Search

The XLF of AGN offers a robust tool to study the evolution and the growth of SMBHs over cosmic time. Owing to the limited area probed by X-ray surveys, optical surveys are routinely used to probe the accretion in the high redshift Universe $z\geq 3$. However, optical surveys may be incomplete because they are strongly affected by dust redenning. In this work, we derive the XLF and its evolution at high redshifts using a large sample of AGNs selected in different fields with various areas and depths covering a wide range of luminosities. Additionally, we put the tightest yet constraints on the absorption function in this redshift regime. In particular, we use more than 600 soft X-ray selected high-z sources in the Chandra Deep fields, the Chandra COSMOS Legacy survey and the XMM-XXL northern field. We derive the X-ray spectral properties for all sources via spectral fitting, using a consistent technique and model. For modeling the parametric form of the XLF and the absorption function, we use a Bayesian methodology allowing us to correctly propagate the uncertainties for the observed X-ray properties of our sources and also the absorption effects. The evolution of XLF is in agreement with a pure density evolution model similar to what is witnessed at optical wavelengths, although a luminosity dependent density evolution model cannot be securely ruled out. A large fraction ($60\%)$ of our sources are absorbed by column densities of $\rm N_H \geq 10^{23} cm^{-2} $, while $17$\% of the sources are CTK. Our results favor a scenario where both the ISM of the host and the AGN torus contribute to the obscuration. The derived BHAD is in agreement with the simulations, if one takes into account that the X-ray AGN are hosted by massive galaxies, while it differs from the one derived using JWST data. The latter could be due to the differences in the AGN and host-galaxy properties., Comment: Accepted for publication in A&A. 25 pages, 15 figures (+4 in Appendix), 2 tables (+1 in Appendix)

Recent observations have revealed an interesting active galactic nuclei (AGN) subclass that shows strong activity at large scales ($\sim1$ kpc) but weaker at small scales ($<10$ pc), suggesting a strong change in the mass accretion rate of the central engine in the past $10^{3-4}$ yr. We systematically search for such declining or fading AGN by cross-matching the SDSS type-1 AGN catalog at $z<0.4$, covering the [OIII]$\lambda5007$ emission line which is a tracer for the narrow-line region (NLR) emission, with the WISE mid-infrared (MIR) catalog covering the emissions from the dusty tori. Out of the 7,653 sources, we found 57 AGN whose bolometric luminosities estimated from the MIR band are at least one order of magnitude fainter than those estimated from the [OIII]$\lambda5007$ emission line. This luminosity declining AGN candidate population shows four important properties: 1) the past AGN activity estimated from the [OIII]$\lambda5007$ line reaches around the Eddington-limit, 2) more than 30% of the luminosity declining AGN candidates show a large absolute variability of $\Delta W1 > 0.45$ mag in the previous $\sim10$ yr at the WISE 3.4 $\mu$m band, 3) the median ratio of $\log$([NII]$\lambda6584/ \mathrm{H}\alpha\lambda6563)=-0.52$, suggesting a lower gas metallicity and/or higher ionization parameter compared to other AGN populations. 4) the second epoch spectra of the population indicate a spectral type change for 15% of the sources. This population provides insights on the possible connection between the luminosity decline which started $\sim10^{3-4}$ yr ago and the decline in the recent $10$ yr., Comment: 21 pages, 10 figures, 2 tables (1 MRT), accepted for publication in ApJ

