arXiv subjects

B. Semelin

Publications and source records attributed to B. Semelin.

At least 19 recordsLinked to original sources

arXiv2025

YOLO-CIANNA: Galaxy detection with deep learning in radio data: II. Winning the SKA SDC2 using a generalized 3D-YOLO network

As the scientific exploitation of the Square Kilometre Array (SKA) approaches, there is a need for new advanced data analysis and visualization tools capable of processing large high-dimensional datasets. In this study, we aim to generalize the YOLO-CIANNA deep learning source detection and characterization method for 3D hyperspectral HI emission cubes. We present the adaptations we made to the regression-based detection formalism and the construction of an end-to-end 3D convolutional neural network (CNN) backbone. We then describe a processing pipeline for applying the method to simulated 3D HI cubes from the SKA Observatory Science Data Challenge 2 (SDC2) dataset. The YOLO-CIANNA method was originally developed and used by the MINERVA team that won the official SDC2 competition. Despite the public release of the full SDC2 dataset, no published result has yet surpassed MINERVA's top score. In this paper, we present an updated version of our method that improves our challenge score by 9.5%. The resulting catalog exhibits a high detection purity of 92.3%, best-in-class characterization accuracy, and contains 45% more confirmed sources than concurrent classical detection tools. The method is also computationally efficient, processing the full ~1TB SDC2 data cube in 30 min on a single GPU. These state-of-the-art results highlight the effectiveness of 3D CNN-based detectors for processing large hyperspectral data cubes and represent a promising step toward applying YOLO-CIANNA to observational data from SKA and its precursors.

astro-ph.GA↗

arXiv2025

Improved upper limits on the 21-cm signal power spectrum at $z=17.0$ and $z=20.3$ from an optimal field observed with NenuFAR

We report the deepest upper limits to date on the power spectrum of the 21-cm signal during the Cosmic Dawn (redshifts: $z>15$), using four nights of observations with NenuFAR. The limits are derived from two redshift bins, centred at $z=20.3$ and $z=17.0$, with integration times of 26.1 h and 23.6 h, from observations of an optimal target field chosen to minimise sidelobe leakage from bright sources. Our analysis incorporates improvements to the data processing pipeline, particularly in subtracting strong radio sources in the primary beam sidelobes and mitigating low-level radio frequency interference, yielding a 50-fold reduction in the excess variance compared to a previous analysis of the north celestial pole field. At $z=20.3$, we achieve a best $2\sigma$ upper limit of $\Delta^{2}_{21}<4.6 \times 10^5 \, \textrm{mK}^{2}$ at $k=0.038$ $h\, \mathrm{cMpc}^{-1}$, while at $z=17.0$, the best limit is $\Delta^{2}_{21}<5.0 \times 10^6 \, \textrm{mK}^{2}$ at $k=0.041$ $h\, \mathrm{cMpc}^{-1}$. These are the strongest constraints on the 21-cm power spectrum at the respective redshifts, with the limit at $z = 20.3$ being deeper by more than an order of magnitude over all previous Cosmic Dawn power spectrum limits. Comparison against simulated exotic 21-cm signals shows that while the $z=20.3$ limits begin to exclude the most extreme models predicting signals stronger than the EDGES detection, an order-of-magnitude improvement would constrain signals compatible with EDGES. A coherence analysis reveals that the excess variance is largely incoherent across nights for the $z=20.3$ redshift bin, suggesting that deeper integrations could yield significantly stronger constraints on the 21-cm signal from the Cosmic Dawn.

astro-ph.CO↗

arXiv2025

Near-field imaging of local interference in radio interferometric data: Impact on the redshifted 21 cm power spectrum

