SILVERRUSH. II. First catalogs and properties of ∼2000 Lyα emitters and blobs at z ∼ 6–7 identified over the 14–21 deg2 sky

SILVERRUSH. II. First catalogs and properties of ∼2000 Lyα emitters and blobs at z ∼ 6–7... Abstract We present an unprecedentedly large catalog consisting of 2230 ≳ L* Lyα emitters (LAEs) at z = 5.7 and 6.6 on the 13.8 and 21.2 deg2 sky, respectively, that are identified by the SILVERRUSH program with the first narrow-band imaging data of the Hyper Suprime-Cam (HSC) survey. We confirm that the LAE catalog is reliable on the basis of 96 LAEs whose spectroscopic redshifts are already determined by this program and previous studies. This catalogue is also available online. Based on this catalogue, we derive the rest-frame Lyα equivalent-width distributions of LAEs at z ≃ 5.7–6.6 that are reasonably explained by the exponential profiles with scale lengths of ≃ 120–170 Å, showing no significant evolution from z ≃ 5.7 to z ≃ 6.6. We find that 275 LAEs with large equivalent widths (LEWs) of >240 Å are candidates of young metal poor galaxies and AGNs. We also find that the fraction of LEW LAEs to all LAEs is 4% and 21% at z ≃ 5.7 and z ≃ 6.6, respectively. Our LAE catalog includes 11 Lyα blobs (LABs) that are LAEs with spatially extended Lyα emission with a profile that is clearly distinguished from those of stellar objects at the ≳3σ level. The number density of the LABs at z = 6–7 is ∼10−7–10−6 Mpc−3, being ∼10–100 times lower than those claimed for LABs at z ≃ 2–3, suggestive of disappearing LABs at z ≳ 6, albeit with the different selection methods and criteria for the low and high-z LABs. 1 Introduction Lyα emitters (LAEs) are one of the important populations of high-z star-forming galaxies in the paradigm of galaxy formation and evolution. Such galaxies are thought to be typically young (an order of 100 Myr; e.g., Finkelstein et al. 2007; Gawiser et al. 2007), compact (an effective radius of <1 kpc; e.g., Taniguchi et al. 2009; Bond et al. 2012), less massive (a stellar mass of 108–109 M⊙; e.g., Ono et al. 2010; Guaita et al. 2011), metal-poor (≃0.1 of the solar metallicity; e.g., Nakajima et al. 2012, 2013; Nakajima & Ouchi 2014; Kojima et al. 2017), less dusty than Lyman break galaxies (e.g., Blanc et al. 2011; Kusakabe et al. 2015), and possible progenitors of Milky Way mass galaxies (e.g., Dressler et al. 2011). In addition, LAEs are used to probe the cosmic reionizaiton, because ionizing photons that escape from the large number of massive stars formed in LAEs contribute to the ionization of the intergalactic medium (IGM; e.g., Rhoads & Malhotra 2001; Malhotra & Rhoads 2006; Shimasaku et al. 2006; Kashikawa et al. 2006, 2011; Ouchi et al. 2008, 2010; Cowie et al. 2010; Hu et al. 2010; Shibuya et al. 2012; Konno et al. 2014; Matthee et al. 2015; Ota et al. 2017; Zheng et al. 2017). LAEs have been surveyed by imaging observations with dedicated narrow-band (NB) filters for a prominent redshifted Lyα emission (e.g., Ajiki et al. 2002; Malhotra & Rhoads 2004; Kodaira et al. 2003; Taniguchi et al. 2005; Gronwall et al. 2007; Erb et al. 2011; Ciardullo et al. 2012). In a large LAE sample constructed by the NB observations, two rare Lyα-emitting populations have been identified: large equivalent width (LEW) LAEs, and spatially extended Lyα LAEs, Lyα blobs (LABs). LEW LAEs are objects with a large Lyα equivalent width (EW) of ≳240 Å which are not reproduced with the normal Salpeter (1955) stellar initial mass function (e.g., Malhotra & Rhoads 2002). Such an LEW is expected to have originated from complicated physical processes such as (i) photoionization by young and/or low-metallicity star-formation, (ii) photoionization by active galactic nucleus (AGN), (iii) photoionization by external UV sources (QSO fluorescence), (iv) collisional excitation due to strong outflows (shock heating), (v) collisional excitation due to gas inflows (gravitational cooling), and (vi) clumpy ISM (see e.g., Hashimoto et al. 2017). The highly complex radiative transfer of Lyα in the interstellar medium (ISM) makes it difficult to understand the Lyα emitting mechanism (Neufeld 1991; Hansen & Oh 2006; Finkelstein et al. 2008; Laursen et al. 2009, 2013; Laursen & Sommer-Larsen 2007; Zheng et al. 2010; Yajima et al. 2014; Duval et al. 2014; Zheng & Wallace 2014). LABs are spatially extended Lyα gaseous nebulae in the high-z universe (e.g., Steidel et al. 2000; Matsuda et al. 2004, 2009, 2011; Prescott et al. 2009, 2012a, 2012b, 2013, 2015; Cantalupo et al. 2014; Arrigoni Battaia et al. 2015a, 2015b; Hennawi et al. 2015; Cai et al. 2017). The origins of LABs (LAEs with a diameter ≃20–400 kpc) are also explained by several mechanisms: (1) resonant scattering of Lyα photons emitted from central sources in dense and extended neutral hydrogen clouds (e.g., Hayes et al. 2011), (2) cooling radiation from gravitationally heated gas in collapsed halos (e.g., Haiman et al. 2000), (3) shock heating by galactic superwind originating from starbursts and/or AGN activity (e.g., Taniguchi & Shioya 2000), (4) galaxy major mergers (e.g., Yajima et al. 2013), and (5) photoionization by external UV sources (QSO fluorescence; e.g., Cantalupo et al. 2005). Moreover, LABs have often been discovered in over-dense regions at z ≃ 2–3 (e.g., Yang et al. 2009, 2010; Matsuda et al. 2011). Thus, such LABs could be closely related to the galaxy environments, and might be linkd to the formation mechanisms of central massive galaxies in galaxy protoclusters. During the previous decades, Suprime-Cam (SCam) on the Subaru telescope has led the world on identifying such rare Lyα-emitting populations at z ≳ 6 (LEW LAEs; e.g., Nagao et al. 2008; Kashikawa et al. 2012; LABs; e.g., Ouchi et al. 2009; Sobral et al. 2015). However, the formation mechanisms of these rare Lyα-emitting populations are still controversial due to the small statistics. While LEW LAEs and LABs at z ≃ 2–5 have been studied intensively with a sample of ≳ 100 sources, only a few sources have been found so far at z ≳ 6. Large-area NB data are required to carry out a statistical study on LEW LAEs and LABs at z ≳ 6. In 2014 March, the Subaru telescope started a large-area NB survey using a new wide field of view (FoV) camera, the Hyper Suprime-Cam (HSC) as part of a Subaru strategic program (SSP: Aihara et al. 2018a). In the five-year project, HSC, equipped with four NB filters of NB387, NB816, NB921, and NB101, will survey for LAEs at z ≃ 2.2, 5.7, 6.6, and 7.3, respectively. The HSC-SSP NB survey data consist of two layers; Ultradeep (UD), covering two fields (UD-COSMOS, UD-SXDS), and Deep (D), covering four fields (D-COSMOS, D-SXDS, D-DEEP2-3, D-ELAIS-N1). The NB816, NB921, and NB101 images will be taken for the UD fields. The NB387, NB816, and NB921 observations will be conducted in 15 HSC-pointing D fields. Using the large HSC NB data complemented by optical and near-infrared (NIR) spectroscopic observations, we launch a research project for Lyα-emitting objects: the Systematic Identification of LAEs for Visible Exploration and Reionization Research Using Subaru HSC (SILVERRUSH). The large LAE samples provided by SILVERRUSH enable us to investigate, e.g., LAE clustering (Ouchi et al. 2018), LEW LAEs and LABs (this work), the spectroscopic properties of bright LAEs (Shibuya et al. 2018), Lyα luminosity functions (Konno et al. 2018), and LAE overdensity (R. Higuchi et al. in preparation). The LAE survey strategy is given by Ouchi et al. (2018). This program is one of the twin programs. Another program is the study for dropouts, the Great Optically Luminous Dropout Research Using Subaru HSC (GOLDRUSH), which is detailed in Ono et al. (2018), Harikane et al. (2018), and Toshikawa et al. (2018). This is the second paper in the SILVERRUSH project. In this paper, we present LAE selection processes and machine-readable catalogs of the LAE candidates at z ≃ 5.7–6.6. Using the large LAE sample obtained from the first HSC NB data, we examine the redshift evolutions of Lyα EW distributions and LAB number density. This paper has the following structure. In section 2, we describe the details of the SSP HSC data. Section 3 presents the LAE selection processes. In section 4, we check the reliability of our LAE selection. Section 5 presents Lyα EW distributions and LABs at z ≃ 6–7. In section 6, we discuss the physical origins of LEW LAEs and LABs. We summarize our findings in section 7. Throughout this page, we adopt the concordance cosmology with (Ωm, ΩΛ, h) = (0.3, 0.7, 0.7) (Planck Collaboration 2016). All magnitudes are given in the AB system (Oke & Gunn 1983). 