In this work, we aimed at investigating the star formation rate of active galactic nuclei host galaxies in the early Universe. To this end, we constructed a sample of 149 luminous ($\rm L_{2-10keV} > 10^{44}\,erg\,s^{-1}$) X-ray AGNs at $\rm z \geq3.5$ selected in three fields with different depths and observed areas (Chandra COSMOS Legacy survey, XMM-XXL North and eFEDS). We built their spectral energy distributions (SED) using available multi-wavelength photometry from X-rays up to far-IR. Then, we estimated the stellar mass, M$_{*}$, and the SFR of the AGNs using the X-CIGALE SED fitting algorithm. After applying several quality criteria, we ended up with 89 high-z sources. More than half (55\%) of the X-ray sample have spectroscopic redshifts. Based on our analysis, our high-z X-ray AGNs live in galaxies with median $\rm M_{*}=5.6 \times10^{10}~M_\odot$ and $\rm SFR_{*}\approx240\,M_\odot yr^{-1}$. The majority of the high-z sources ($\sim89$\%) were found inside or above the main sequence (MS) of star-forming galaxies. Estimation of the normalised SFR, $\rm SFR_{NORM}$, defined as the ratio of the SFR of AGNs to the SFR of MS galaxies, showed that the SFR of AGNs is enhanched by a factor of $\sim 1.8$ compared to non-AGN star-forming systems. Combining our results with previous studies at lower redshifts, we confirmed that $\rm SFR_{NORM}$ does not evolve with redshift. Using the specific BHAR (i.e., $\rm L_X$ divided by $\rm M_{*}$), $\rm \lambda _{BHAR}$, that can be used as a tracer of the Eddington ratio, we found that the bulk of AGNs that lie inside or above the MS have higher specific accretion rates compared to sources below the MS. Finally, we found indications that the SFR of the most massive AGN host galaxies ($\rm log\,(M_{*}/ M_\odot) >10^{11.5-12}$) remains roughly constant as a function of M$_*$, in agreement with the SFR of MS star-forming galaxies., Comment: Accepted for publication in A&A. 19 pages, 12 figures (+6 in Appendix), 4 tables

X-rays offer a reliable method to identify Active Galactic Nuclei (AGN). However, in the high-redshift Universe X-ray AGN are poorly sampled due to their relatively low space density and the small areas covered by X-ray surveys. In addition to wide-area X-ray surveys, it is important to have deep optical data in order to locate the optical counterparts and determine their redshifts. In this work, we build a high-redshift ($z\geqslant3.5$) X-ray selected AGN sample in the XMM-XXL northern field using the most updated [0.5-2 keV] catalogue along with a plethora of new spectroscopic and multi-wavelength catalogues, including the deep optical Subaru Hyper Suprime-Cam (HSC) data reaching magnitude limits i$\sim$26 mag. We select all the spectroscopically confirmed AGN and complement this sample with high-redshift candidates that are HSC g- and r-band dropouts. To confirm the dropouts, we derive their photometric redshifts using spectral energy distribution techniques. We end up with a sample of 54 high-z sources (28 with spec-z), the largest in this field so far (almost three times larger than in previous studies), and we estimate the possible contamination and completeness. We calculate the number counts (logN-logS) in different redshift bins and compare our results with previous studies and models. We provide the strongest high-redshift AGN constraints yet at bright fluxes ($\rm f_{0.5-2~keV} > 10^{-15} erg~s^{-1}~cm^{-2}$). The samples of $z\geqslant3.5$, $z\geqslant4$ and $z\geqslant5$ are in agreement with an exponential decline model similar to that witnessed at optical wavelengths. Our work emphasizes the importance of using wide area X-ray surveys with deep optical data to uncover high redshift AGN., Comment: Accepted for publication in A&A. 16 pages, 12 figures, 3 tables