Radio-frequency interference (RFI) is a major systematic limitation in radio astronomy, particularly for science cases requiring high sensitivity, such as 21 cm cosmology. Traditionally, RFI is dealt with by identifying its signature in the dynamic spectra of visibility data and flagging strongly affected regions. However, for RFI sources that do not occupy narrow regions in the time-frequency space, such as persistent local RFI, modeling these sources could be essential to mitigating their impact. This paper introduces two methods for detecting and characterizing local RFI sources from radio interferometric visibilities: matched filtering and maximum a posteriori (MAP) imaging. These algorithms use the spherical wave equation to construct three-dimensional near-field image cubes of RFI intensity from the visibilities. The matched filter algorithm can generate normalized maps by cross-correlating the expected contributions from RFI sources with the observed visibilities, while the MAP method performs a regularized inversion of the visibility equation in the near field. We developed a full polarization simulation framework for RFI and demonstrated the methods on simulated observations of local RFI sources. The stability, speed, and errors introduced by these algorithms were investigated, and, as a demonstration, the algorithms were applied to a subset of NenuFAR observations to perform spatial, spectral, and temporal characterization of two local RFI sources. We used simulations to assess the impact of local RFI on images, the uv plane, and cylindrical power spectra, and to quantify the level of bias introduced by the algorithms in order to understand their implications for the estimated 21 cm power spectrum with radio interferometers. The near-field imaging and simulation codes are publicly available in the Python library nfis.

astro-ph.IM↗

arXiv2024

YOLO-CIANNA: Galaxy detection with deep learning in radio data. I. A new YOLO-inspired source detection method applied to the SKAO SDC1

The upcoming Square Kilometer Array (SKA) will set a new standard regarding data volume generated by an astronomical instrument, which is likely to challenge widely adopted data-analysis tools that scale inadequately with the data size. The aim of this study is to develop a new source detection and characterization method for massive radio astronomical datasets based on modern deep-learning object detection techniques. For this, we seek to identify the specific strengths and weaknesses of this type of approach when applied to astronomical data. We introduce YOLO-CIANNA, a highly customized deep-learning object detector designed specifically for astronomical datasets. In this paper, we present the method and describe all the elements introduced to address the specific challenges of radio astronomical images. We then demonstrate the capabilities of this method by applying it to simulated 2D continuum images from the SKA observatory Science Data Challenge 1 (SDC1) dataset. Using the SDC1 metric, we improve the challenge-winning score by +139% and the score of the only other post-challenge participation by +61%. Our catalog has a detection purity of 94% while detecting 40 to 60% more sources than previous top-score results, and exhibits strong characterization accuracy. The trained model can also be forced to reach 99% purity in post-process and still detect 10 to 30% more sources than the other top-score methods. It is also computationally efficient, with a peak prediction speed of 500 images of 512x512 pixels per second on a single GPU. YOLO-CIANNA achieves state-of-the-art detection and characterization results on the simulated SDC1 dataset and is expected to transfer well to observational data from SKA precursors.

astro-ph.IM↗

arXiv2023

First upper limits on the 21 cm signal power spectrum from cosmic dawn from one night of observations with NenuFAR

The redshifted 21 cm signal from neutral hydrogen is a direct probe of the physics of the early universe and has been an important science driver of many present and upcoming radio interferometers. In this study we use a single night of observations with the New Extension in Nan\c{c}ay Upgrading LOFAR (NenuFAR) to place upper limits on the 21 cm power spectrum from cosmic dawn at a redshift of $z$ = 20.3. NenuFAR is a new low-frequency radio interferometer, operating in the 10-85 MHz frequency range, currently under construction at the Nan\c{c}ay Radio Observatory in France. It is a phased array instrument with a very dense uv coverage at short baselines, making it one of the most sensitive instruments for 21 cm cosmology analyses at these frequencies. Our analysis adopts the foreground subtraction approach, in which sky sources are modeled and subtracted through calibration and residual foregrounds are subsequently removed using Gaussian process regression. The final power spectra are constructed from the gridded residual data cubes in the uv plane. Signal injection tests are performed at each step of the analysis pipeline, the relevant pipeline settings are optimized to ensure minimal signal loss, and any signal suppression is accounted for through a bias correction on our final upper limits. We obtain a best 2$\sigma$ upper limit of $2.4\times 10^7$ $\text{mK}^{2}$ at $z$ = 20.3 and $k$ = 0.041 $h\,\text{cMpc}^{-1}$. We see a strong excess power in the data, making our upper limits two orders of magnitude higher than the thermal noise limit. We investigate the origin and nature of this excess power and discuss further improvements to the analysis pipeline that can potentially mitigate it and consequently allow us to reach thermal noise sensitivity when multiple nights of observations are processed in the future.