2 HSC-SSP imaging data We use the HSC-SSP S16A data products of g, r, i, z, and y broad-band (BB: Kawanomoto 2017), NB921, and NB816 (Ouchi et al. 2018) images that were obtained between 2014–2016. It should be noted that this HSC-SSP S16A data set is significantly larger than that of the first-data release in Aihara et al. (2018b). The NB921 (NB816) filter has a central wavelength of λc = 9215 Å (8177 Å) and an FWHM of Δλ = 135 Å (113 Å), all of which are the area-weighted mean values. The NB921 and NB816 filters trace the redshifted Lyα emission lines at z = 6.580 ± 0.056 and z = 5.726 ± 0.046, respectively. The NB filter transmission curves are shown in figure 1. The central wavelength, FWHM, and the bandpass shape for these NB filters are almost uniform over the HSC FoV. The deviation of the λc and FWHM values are typically within ≃0.3% and ≃10%, respectively. Thus, we use the area-weighted mean transmission curves in this study. The detailed specifications of these NB filters are given in Ouchi et al. (2018). Fig. 1. View largeDownload slide Filter transmission curves of the NB and BB filters. The red and blue curves represent the NB921 and NB816 filters, respectively. The red and blue ticks show the NB central wavelengths with the same color coding as for the NB filter transmission curves. The black solid curves indicate the i-, z-, and y-band filters, from left to right. The gray line denotes the OH sky lines. The bandpass of these NB and BB filters corresponds to the area-weighted mean transmission curves.1 The transmission curves are derived by taking into account (1) the quantum efficiency of CCD, the transmittance of (2) the dewar window and (3) the HSC primary focus unit (POpt2), (4) the reflectivity of the primary mirror, and (5) the sky transparency (see Aihara et al. 2018b). The upper x-axis corresponds to the redshift of Lyα. (Color online) Fig. 1. View largeDownload slide Filter transmission curves of the NB and BB filters. The red and blue curves represent the NB921 and NB816 filters, respectively. The red and blue ticks show the NB central wavelengths with the same color coding as for the NB filter transmission curves. The black solid curves indicate the i-, z-, and y-band filters, from left to right. The gray line denotes the OH sky lines. The bandpass of these NB and BB filters corresponds to the area-weighted mean transmission curves.1 The transmission curves are derived by taking into account (1) the quantum efficiency of CCD, the transmittance of (2) the dewar window and (3) the HSC primary focus unit (POpt2), (4) the reflectivity of the primary mirror, and (5) the sky transparency (see Aihara et al. 2018b). The upper x-axis corresponds to the redshift of Lyα. (Color online) Table 1 summarizes the survey areas, exposure time, and depth of the HSC-SSP S16A NB data. The current HSC-SSP S16A NB data covers UD-COSMOS, UD-SXDS, D-COSMOS, D-DEEP2-3, and D-ELAIS-N1 for z ≃ 6.6, and UD-COSMOS, UD-SXDS, D-DEEP2-3, and D-ELAIS-N1 for z ≃ 5.7. The effective survey areas of the NB921 and NB816 images are 21.2 and 13.8 arcmin2, corresponding to survey volumes of ≃1.9 × 107 and ≃1.2 × 107 Mpc3, respectively. The area of these HSC NB fields are covered by the observations of all the BB filters. The typical limiting magnitudes of BB filters are g ≃ 26.9, r ≃ 26.5, r ≃ 26.3, z ≃ 25.7, and y ≃ 25.0 (g ≃ 26.6, r ≃ 26.1, r ≃ 25.9, z ≃ 25.2, and y ≃ 24.4) in a 1$${^{\prime\prime}_{.}}$$5 aperture at 5σ for the UD(D) fields. The FWHM size of the point spread function in the HSC images is typically ≃0$${^{\prime\prime}_{.}}$$8 (Aihara et al. 2018b). Table 1. Properties of the HSC-SSP S16A NB data.* Field  RA  Dec  Area  Texp  mlim(5σ, 1$${^{\prime\prime}_{.