In the SSA22 field which exhibits a large-scale proto-cluster at $z=3.1$, we carried out a spectroscopic survey for Lyman Break Galaxies (LBGs) with the VLT/VIMOS and identified 78 confident LBGs at $z=2.5$--4. We stacked their spectra in the observer's frame by using a sophisticated method. Analyzing the composite spectrum, we have revealed that the large-scale proto-cluster at $z=3.1$ has a strong HI absorption dip of rest-frame equivalent width of $-1.7$ A. Another strong absorption dip found at $z=3.28$ is associated with a modestly high-density LBG peak, similar to that at $z=3.1$. We have also detected an HI transparency peak at $z=2.98$ in the composite spectrum, coincident with a void in the LBG distribution. In this paper, we also investigated the relation between LBGs, HI gas and AGNs at $z=3$--4 in the SSA22 field. Two AGNs at $z=3.353$ and 3.801 are, respectively, associated with the LBG concentration of an overdensity factor $\delta_{\rm LBG}\simeq2$ in the present statistics. Another structure at $z=3.453$ is remarkable: 20 comoving Mpc scale dense HI gas which is not associated with any apparent LBG overdensity but involving a pair of AGNs. Such structure may be a new type of the AGN-matter correlation. If the inhomogeneous structures over a comoving Gpc scale found in this paper are confirmed with sufficient statistics in the future, the SSA22 field will become a key region to test the standard cold dark matter structure formation scenario., Comment: MNRAS accepted, 24 pages, 20 figures, 8 tables

We present near-IR spectroscopy in J- and H-band for a large sample of 243 X-ray selected moderate-luminosity type-1 AGN in the COSMOS, SXDS and E-CDF-S survey fields using the multi-object spectrograph Subaru/FMOS. Our sample covers the redshift range 0.50.4 dex). We extend a recently proposed correction based on the CIV blueshift to lower luminosities and black hole masses. While our sample shows an improvement in their CIV black hole mass estimates, the deficit of high blueshift sources reduces its overall importance for moderate-luminosity AGN, compared to the most luminous quasars. In addition, we revisit luminosity correlations between $L_{\rm{bol}}$, $L_X$, $L_{\rm{[OIII]}}$, $L_{5100}$ and $L_{\rm{H}\alpha}$ and find them to be consistent with a simple empirical model, based on a small number of well-established scaling relations. We finally highlight our highest redshift AGN, CID 781, at z=4.6 which shows the lowest black hole mass ($\sim10^8$M$_\odot$) among current near-IR samples at this redshift, and is in a state of fast growth., Comment: 25 pages, 15 figures, accepted for publication in ApJS

We report discovery of 41 new high-z quasars and luminous galaxies, which were spectroscopically identified at 5.7 < z < 6.9. This is the fourth in a series of papers from the Subaru High-z Exploration of Low-Luminosity Quasars (SHELLQs) project, based on the deep multi-band imaging data collected by the Hyper Suprime-Cam (HSC) Subaru Strategic Program survey. We selected the photometric candidates by a Bayesian probabilistic algorithm, and then carried out follow-up spectroscopy with the Gran Telescopio Canarias and the Subaru Telescope. Combined with the sample presented in the previous papers, we have now spectroscopically identified 137 extremely-red HSC sources over about 650 deg2, which include 64 high-z quasars, 24 high-z luminous galaxies, 6 [O III] emitters at z ~ 0.8, and 43 Galactic cool dwarfs (low-mass stars and brown dwarfs). The new quasars span the luminosity range from M1450 ~ -26 to -22 mag, and continue to populate a few magnitude lower luminosities than have been probed by previous wide-field surveys. In a companion paper, we derive the quasar luminosity function at z ~ 6 over an unprecedentedly wide range of M1450 ~ -28 to -21 mag, exploiting the SHELLQs and other survey outcomes., Comment: ApJS in press. arXiv admin note: substantial text overlap with arXiv:1704.05854