astro-ph.CO↗

arXiv2023

SKA Science Data Challenge 2: analysis and results

The Square Kilometre Array Observatory (SKAO) will explore the radio sky to new depths in order to conduct transformational science. SKAO data products made available to astronomers will be correspondingly large and complex, requiring the application of advanced analysis techniques to extract key science findings. To this end, SKAO is conducting a series of Science Data Challenges, each designed to familiarise the scientific community with SKAO data and to drive the development of new analysis techniques. We present the results from Science Data Challenge 2 (SDC2), which invited participants to find and characterise 233245 neutral hydrogen (Hi) sources in a simulated data product representing a 2000~h SKA MID spectral line observation from redshifts 0.25 to 0.5. Through the generous support of eight international supercomputing facilities, participants were able to undertake the Challenge using dedicated computational resources. Alongside the main challenge, `reproducibility awards' were made in recognition of those pipelines which demonstrated Open Science best practice. The Challenge saw over 100 participants develop a range of new and existing techniques, with results that highlight the strengths of multidisciplinary and collaborative effort. The winning strategy -- which combined predictions from two independent machine learning techniques to yield a 20 percent improvement in overall performance -- underscores one of the main Challenge outcomes: that of method complementarity. It is likely that the combination of methods in a so-called ensemble approach will be key to exploiting very large astronomical datasets.

astro-ph.IM↗

arXiv2021

Exploring the Cosmic Dawn with NenuFAR

The exploration of the Cosmic Dawn, the period of the Universe during which the first stars and galaxies were formed, is one of the last frontiers of modern astronomy and cosmology. The redshifted 21-cm line emission from neutral hydrogen is a unique probe and can open this era for astrophysical and cosmological studies. The tentative detection of the 21-cm global signal by the EDGES team at $z \sim 17$ underlines the need for an interferometric detection of this signal to discriminate between the numerous models trying to explain this unexpected discovery. The NenuFAR Cosmic-Dawn Key-Science Program aims to perform this detection in the redshift range $z \sim 15 - 31$ with a novel SKA precursor, the NenuFAR radio telescope located at the Station de Radioastronomie de Nan\c{c}ay and that started operating in 2019. Due to its compactness it is particularly sensitive to the large scale of the 21-cm signal. Only 100 hours of observation are needed to reach the level of the most extreme models, while 1000 hours are needed for the more standard models. Observations have already started, accumulating to almost 500 hours on the North Celestial Pole field. In this contribution, we introduce the project, our first results and the developments in calibration and RFI mitigation specific to this new instrument.

astro-ph.CO↗

arXiv2017

What the Milky Way bulge reveals about the initial metallicity gradients in the disc

We examine the metallicity trends in the Milky Way (MW) bulge - using APOGEE DR13 data - and explore their origin by comparing two N-body models of isolated galaxies which develop a bar and a boxy/peanut (b/p) bulge. Both models have been proposed as scenarios for reconciling a disc origin of the MW bulge with a negative vertical metallicity gradient. The first model is a superposition of co-spatial, i.e. overlapping, disc populations with different scaleheights, kinematics and metallicities. In this model the thick, metal-poor, and centrally concentrated disc populations contribute significantly to the stellar mass budget in the inner galaxy. The second model is a single disc with an initial steep radial metallicity gradient, which is mapped by the bar into the b/p bulge in such a way that the vertical metallicity gradient of the MW bulge is reproduced -- as shown already in previous works in the literature. However, as we show here, the latter model does not reproduce the positive longitudinal metallicity gradient of the inner disc, nor the metal-poor innermost regions of the Bulge seen in the data. On the other hand, the model with co-spatial thin and thick disc populations reproduces all the aforementioned trends. We therefore see that it is possible to reconcile a (primarily) disc origin for the MW bulge with the observed trends in metallicity by mapping the inner thin and thick discs of the MW into a b/p. For this scenario to reproduce the observations, the $α$-enhanced, metal-poor, thick disc populations must have a significant mass contribution in the inner regions -- as has been suggested for the Milky Way.

astro-ph.GA↗

arXiv2017

Bars & boxy/peanut bulges in thin & thick discs: I. Morphology and line-of-sight velocities of a fiducial model