}}$$5ϕ)  NLAE,ALL  NLAE,F    (J2000.0)  (J2000.0)  (deg2)  (hr)  (mag)      (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  NB921 (z ≃ 6.6)  UD-COSMOS  10h00m28s  +02°12΄21″  2.05  11.25  25.6  338  116  UD-SXDS  02h18m00s  −05°00΄00″  2.02  7.25  25.5  58  23  D-COSMOS  10h00m60s  +02°13΄53″  5.31  2.75  25.3  244†  47†  D-DEEP2-3  23h30m22s  −00°44΄38″  5.76  1.00  24.9  164  35  D-ELAIS-N1  16h10m00s  +54°17΄51″  6.08  1.75  25.3  349  48  Total  —  —  21.2  24.00  —  1153  269  NB816 (z ≃ 5.7)  UD-COSMOS  10h00m28s  +02°12΄21″  1.97  5.50  25.7  201  176  UD-SXDS  02h18m00s  −05°00΄00″  1.93  3.75  25.5  224  188  D-DEEP2-3  23h30m22s  −00°44΄38″  4.37  1.00  25.2  423  282  D-ELAIS-N1  16h10m00s  +54°17΄51″  5.56  1.00  25.3  229  130  Total  —  —  13.8  11.25  —  1077  776  Field  RA  Dec  Area  Texp  mlim(5σ, 1$${^{\prime\prime}_{.}}$$5ϕ)  NLAE,ALL  NLAE,F    (J2000.0)  (J2000.0)  (deg2)  (hr)  (mag)      (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  NB921 (z ≃ 6.6)  UD-COSMOS  10h00m28s  +02°12΄21″  2.05  11.25  25.6  338  116  UD-SXDS  02h18m00s  −05°00΄00″  2.02  7.25  25.5  58  23  D-COSMOS  10h00m60s  +02°13΄53″  5.31  2.75  25.3  244†  47†  D-DEEP2-3  23h30m22s  −00°44΄38″  5.76  1.00  24.9  164  35  D-ELAIS-N1  16h10m00s  +54°17΄51″  6.08  1.75  25.3  349  48  Total  —  —  21.2  24.00  —  1153  269  NB816 (z ≃ 5.7)  UD-COSMOS  10h00m28s  +02°12΄21″  1.97  5.50  25.7  201  176  UD-SXDS  02h18m00s  −05°00΄00″  1.93  3.75  25.5  224  188  D-DEEP2-3  23h30m22s  −00°44΄38″  4.37  1.00  25.2  423  282  D-ELAIS-N1  16h10m00s  +54°17΄51″  5.56  1.00  25.3  229  130  Total  —  —  13.8  11.25  —  1077  776  *(1) Field. (2) Right ascension. (3) Declination. (4) Survey area with the HSC SQL parameters in table 2. (5) Total exposure time of the NB imaging observation. (6) Limiting magnitude of the NB image defined by a 5σ sky noise in a 1$${^{\prime\prime}_{.}}$$5 diameter circular aperture. (7) Number of the LAE candidates in the ALL (unforced + forced) catalog. (8) Number of the LAE candidates in the forced catalog. †This value of NLAE,ALL (NLAE,F) includes 30 (7) LAEs selected in UD-COSMOS. View Large The HSC images were reduced with the HSC pipeline, hscPipe 4.0.2 (Bosch et al. 2018), which is a code from the Large Synoptic Survey Telescope (LSST) software pipeline (Ivezic et al. 2008; Axelrod et al. 2010; Jurić et al. 2015). The HSC pipeline performs CCD-by-CCD reduction, calibration for astrometry, and photometric zero-point determination. The pipeline then conducts mosaic-stacking that combines reduced CCD images into a large coadd image, and creates source catalogs by detecting and measuring sources on the coadd images. The photometric calibration is carried out with the PanSTARRS1 processing version 2 imaging survey data (Magnier et al. 2013; Schlafly et al. 2012; Tonry et al. 2012). The details of the HSC-SSP survey, data reduction, and source detection and photometric catalog construction are provided in Aihara et al. (2018a, 2018b) and Bosch et al. (2018). In the HSC images, source detection and photometry were carried out via two methods: unforced and forced. The unforced photometry is a method that performs measurements of coordinates, shapes, and fluxes individually in each band image for an object. The forced photometry is a method that carries out photometry by fixing centroid and shape determined in a reference band and applying them to all the other bands. The algorithm of the forced detection and photometry is similar to the double-image mode of SExtractor (Bertin & Arnouts 1996) which is used in most of the previous studies for high-z galaxies. According to which of them depends on magnitudes, S/N, positions, and profiles for detected sources, either the BB or the NB filter is regarded as a reference band. For merging the catalogs of each band, the object-matching radius is not a specific value; it depends on the area of a region with a >5σ sky noise level. We refer the detailed algorithm to choose the reference filter and filter priority to Bosch et al. (2018). In the hscPipe detection and photometry, an NB filter is basically chosen as a reference band for the NB-bright and BB-faint sources