cm signal from the Cosmic Dawn is a major goal for several existing and upcoming radio interferometers such as NenuFAR and SKA. During this era before the beginning of the Epoch of Reionisation, the signal is more difficult to observe due to brighter foregrounds, but it reveals additional information on the underlying astrophysical processes encoded in the spatial fluctuations of the spin temperature of hydrogen. To interpret future measurements, controlling the level of accuracy of the Lyman-α flux modelling is mandatory. In this work, we evaluate the impact of various approximations that exist in the main fast modelling approach compared to the results of a costly full radiative transfer simulation. The fast SPINTER code, presented in this work, computes the Lyman-α flux including the effect of wing scatterings for an inhomogeneous emissivity field, but assuming an otherwise homogeneous expanding universe. The LICORICE code computes the full radiative transfer in the Lyman-α line without any substantial approximation. We find that the difference between homogeneous and inhomogeneous gas density and temperature is very small for the computed flux. On the contrary, neglecting the effect of gas velocities produces a significant change in the computed flux. We identify the causes (mainly Doppler shifts due to velocity gradients) and quantify the magnitude of the effect in both an idealised setup and a realistic cosmological situation. We find that the amplitude of the effect, up to a factor of ∼2 on the 21 cm signal power spectrum on some scales (depending on both other model parameters and the redshift), can be easily discriminated with an SKA-like survey and can already be approached, particularly for exotic signals, by the ongoing NenuFAR Cosmic Dawn Key Science Program.

Image of the month : Cosmic Dawn III, a fully coupled radiative hydrodynamic simulation of the first billion years (May 2022)

Posted on by
Reply

8 cMpc/h sub-region of the simulation, at z=7.3. White : gas density. Blue : ionizing photon density. Red : hot gas, heated by supernovae and shocks.

Cosmic Dawn III (CoDa III) is the latest iteration of the cosmological simulations of the EoR developed at the Observatoire astronomique de Strasbourg. It was run on Summit at the Oak ridge Leadership Computing Facility, the most powerful computing center in the world until June 2020. It required 131072 CPUs, 24576 GPUs, and produced roughly 20 PB of data !

Alongside gravity, hydrodynamics, star formation, and radiative transfer, the ~94 cMpc simulation includes processes such as supernovae feedback and chemical enrichment. The mass resolution in CoDa III is 8 times finer than in CoDa II (2020) which allows for a better description of galactic haloes and absorbers in the intergalactic medium. Coda III features improved models for star formation, ionizing emissivity, and dust, with respect to CoDa II.

Movies of the evolution of the simulation on different scales are available on Pierre Ocvirk’s Youtube Channel.

Ref : Lewis, J. S., Ocvirk, P., et al (2022). The short ionizing photon mean free path at z= 6 in Cosmic Dawn III, a new fully-coupled radiation-hydrodynamical simulation of the Epoch of Reionization. arXiv preprint arXiv:2202.05869.

Credits : P. Ocvirk, D. Aubert, J. Lewis, N. Gillet, J. Chardin

.

NenuFAR Cosmic Dawn Workshop (April 2022)

Image

Reply

The NenuFAR Cosmic Dawn team met in Paris on 26 April for the 3rd Cosmic Dawn workshop. In a milder weather than at our previous meeting in November, we had a very rich afternoon of news about the NenuFAR telescope, updates on Cosmic Dawn observations and data processing, the latest news on the ACE project, and from 21-cm simulations. The meeting was contiguous with the LOFAR-EoR plenary meeting which allowed the two teams to meet and share.

NenuFAR image of the 3C196 field

Posted on by
Reply

As human activity is slowing all around the world, NenuFAR keeps observing. Over three days mid-March, we pointed the telescope toward 3C196, one of the main target fields of the LOFAR-EoR project. These observations which are part of stage 1 of our project were done to look at the impact of foregrounds and instrumental response to the power-spectra relative to what we observe on our main target field, the North Celestial Pole. The data was quickly processed and we produced wide-field images covering the full 30-85 MHz bandwidth. For each of these images, we accumulated about 7 hours of observation. The dashed black circle shows the full-width at half maximum of the instrument’s primary beam (note the changing field of view for each panel). We will now consolidate these results and compare them to our NCP observations.

Credit : Florent Mertens

Diffuse Galactic Emission around the North Celestial Pole with NenuFAR

Posted on by
Reply

This animated image shows Stokes I 40×40 degrees images centered on the North Celestial Pole as seen by NenuFAR at four frequencies ranges: 29-41.5 MHz, 41.5 -54 MHz, 54-66.5 MHz, 66.5-82.1 MHz. For comparison, an AARTFAAC HBA image at 122.1 MHz produced by Bharat Gehlot (Large-scale 21-cm Cosmology with LOFAR and AARTFAAC, Chapter 5, Fig. 5.2) is also included. The current 400-meter maximum baseline of NenuFAR limits the resolution to 1.2 degrees at the lowest frequency and 0.5 degrees at the highest frequency, while the AARTFAAC image has a resolution of 0.2 degrees. The upcoming addition of remote mini-arrays to NenuFAR will push the resolution down to 4 arcmins at 85 MHz. The Field of View NenuFAR is set by the mini-array diameter (25 meters) which gives a primary-beam FWHM of 20 and 9 degrees at the lowest and highest frequency respectively (26 degrees for AARTFAAC HBA image at 122.1 MHz). Despite the very different resolution and Field of View, the similarities in the images are striking. The smallest baselines were included in making these images at which large-scale diffuse Galactic emission, clearly and consistently seen in these images, starts to dominate. Modeling and including this diffuse emission in the calibration sky-model can significantly enhance calibration of the instrument.

Credit : Florent Mertens

Wide-Field view towards the North Celestial Pole with NenuFAR (February 2020)

Posted on by
Reply

The first phase of observation of the NenuFAR Cosmic Dawn KSP is now behind us. Between the 13th of July and 25 of December 2019, we have accumulated in total 340 hours of observation on 18 spectral windows covering 55 MHz of bandwidth between 30 MHz and 85 MHz. More than half of these datasets have now been fully processed from which we obtained a wide-field image-cube centred on the North Celestial Pole (NCP) with a spatial resolution between 1 and 0.5 degrees. In these images, the field around the NCP is confusion limited, but we also observe many bright radio sources at distances up to 70 degrees from the NCP from which I annotated some of the most famous ones in the above figure. The brightest ones, Cygnus A and Cassiopeia A have been subtracted. After completion and verification, this image cube will be released to the CD KSP team. Besides finalizing the analysis of phase 1, we are now fully focused on preparing the phase 2 observations, which will bring higher spatial (new remote mini-arrays) and frequency resolution (new correlator) and with which we will start our deep integration toward the detecting of the 21-cm signal from the Cosmic Dawn!

Credit : Florent Mertens