We explore trends in the morphology and line-of-sight (los) velocity of stellar populations in the inner regions of disc galaxies, using N-body simulations with both a thin (kinematically cold) and a thick (kinematically hot) disc which form a bar and boxy/peanut (b/p) bulge. The bar in the thin disc component is $\sim$50\% stronger than the thick disc bar and is more elongated, with an axis ratio almost half that of the thick disc bar. The thin disc b/p bulge has a pronounced X-shape, while the thick disc b/p is weaker with a rather boxy shape. This leads to the signature of the b/p bulge in the thick disc to be weaker and further away from the plane than in the thin disc. Regarding the kinematics, we find that the los velocity of thick disc stars in the outer parts of the b/p bulge can be \emph{larger} than that of thin disc stars, by up to 40\% and 20\% for side-on and Milky Way-like orientations of the bar respectively. This is due to the different orbits followed by thin and thick disc stars in the bar-b/p region, which are affected by the fact that: i) thin disc stars are trapped more efficiently in the bar - b/p instability and thus lose more angular momentum than their thick disc counterparts and ii) thick disc stars have large radial excursions and therefore stars from large radii with high angular momenta can be found in the bar region. We also find that the difference between the los velocities of the thin and thick disc in the b/p bulge ($Δv_{los}$) correlates with the initial difference between the radial velocity dispersions of the two discs ($Δσ$) . We therefore conclude that stars in the bar - b/p bulge will have considerably different morphologies and kinematics depending on the kinematic properties of the disc population they originate from.

astro-ph.GA↗

arXiv2016

On the kinematic detection of accreted streams in the Gaia era: a cautionary tale

The $Λ$CDM cosmological scenario predicts that our Galaxy should contain hundreds of stellar streams at the solar vicinity, fossil relics of the merging history of the Milky Way and more generally of the hierarchical growth of galaxies. Because of the mixing time scales in the inner Galaxy, it has been claimed that these streams should be difficult to detect in configuration space but can still be identifiable in kinematic-related spaces like the energy/angular momenta spaces, E-Lz and Lperp-Lz, or spaces of orbital/velocity parameters. By means of high-resolution, dissipationless N-body simulations, containing between 25$\times10^6$ and 35$\times10^6$ particles, we model the accretion of a series of up to four 1:10 mass ratio satellites then up to eight 1:100 satellites and we search systematically for the signature of these accretions in these spaces. In all spaces considered (1) each satellite gives origin to several independent overdensities; (2) overdensities of multiple satellites overlap; (3) satellites of different masses can produce similar substructures; (4) the overlap between the in-situ and the accreted population is considerable everywhere; (5) in-situ stars also form substructures in response to the satellite(s) accretion. These points are valid even if the search is restricted to kinematically-selected halo stars only. As we are now entering the 'Gaia era', our results warn that an extreme caution must be employed before interpreting overdensities in any of those spaces as evidence of relics of accreted satellites. Reconstructing the accretion history of our Galaxy will require a substantial amount of accurate spectroscopic data, that, complemented by the kinematic information, will possibly allow us to (chemically) identify accreted streams and measure their orbital properties. (abridged)

astro-ph.GA↗

arXiv2015

The Cosmic Dawn and Epoch of Reionization with the Square Kilometre Array

Concerted effort is currently ongoing to open up the Epoch of Reionization (EoR) ($z\sim$15-6) for studies with IR and radio telescopes. Whereas IR detections have been made of sources (Lyman-$\alpha$ emitters, quasars and drop-outs) in this redshift regime in relatively small fields of view, no direct detection of neutral hydrogen, via the redshifted 21-cm line, has yet been established. Such a direct detection is expected in the coming years, with ongoing surveys, and could open up the entire universe from $z\sim$6-200 for astrophysical and cosmological studies, opening not only the EoR, but also its preceding Cosmic Dawn ($z\sim$30-15) and possibly even the later phases of the Dark Ages ($z\sim$200-30). All currently ongoing experiments attempt statistical detections of the 21-cm signal during the EoR, with limited signal-to-noise. Direct imaging, except maybe on the largest (degree) scales at lower redshifts, as well as higher redshifts will remain out of reach. The Square Kilometre Array(SKA) will revolutionize the field, allowing direct imaging of neutral hydrogen from scales of arc-minutes to degrees over most of the redshift range $z\sim$6-28 with SKA1-LOW, and possibly even higher redshifts with the SKA2-LOW. In this SKA will be unique, and in parallel provide enormous potential of synergy with other upcoming facilities (e.g. JWST). In this chapter we summarize the physics of 21-cm emission, the different phases the universe is thought to go through, and the observables that the SKA can probe, referring where needed to detailed chapters in this volume (Abridged).