such as LAEs. However, a BB filter is used as a reference band in the case that sources are bright in the BB image. The current version of hscPipe has not implemented the NB-reference forced photometry for BB-bright sources. In this specification, there is a possibility that we miss BB-bright sources with a spatial offset between centroids of BB and NB by using only the forced photometry. Thus, we combine the unforced or forced photometry for BB − NB colors to identify such BB-bright objects with a spatial offset between centroids of BB and NB (e.g., Shibuya et al. 2014a). See section 3 for details of the LAE selection criteria. We use cmodel magnitudes for estimating total magnitudes of sources. The cmodel magnitude is a weighted combination of exponential and de Vaucouleurs fits to the light profiles of each object. The detailed algorithm of the cmodel photometry are presented in Bosch et al. (2018). To measure the S/N values for source detections, we use 1$${^{\prime\prime}_{.}}$$5-diameter aperture magnitudes. 3 LAE selection Using the HSC data, we perform a selection for LAEs at z ≃ 6.6 and ≃ 5.7. Basically, we select objects showing a significant flux excess in the NB images and a spectral break at the wavelength of redshifted Lyα emission. In this study, we create two LAE catalogs: an HSC LAE ALL (forced+unforced) catalog and an HSC LAE forced catalog. The HSC LAE ALL catalog is constructed using a combination of the forced and unforced photometry. We use this HSC LAE ALL catalog for identifying objects with a spatial offset between centroids of BB and NB (see section 2). On the other hand, the HSC LAE forced catalog consists of LAEs that meet only the selection criteria of the forced photometry. We use this HSC LAE forced catalog for statistical studies of LAEs [e.g., Lyα luminosity functions (LFs)]. The HSC LAE forced catalog is a subsample of the ALL one. Figure 2 shows the flow chart of the LAE selection process. We carry out the following processes: (1) SQL selection, (2) visual inspections of the object images, (3) rejections of variable and moving objects with the multi-epoch images, and (4) forced selection. The details are described as below. SQLselection: We retrieve detection and photometric catalogs from postgreSQL database tables. Using SQL scripts, we select objects meeting the following criteria of (i) magnitude and color selections and (ii) hscPipe parameters and flags. Magnitude and color selection: To identify objects with an NB magnitude excess in the HSC catalog, we apply magnitude and color selection criteria that are similar to e.g., Ouchi et al. (2008, 2010):   \begin{eqnarray} &&{\mathit {NB921}^\mathrm{ap}_\mathrm{frc} < \mathit {NB921}_{5\sigma }} \nonumber \\ && \&\&\ \left({\it g}_\mathrm{frc} > {\it g}_{3\sigma } \,{\|}\, {\it g}^\mathrm{ap}_\mathrm{frc} > {\it g}_{3\sigma } \right) \nonumber \\ && \&\&\ \left({\it r}_\mathrm{frc} > {\it r}_{3\sigma } \,{\|}\, {\it r}^\mathrm{ap}_\mathrm{frc} > {\it r}_{3\sigma } \right) \nonumber \\ && \&\&\ \left( {\it z}_\mathrm{frc} \!-\! {\it NB921}_\mathrm{frc} \!>\! 1.0 \,{\|}\, {\it z}_\mathrm{unf} \!-\! {\it NB921}_\mathrm{unf} \!>\! 1.0 \right) \nonumber \\ && \&\&\ \left\lbrace [ ({\it z}_\mathrm{frc} < {\it z}_{3\sigma } \,{\|}\, {\it z}^\mathrm{ap}_\mathrm{frc} < {\it z}_{3\sigma }) \right. \nonumber \\ &&\qquad\quad \&\&\ ({\it i}_\mathrm{frc} - {\it z}_\mathrm{frc} > 1.3 \,{\|}\, {\it i}_\mathrm{unf} - {\it z}_\mathrm{unf} > 1.3) ] \nonumber\\ &&\qquad\quad \left. {\|}\, ({\it z}_\mathrm{frc} > {\it z}_{3\sigma } \,{\|}\, {\it z}^\mathrm{ap}_\mathrm{frc} > {\it z}_{3\sigma }) \right\rbrace \!, \end{eqnarray} (1)for z ≃ 6.6, and   \begin{eqnarray} &&{\mathit {NB816}^\mathrm{ap}_\mathrm{frc} < {\it NB816}_{5\sigma }} \nonumber \\ && \&\&\ \left({\it g}_\mathrm{frc} > {\it g}_{3\sigma } \,{\|}\, {\it g}^\mathrm{ap}_\mathrm{frc} > {\it g}_{3\sigma } \right) \nonumber \\ && \&\&\ \left( {\it i}_\mathrm{frc} \!-\! {\it NB816}_\mathrm{frc} > 1.2 \,{\|}\, {\it i}_\mathrm{unf} \!-\! {\it NB816}_\mathrm{unf} \!>\! 