We investigate the X-ray active galactic nucleus (AGN) properties of millimeter galaxies in the Great Observatories Origins Deep Survey South (GOODS-S) field detected with the Atacama Large Millimeter/submillimeter Array (ALMA), by utilizing the Chandra 7-Ms data, the deepest X-ray survey to date. Our millimeter galaxy sample comes from the ASAGAO survey covering 26 arcmin$^2$ (12 sources at a 1.2-mm flux-density limit of $\approx$0.6 mJy), supplemented by the deeper but narrower 1.3-mm survey of a part of the ASAGAO field by Dunlop et al.\ (2017). Fourteen out of the total 25 millimeter galaxies have Chandra counterparts. The observed AGN fractions at $z=1.5-3$ is found to be 90$^{+8}_{-19}$\% and $57^{+23}_{-25}$\% for the ultra/luminous infrared galaxies with $\log L_{\rm IR}/L_{\odot} = 12-12.8$ and $\log L_{\rm IR}/L_{\odot} = 11.5-12$, respectively. The majority ($\sim$2/3) of the ALMA and/or Herschel detected X-ray AGNs at $z=1.5-3$ appear to be star-formation dominant populations, having $L_{\rm X}$/ $L_{\rm IR}$ ratios smaller than the "simultaneous evolution" value expected from the local black-hole mass to stellar mass ($M_{\rm BH}$-$M_*$) relation. On the basis of the $L_{\rm X}$ and stellar mass relation, we infer that a large fraction of star-forming galaxies at $z=1.5-3$ have black hole masses smaller than those expected from the local $M_{\rm BH}$-$M_*$ relation. This is opposite to previous reports on luminous AGN at same redshifts detected in wider and shallower surveys, which are subject to selection biases against lower luminosity AGN. Our results are consistent with an evolutionary scenario that star formation occurs first, and an AGN-dominant phase follows later, in objects finally evolving into galaxies with classical bulges., Comment: 14 pages including 5 figures and 1 table, accepted for publication in ApJ

We present the result of our spectroscopic follow-up observation for faint quasar candidates at z~5 in a part of the Canada-France-Hawaii Telescope Legacy Survey wide field. We select nine photometric candidates and identify three z~5 faint quasars, one z~4 faint quasar, and a late-type star. Since two faint quasar spectra show Civ emission line without suffering from a heavy atmospheric absorption, we estimate the black hole mass (M$_{BH}$) and Eddington ratio (L/L$_{Edd}$) of them. The inferred log M$_{BH}$ are 9.04+/-0.14 and 8.53+/-0.20, respectively. In addition, the inferred log (L/L$_{Edd}$) are -1.00+/-0.15 and -0.42+/-0.22, respectively. If we adopt that L/L$_{Edd}$= constant or $\propto$ (1+z)^2, the seed black hole masses (M$_{seed}$) of our z~5 faint quasars are expected to be >10^5 M$_\odot$ in most cases. We also compare the observational results with a mass accretion model where angular momentum is lost due to supernova explosions (Kawakatu & Wada 2008). Accordingly, M$_{BH}$ of the z~5 faint quasars in our sample can be explained even if M$_{seed}$ is ~10^3M$_\odot$. Since z~6 luminous qusars and our z~5 faint quasars are not on the same evolutionary track, z~6 luminous quasars and our z~5 quasars are not the same populations but different populations, due to the difference of a period of the mass supply from host galaxies. Furthermore, we confirm that one can explain M$_{BH}$ of z~6 luminous quasars and our z~5 faint quasars even if their seed black holes of them are formed at z~7., Comment: 16 pages, 8 figures, Accepted for publication in The Astrophysical Journal

We discuss images of the star clusters GLIMPSE C01 (GC01) and GLIMPSE C02 (GC02) that were recorded with the Subaru IRCS. Distortions in the wavefront were corrected with the RAVEN adaptive optics (AO) science demonstrator, allowing individual stars in the central regions of both clusters -- where the fractional contamination from non-cluster objects is lowest -- to be imaged. In addition to J, H, and K' images, both clusters were observed through a narrow-band filter centered near 3.05um; GC01 was also observed through two other narrow-band filters that sample longer wavelengths. Stars in the narrow-band images have a FWHM that is close to the telescope diffraction limit, demonstrating that open loop AO systems like RAVEN can deliver exceptional image quality. The near-infrared color magnitude diagram of GC01 is smeared by non-uniform extinction with a dispersion +/- 0.13 magnitudes in A_K. The Red Clump is identified in the K luminosity function (LF) of GC01, and a distance modulus of 13.6 is found. The K LF of GC01 is consistent with a system that is dominated by stars with an age > 1 Gyr. As for GC02, the K LF is flat for K > 16, and the absence of a sub-giant branch argues against an old age if the cluster is at a distance of ~ 7 kpc. Archival SPITZER [3.6] and [4.5] images of the clusters are also examined, and the red giant branch-tip is identified., Comment: To appear in the Astronomical Journal