astro-ph.CO↗

arXiv2014

Why the Milky Way's bulge is not only a bar formed from a cold thin disk

By analyzing a N-body simulation of a bulge formed simply via a bar instability mechanism operating on a kinematically cold stellar disk, and by comparing the results of this analysis with the structural and kinematic properties of the main stellar populations of the Milky Way bulge, we conclude that the bulge of our Galaxy is not a pure stellar bar formed from a pre-existing thin stellar disk, as some studies have recently suggested. On the basis of several arguments emphasized in this paper, we propose that the bulge population which, in the Milky Way, is observed not to be part of the peanut structure corresponds to the old galactic thick disk, thus implying that the Milky Way is a pure thin+thick disk galaxy, with only a possible limited contribution of a classical bulge.

astro-ph.GA↗

arXiv2014

Mapping a stellar disk into a boxy bulge: The outside-in part of the Milky Way bulge formation

By means of idealized, dissipationless N-body simulations which follow the formation and subsequent buckling of a stellar bar, we study the characteristics of boxy/peanut-shaped bulges and compare them with the properties of the stellar populations in the Milky Way bulge. The main results of our modeling, valid for the general family of boxy/peanut shaped bulges, are the following: (i) because of the spatial redistribution in the disk initiated at the epoch of bar formation, stars from the innermost regions to the outer Lindblad resonance of the stellar bar are mapped into a boxy bulge; (ii) the contribution of stars to the local bulge density depends on their birth radius: stars born in the innermost disk tend to dominate the innermost regions of the boxy bulge, while stars originating closer to the OLR are preferentially found in the outer regions of the boxy/peanut structure; (iii) stellar birth radii are imprinted in the bulge kinematics, the larger the birth radii of stars ending up in the bulge, the greater their rotational support and the higher their line-of- sight velocity dispersions (but note that this last trend depends on the bar viewing angle); (iv) the higher the classical bulge-over-disk ratio, the larger its fractional contribution of stars at large vertical distance from the galaxy mid-plane. (ABRIDGED) On the basis of their chemical and kinematic characteristics, the results of our modeling suggests that the populations A, B and C, as defined by the ARGOS survey, can be associated, respectively, with the inner thin disk, to the young thick and to the old thick disk, following the nomenclature recently suggested for stars in the solar neighborhood by Haywood et al. (2013).

astro-ph.GA↗

arXiv2013

Signatures of radial migration in barred galaxies: Azimuthal variations in the metallicity distribution of old stars

By means of N-body simulations, we show that radial migration in galaxy disks, induced by bar and spiral arms, leads to significant azimuthal variations in the metallicity distribution of old stars at a given distance from the galaxy center. Metals do not show an axisymmetric distribution during phases of strong migration. Azimuthal variations are visible during the whole phase of strong bar phase, and tend to disappear as the effect of radial migration diminishes, together with a reduction in the bar strength. These results suggest that the presence of inhomogeneities in the metallicity distribution of old stars in a galaxy disk can be a probe of ongoing strong migration. Such signatures may be detected in the Milky Way by Gaia (and complementary spectroscopic data), as well as in external galaxies, by IFU surveys like CALIFA and ATLAS3D. Mixing - defined as the tendency toward a homogeneous, azimuthally symmetric, stellar distribution in the disk - and migration turns out to be two distinct processes, the effects of mixing starting to be visible when strong migration is over.

astro-ph.GA↗

arXiv2012

Joint Lyα emitters - quasars reionization constraints

We present a novel method to investigate cosmic reionization, using joint spectral information on high redshift Lyman Alpha Emitters (LAE) and quasars (QSOs). Although LAEs have been proposed as reionization probes, their use is hampered by the fact their Lyα line is damped not only by intergalactic HI but also internally by dust. Our method allows to overcome such degeneracy. First, we carefully calibrate a reionization simulation with QSO absorption line experiments. Then we identify LAEs in two simulation boxes at z=5.7 and z=6.6 and we build synthetic images/spectra of a prototypical LAE. At redshift 5.7, we find that the Lyα transmissivity (T_LAE) ~ 0.25, almost independent of the halo mass. This constancy arises from the conspiracy of two effects: (i) the intrinsic Lyα line width and (ii) the infall peculiar velocity. At higher redshift, z=6.6, where the transmissivity is instead largely set by the local HI abundance and LAE transmissivity consequently increases with halo mass from 0.15 to 0.3. Although outflows are present, they are efficiently pressure-confined by infall in a small region around the LAE; hence they only marginally affect transmissivity. Finally, we cast LOS originating from background QSOs passing through foreground LAEs at different impact parameters, and compute the quasar transmissivity (T_QSO). At smaller impact parameters, d < 1 cMpc, a positive correlation between T_QSO and halo mass is found at z = 5.7, which tends to become less pronounced (i.e. flatter) at larger distances. By cross-correlating T_LAE and T_QSO, we can obtain a HI density estimate unaffected by dust. At z= 5.7, the cross-correlation is relatively weak,whereas at z = 6.6 we find a clear positive correlation. We conclude by briefly discussing the perspectives for the application of the method to existing and forthcoming data.