1.2 \right) \nonumber \\ && \&\&\ \left\lbrace [ ({\it r}_\mathrm{frc} < {\it r}_{3\sigma } \,{\|}\, {\it r}^\mathrm{ap}_\mathrm{frc} < {\it r}_{3\sigma }) \right. \nonumber \\ &&\qquad\quad \&\&\ ({\it r}_\mathrm{frc} - {\it i}_\mathrm{frc} > 1.0 \,{\|}\, {\it r}_\mathrm{unf} - {\it i}_\mathrm{unf} > 1.0) ] \nonumber \\ &&\qquad\quad \left. {\|}\, ({\it r}_\mathrm{frc} > {\it r}_{3\sigma } \,{\|}\, {\it r}^\mathrm{ap}_\mathrm{frc} > {\it r}_{3\sigma }) \right\rbrace \!, \end{eqnarray} (2)for z ≃ 5.7, where the indices of “frc” and “unf” represent the forced and unforced photometry, respectively. The subscript of 5σ(3σ) indicates the 5σ(3σ) limiting magnitude for a given filter. The values with and without a superscript of “ap” indicate the aperture and total magnitudes, respectively. These magnitudes are derived with the hscPipe software (see section 2; Bosch et al. 2018). The limits of the $$i-\mathit {NB816}$$ and $$z- \mathit {NB921}$$ colors are the same as those of Ouchi et al. (2008, 2010), respectively. To exploit the survey capability of HSC in identifying rare objects, we use the 3σ g and r limiting magnitude (instead of the value of 2σ used in Ouchi et al. 2008) for the criteria of Lyman break off-band non-detection. In process (4), we replace 3σ with 2σ for the g- and r-magnitude criteria for consistency with the previous studies. Note that we do not apply the flags_pixel_bright_object_[center/any] masking to the LAE ALL catalog in order to maximize LAE targets for future follow-up observations (Aihara et al. 2018b). These object masking flags are used in process (4). Parameters and flags: Similar to Ono et al. (2018), we set several hscPipe parameters and flags in the HSC catalog to exclude e.g., blended sources, objects affected by saturated pixels, and nearby bright source halos. We also mask regions where exposure times are relatively short by using the countinputs parameter, Nc, which denotes the number of exposures at a source position for a given filter. Table 2 summarizes the values and provides brief explanations of the hscPipe parameters and flags used for our LAE selection. The full details of these parameters and flags are presented in Aihara et al. (2018b). To search for LAEs in large areas of the HSC fields, we do not apply the countinputs parameter to the BB images. The number of objects selected in this process is nSQL ≃ 121000. Visual inspections for object images: To exclude cosmic rays, cross-talks, compact stellar objects, and artificial diffuse objects, we perform visual inspections for the BB and NB images of all the objects selected in the process (1). Most spurious sources are diffuse components near bright stars and extended nearby galaxies. The hscPipe software conducts the cmodel fit to broad light profiles of such diffuse sources in the NB images, which enhances the BB − NB colors. For this reason, the samples constructed in the current SQL selection are contaminated by many diffuse components. Due to the clear difference of the appearance between LAE candidates and diffuse components, such spurious sources can be easily excluded through the visual inspections. The number of objects selected in this process is nvis ≃ 10900. The visual inspection processes have mainly been conducted by one of the authors. As a reliability check, four authors in this paper have individually carried out such visual inspections for ≃5300 objects in the UD-COSMOS NB816 fields, and compared the results of the LAE selection. The difference in the number of selected LAEs is within  ± 5 objects. Thus, we do not find a large difference in our visual inspection results. Rejection of variable and moving objects with multi-epoch images: We exclude variable and moving objects such as supernovae, AGNs, satellite trails, and asteroids using multi-epoch NB images. The NB images were typically taken a few months to years after the BB imaging observations. For this reason, there is a possibility that sources with an NB flux excess are variable or moving objects which happened to enhance the luminosities during the NB imaging observations. The NB images are created by coadding ≃10–20 and ≃3–5 frames of 15 min exposures for the current HSC UD and