Aims. We present a spectroscopic study of the properties of 64 Balmer break galaxies that show signs of star formation. The studied sample of star-forming galaxies spans a redshift range from 0.094 to 1.475 with stellar masses in the range 10$^{8}-$10$^{12}$ $M_{\odot}$. The sample also includes eight broad emission line galaxies with redshifts between 1.5 $

We have conducted 1.1 mm ALMA observations of a contiguous $105'' \times 50''$ or 1.5 arcmin$^2$ window in the SXDF-UDS-CANDELS. We achieved a 5$\sigma$ sensitivity of 0.28 mJy, providing a flat sensus of dusty star-forming galaxies with $L_{\rm IR} \sim6\times10^{11}$ $L_\odot$ (for $T_{\rm dust}$ =40K) up to $z\sim10$ thanks to the negative K-correction at this wavelength. We detected 5 brightest sources (S/N$>$6) and 18 low-significance sources (5$>$S/N$>$4; these may contain spurious detections, though). One of the 5 brightest ALMA sources ($S_{\rm 1.1mm} = 0.84 \pm 0.09$ mJy) is extremely faint in the WFC3 and VLT/HAWK-I images, demonstrating that a contiguous ALMA imaging survey is able to uncover a faint dust-obscured population that is invisible in deep optical/near-infrared surveys. We found a possible [CII]-line emitter at $z=5.955$ or a low-$z$ CO emitting galaxy within the field, which may allow us to constrain the [CII] and/or the CO luminosity functions across the history of the universe., Comment: 4 pages, 2 figures, 1 table, to appear in the proceedings of IAU Symposium 319 "Galaxies at High Redshift and Their Evolution over Cosmic Time", eds. S. Kaviraj & H. Ferguson

Near-infrared (NIR) spectra that have an angular resolution of ~ 0.15 arcsec are used to examine the stellar content of the central regions of the nearby elliptical galaxy Maffei 1. The spectra were recorded at the Subaru Telescope, with wavefront distortions corrected by the RAVEN Multi-Object Adaptive Optics science demonstrator. The Ballick-Ramsey C_2 absorption bandhead near 1.76 microns is detected, and models in which 10 - 20% of the light near 1.8 microns originates from stars of spectral type C5 reproduce this feature. Archival NIR and mid-infrared images are also used to probe the structural and photometric properties of the galaxy. Comparisons with models suggest that an intermediate age population dominates the spectral energy distribution between 1 and 5 microns near the galaxy center. This is consistent not only with the presence of C stars, but also with the large HBeta index that has been measured previously for Maffei 1. The J-K color is more-or-less constant within 15 arcsec of the galaxy center, suggesting that the brightest red stars are well-mixed in this area., Comment: To appear in The Astrophysical Journal

In order to investigate the origin of quasars, we estimate the bias factor for low-luminosity quasars at high redshift for the first time. In this study, we use the two-point angular cross-correlation function (CCF) for both low-luminosity quasars at $-24

We construct the rest-frame 2--10 keV intrinsic X-ray luminosity function of Active Galactic Nuclei (AGNs) from a combination of X-ray surveys from the all-sky Swift BAT survey to the Chandra Deep Field-South. We use ~3200 AGNs in our analysis, which covers six orders of magnitude in flux. The inclusion of the XMM and Chandra COSMOS data has allowed us to investigate the detailed behavior of the XLF and evolution. In deriving our XLF, we take into account realistic AGN spectrum templates, absorption corrections, and probability density distributions in photometric redshift. We present an analytical expression for the overall behavior of the XLF in terms of the luminosity-dependent density evolution, smoothed two power-law expressions in 11 redshift shells, three-segment power-law expression of the number density evolution in four luminosity classes, and binned XLF. We observe a sudden flattening of the low luminosity end slope of the XLF slope at z>~0.6. Detailed structures of the AGN downsizing have been also revealed, where the number density curves have two clear breaks at all luminosity classes above log LX>43. The two break structure is suggestive of two-phase AGN evolution, consisting of major merger triggering and secular processes., Comment: 39 Pages, 9 figures. ApJ in press