astro-ph.CO↗

arXiv2011

Distinctive rings in the 21 cm signal of the epoch of reionization

It is predicted that sources emitting UV radiation in the Lyman band during the epoch of reionization (EoR) showed a series of discontinuities in their Ly-alpha flux radial profile as a consequence of the thickness of the Lyman line series in the primeval intergalactic medium. Through unsaturated Wouthuysen-Field coupling, these spherical discontinuities are also present in the 21 cm emission of the neutral IGM. In this article, we study the effects these discontinuities have on the differential brightness temperature of the 21 cm signal of neutral hydrogen in a realistic setting including all other sources of fluctuations. We focus on the early phases of the EoR, and we address the question of the detectability by the planned Square Kilometre Array. Such a detection would be of great interest, because these structures could provide an unambiguous diagnostic for the cosmological origin of the signal remaining after the foreground cleaning procedure. Also, they could be used as a new type of standard rulers. We determine the differential brightness temperature of the 21 cm signal in the presence of inhomogeneous Wouthuysen-Field effect using simulations which include (hydro)dynamics and both ionizing and Lyman lines 3D radiative transfer with the code LICORICE. We find that the Lyman horizons are clearly visible on the maps and radial profiles around the first sources of our simulations, but for a limited time interval, typically Δz \approx 2 at z \sim 13. Stacking the profiles of the different sources of the simulation at a given redshift results in extending this interval to Δz \approx 4. When we take into account the implementation and design planned for the SKA (collecting area, sensitivity, resolution), we find that detection will be challenging. It may be possible with a 10 km diameter for the core, but will be difficult with the currently favored design of a 5 km core.

astro-ph.CO↗

arXiv2011

Gas inflows, star formation and metallicity evolution in galaxy pairs

It has been known since many decades that galaxy interactions can induce star formation (hereafter SF) enhancements and that one of the driving mechanisms of this enhancement is related to gas inflows into the central galaxy regions, induced by asymmetries in the stellar component, like bars. In the last years many evidences have been accumulating, showing that interacting pairs have central gas-phase metallicities lower than those of field galaxies, by {\sim} 0.2-0.3 dex on average. These diluted ISM metallicities have been explained as the result of inflows of metal-poor gas from the outer disk to the galaxy central regions. A number of questions arises: What's the timing and the duration of this dilution? How and when does the SF induced by the gas inflow enrich the circumnuclear gas with re-processed material? Is there any correlation between the timing and strength of the dilution and the timing and intensity of the SF? By means of Tree-SPH simulations of galaxy major interactions, we have studied the effect that gas inflows have on the ISM dilution, and the effect that the induced SF has, subsequently, in re-enriching the nuclear gas. In this contribution, we present the main results of this study.

astro-ph.CO↗

arXiv2010

Reionization by UV or X-ray sources

We present simulations of the 21-cm signal during the epoch of reionization. We focus on properly modeling the absorption regime in the presence of inhomogeneous Wouthuysen-Field effect and X-ray heating. We ran radiative transfer simulations for three bands in the source spectrum (Lyman, UV, and X-ray) to fully account for these processes. We find that the brightness temperature fluctuation of the 21 cm signal has an amplitude greater than 100 mK during the early reionization, up to 10 times greater than the typical amplitude of a few 10 mK obtained during the later emission phase. More importantly, we find that even a rather high contribution from QSO-like sources only damps the absorption regime without erasing it. Heating the IGM with X-ray takes time. Our results show that observations of the early reionization will probably benefit from a higher signal-to-noise value than during later stages. After analyzing the statistical properties of the signal (power spectrum and PDF) we find three diagnostics to constrain the level of X-ray, hence the nature of the first sources.

astro-ph.CO↗