D data, respectively. Using the multi-epoch images, we automatically remove the variable and moving objects as follows. First, we measure the flux for individual epoch images, f1epoch, for each object. Next, we obtain an average, fave, and a standard deviation, σepoch, from a set of the f1epoch values after a 2σ flux clipping. Finally, we discard any object having at least a multi-epoch image with a significantly large f1epoch value of f1epoch ≥ fave + Aepoch × σepoch. Here we tune the Aepoch factor based on the depth of the NB fields. The Aepoch value is typically ≃2.0–2.5. Figure 3 shows examples of the spurious sources. We also perform visual inspections for multi-epoch images to remove contaminants which are not excluded in the automatic rejection above. We refer to the objects remaining after this process as the LAE ALL catalog. forcedselection: In the selection criteria of equations (1) and (2), the HSC LAE ALL catalog is obtained in the combination of the forced and unforced colors. In this process, we select LAEs only with the forced color excess to create the forced LAE subsamples from the HSC LAE ALL catalog. In addition, the 3σ limit is replaced with 2σ for the criteria of g- and r-band non-detections. Here we also adopt a new stringent color criterion of $$z- \mathit {NB921} >1.8$$ for z ≃ 6.6 LAEs. Due to the difference of the z-band transmission curves between SCam and HSC, the criterion of $$z- \mathit {NB921} >1.0$$ in equation (1) does not allow us to select LAEs whose EW0,Lyα is similar to those of previous SCam studies. The BB − NB color criteria in in the forced selection correspond to the rest-frame Lyα EW of EW0,Lyα > 14 Å and >10 Å for z ≃ 6.6 and z ≃ 5.7 LAEs, respectively. These EW0,Lyα limits are comparable to those of the previous SCam studies (e.g., Ouchi et al. 2010). The relation between EW0,Lyα and BB − NB colors is described in Konno et al. (2018) in detail. Moreover, we remove the objects in masked regions defined by the flags_pixel_bright_object_[center/any] parameters (Aihara et al. 2018b). We refer to the set of objects remaining after this process as the forced LAE catalog. This forced LAE catalog is used for studies of LAE statistics, such as measurements of Lyα EW scale lengths. The LAE candidates selected in this forced selection are referred to as the forced LAEs. On the other hand, we refer to the remaining LAE candidates in the HSC LAE ALL catalog as the unforced LAEs. The examples of forced and unforced LAEs are shown in figure 3. As shown in the top right-hand panels of figure 3, the unforced LAEs have a ≃0$${^{\prime\prime}_{.}}$$2–0$${^{\prime\prime}_{.}}$$3 spatial offset between centroids in NB and BB. Fig. 2. View largeDownload slide Flow chart of the HSC LAE selection process. See section 3 for more details. (Color online) Fig. 2. View largeDownload slide Flow chart of the HSC LAE selection process. See section 3 for more details. (Color online) Fig. 3. View largeDownload slide Multi-band cutout images of our example LAEs and spurious sources. (a) LAEs at z ≃ 6.6 (top) and z ≃ 5.7 (bottom) in the forced LAE catalog. (b) LAEs at z ≃ 6.6 (top) and z ≃ 5.7 (bottom) in the unforced catalog. In the rightmost cutout images, the yellow solid and cyan dashed circles represent the central positions of the unforced LAEs in the NB and BB images, respectively. The diameters of the yellow solid and dashed circles in the cutout images of the unforced LAEs are 1″ and 0$${^{\prime\prime}_{.}}$$5, respectively. (c) Spurious sources with an NB magnitude-excess similar to that of LAE candidates (four panel sets at the top): 1, variable (e.g., supernova); 2, cosmic ray; 3, cross-talk artifact; 4, moving object (e.g., asteroids) and corresponding multi-epoch images (four panel sets at the bottom). The image size is 4″ × 4″ for the LAEs and spurious sources. (Color online) Fig. 3. View largeDownload slide Multi-band cutout images of our example LAEs and spurious sources. (a) LAEs at z ≃ 6.6 (top) and z ≃ 5.7 (bottom) in the forced LAE catalog. (b) LAEs at z ≃ 6.6 (top) and z ≃ 5.7 (bottom) in the unforced catalog. In the rightmost cutout images, the yellow solid and cyan dashed circles represent the central positions of the unforced LAEs in the NB and BB images, respectively. The diameters of the yellow solid and dashed circles in the cutout images of the unforced LAEs are 1″ and 0$${^{\prime\prime}_{.