We present an investigation into how well the properties of the accretion flow onto a supermassive black hole may be coupled to those of the overlying hot corona. To do so, we specifically measure the characteristic spectral index, Gamma, of a power-law energy distribution, over an energy range of 2 to 10 keV, for X-ray selected, broad-lined radio-quiet AGN up to z~2 in COSMOS and E-CDF-S. We test the previously reported dependence between Gamma and black hole mass, FWHM and Eddington ratio using a sample of AGN covering a broad range in these parameters based on both the Mg ii and H-alpha emission lines with the later afforded by recent near infrared spectroscopic observations using Subaru/FMOS. We calculate the Eddington ratios, lambda_Edd, for sources where a bolometric luminosity (L_Bol) has been presented in the literature, based on SED fitting, or, for sources where these data do not exist, we calculate L_Bol using a bolometric correction to the LX, derived from a relationship between the bolometric correction, and LX/L3000. From a sample of 69 X-ray bright sources (>250 counts), where Gamma can be measured with greatest precision, with an estimate of L_Bol, we find a statistically significant correlation between Gamma and lambda_Edd, which is highly significant with a chance probability of 6.59x10^-8. A statistically significant correlation between Gamma and the FWHM of the optical lines is confirmed, but at lower significance than with lambda_Edd indicating that lambda_Edd is the key parameter driving conditions in the corona. Linear regression analysis reveals that Gamma=(0.32+/-0.05)log10 lambda_Edd+(2.27+/-0.06) and Gamma=(-0.69+/-0.11)log10(FWHM/km/s)+(4.44+/-0.42). Our results on Gamma-lambda_Edd are in very good agreement with previous results. (ABRIDGED)

In order to investigate the growth of super-massive black holes (SMBHs), we construct the black hole mass function (BHMF) and Eddington ratio distribution function (ERDF) of X-ray-selected broad-line AGNs at z~1.4 in the Subaru XMM-Newton Deep Survey field. In this redshift range, a significant part of the accretion growth of SMBHs is thought to be taking place. Black hole masses of X-ray-selected broad-line AGNs are estimated using the width of the broad MgII line and the 3000A monochromatic luminosity. We supplement the MgII FWHM values with the Ha FWHM obtained from our NIR spectroscopic survey. Using the black hole masses of broad-line AGNs at redshifts between 1.18 and 1.68, the binned broad-line AGN BHMF and ERDF are calculated using the Vmax method. To properly account for selection effects that impact the binned estimates, we derive the corrected broad-line AGN BHMF and ERDF by applying the Maximum Likelihood method, assuming that the ERDF is constant regardless of the black hole mass. We do not correct for the non-negligible uncertainties in virial BH mass estimates. If we compare the corrected broad-line AGN BHMF with that in the local Universe, the corrected BHMF at z~1.4 has a higher number density above 10^8 Msolar but a lower number density below that mass range. The evolution may be indicative of a down-sizing trend of accretion activity among the SMBH population. The evolution of broad-line AGN ERDF from z=1.4 to 0 indicates that the fraction of broad-line AGNs with accretion rate close to the Eddington-limit is higher at higher redshifts., Comment: 23 pages with 18 figures. Accepted for publication in ApJ

We investigate the properties (e.g. star formation rate, dust attentuation, stellar mass and metallicity) of a sample of infrared luminous galaxies at z \sim 1 via near-IR spectroscopy with Subaru-FMOS. Our sample consists of Herschel SPIRE and Spitzer MIPS selected sources in the COSMOS field with photometric redshifts in the range 0.7 < z-phot < 1.8, which have been targeted in 2 pointings (0.5 sq. deg.) with FMOS. We find a modest success rate for emission line detections, with candidate H{\alpha} emission lines detected for 57 of 168 SPIRE sources (34 per cent). By stacking the near-IR spectra we directly measure the mean Balmer decrement for the H{\alpha} and H{\beta} lines, finding a value of = 0.51\pm0.27 for = 10^12 Lsol sources at = 1.36. By comparing star formation rates estimated from the IR and from the dust uncorrected H{\alpha} line we find a strong relationship between dust attenuation and star formation rate. This relation is broadly consistent with that previously seen in star-forming galaxies at z ~ 0.1. Finally, we investigate the metallicity via the N2 ratio, finding that z ~ 1 IR-selected sources are indistinguishable from the local mass-metallicity relation. We also find a strong correlation between dust attentuation and metallicity, with the most metal-rich IR-sources experiencing the largest levels of dust attenuation., Comment: 13 pages, 7 figures, MNRAS accepted

The FIBRE-pac (FMOS image-based reduction package) is an IRAF-based reduction tool for the fiber multiple-object spectrograph (FMOS) of the Subaru telescope. To reduce FMOS images, a number of special techniques are necessary because each image contains about 200 separate spectra with airglow emission lines variable in spatial and time domains, and with complicated throughput patterns for the airglow masks. In spite of these features, almost all of the reduction processes except for a few steps are carried out automatically by scripts in text format making it easy to check the commands step by step. Wavelength- and flux-calibrated images together with their noise maps are obtained using this reduction package., Comment: Accepted for publication in PASJ. 9 pages, 14 figures

XMM and Chandra opened a new area for the study of clusters of galaxies. Not only for cluster physics but also, for the detection of faint and distant clusters that were inaccessible with previous missions. This article presents 66 spectroscopically confirmed clusters (0.05

(Abridged) Deep NB359 imaging with Subaru by Iwata et al. have detected surprisingly strong Lyman continuum (LyC; ~900A in the rest-frame) from some LAEs at z=3.1. However, the redshifts might be misidentified due to a narrow wavelength coverage in previous spectroscopy. We here present new deep spectroscopy covering the observed 4,000-7,000A with VLT/VIMOS and Subaru/FOCAS of 8 LAEs detected in NB359. All the 8 objects have only one detectable emission line around 4,970A which is most likely to be Ly-A at z=3.1, and thus, the objects are certainly LAEs at the redshift. However, 5 of them show a ~0.''8 spatial offset between the Ly-A emission and the source detected in NB359. No indications of the redshifts of the NB359 sources are found although it is statistically difficult that all the 5 LAEs have a foreground object accounting for the NB359 flux. The rest 3 LAEs show no significant offset from the NB359 position. Therefore, they are truly LyC emitting LAEs at z=3.1. We also examine the stellar population which simultaneously accounts for the strength of the LyC and the spectral slope of non-ionizing ultraviolet of the LAEs. We consider the latest statistics of Lyman limit systems to estimate the LyC optical depth in the IGM and an additional contribution of the bound-free LyC from photo-ionized nebulae to the LyC emissivity. As a result, we find that stellar populations with metallicity Z>=1/50Z_sun can explain the observed LyC strength only with a very top-heavy initial mass function (IMF; ~50 M_sun). However, the critical metallicity for such an IMF is expected to be much lower. A very young (~1 Myr) and massive (~100 M_sun) extremely metal-poor (Z<=5e-4Z_sun) or metal-free (so-called Population III) stellar population can reproduce the observed LyC strength. The required mass fraction of such `primordial' stellar population is ~1--10% in total stellar mass of the LAEs., Comment: MNRAS accepted (18 pages, 14 figures, 5 tables)