}}$$5, respectively. (c) Spurious sources with an NB magnitude-excess similar to that of LAE candidates (four panel sets at the top): 1, variable (e.g., supernova); 2, cosmic ray; 3, cross-talk artifact; 4, moving object (e.g., asteroids) and corresponding multi-epoch images (four panel sets at the bottom). The image size is 4″ × 4″ for the LAEs and spurious sources. (Color online) Table 2. HSC SQL parameters and flags for our LAE selection. Parameter or flag  Value  Band  Comment  detect_is_tract_inner  True  —  Object is in an inner region of a tract and not in the overlapping region with adjacent tracts  detect_is_patch_inner  True  —  Object is in an inner region of a patch and not in the overlapping region with adjacent patches  countinputs  >=3  NB  Number of visits at a source position for a given filter  flags_pixel_edge  False  grizy, NB  Locate within images  flags_pixel_interpolated_center  False  grizy, NB  None of the central 3 × 3 pixels of an object is interpolated  flags_pixel_saturated_center  False  grizy, NB  None of the central 3 × 3 pixels of an object is saturated  flags_pixel_cr_center  False  grizy, NB  None of the central 3 × 3 pixels of an object is masked as cosmic ray  flags_pixel_bad  False  grizy, NB  None of the pixels in the footprint of an object is labelled as bad  Parameter or flag  Value  Band  Comment  detect_is_tract_inner  True  —  Object is in an inner region of a tract and not in the overlapping region with adjacent tracts  detect_is_patch_inner  True  —  Object is in an inner region of a patch and not in the overlapping region with adjacent patches  countinputs  >=3  NB  Number of visits at a source position for a given filter  flags_pixel_edge  False  grizy, NB  Locate within images  flags_pixel_interpolated_center  False  grizy, NB  None of the central 3 × 3 pixels of an object is interpolated  flags_pixel_saturated_center  False  grizy, NB  None of the central 3 × 3 pixels of an object is saturated  flags_pixel_cr_center  False  grizy, NB  None of the central 3 × 3 pixels of an object is masked as cosmic ray  flags_pixel_bad  False  grizy, NB  None of the pixels in the footprint of an object is labelled as bad  View Large In total, we identify 2230 and 1045 LAE candidates in the HSC LAE ALL and forced catalogs, respectively. Table 1 presents the numbers of LAE candidates in each field. The machine-readable catalogs of all the LAE candidates will be provided on our project website.2 The photometric properties of example LAE candidates are shown in table 3. Table 3. Photometric properties of example LAE candidates.* Object ID  NB  g  r  i  z  y    (mag)  (mag)  (mag)  (mag)  (mag)  (mag)  (1)  (2)  (3)  (4)  (5)  (6)  (7)  UD-SXDS (NB921)  HSC J021601−041442  23.85 ± 0.10  26.89 ± 0.45  27.03 ± 0.62  26.65 ± 0.63  25.28 ± 0.31  25.29 ± 0.53  HSC J021754−051454  24.01 ± 0.12  >27.6  >27.3  >26.9  26.09 ± 0.57  25.21 ± 0.50  HSC J021702−050604  24.64 ± 0.21  >27.6  >27.3  >26.9  >26.5  >25.8  HSC J021638−043228  24.74 ± 0.23  >27.6  >27.3  >26.9  26.17 ± 0.60  >25.8  HSC J021609−050236  24.90 ± 0.26  27.53 ± 0.72  27.29 ± 0.75  >26.9  26.32 ± 0.67  >25.8  UD-COSMOS (NB816)  HSC J100243+024551  23.69 ± 0.08  >27.6  >27.3  26.49 ± 0.53  >26.6  >25.8  HSC J100239+022806  24.14 ± 0.13  >27.6  >27.3  26.76 ± 0.64  26.12 ± 0.54  >25.8  HSC J100243+015931  24.63 ± 0.19  >27.6  >27.3  >27.0  >26.6  >25.8  HSC J095936+014108  25.02 ± 0.26  >27.6  >27.3  >27.0  >26.6  >25.8  HSC J100245+021536  25.15 ± 0.29  >27.6  >27.3  >27.0  >26.6  >25.8  Object ID  NB  g  r  i  z  y    (mag)  (mag)  (mag)  (mag)  (mag)  (mag)  (1)  (2)  (3)  (4)  (5)  (6)  (7)  UD-SXDS (NB921)  HSC J021601−041442  23.85 ± 0.10  26.89 ± 0.45  27.03 ± 0.62  26.65 ± 0.63  25.28 ± 0.31  25.29 ± 0.53  HSC J021754−051454  24.01 ± 0.12  >27.6  >27.3  >26.9  26.09 ± 0.57  25.21 ± 0.50  HSC J021702−050604  24.64 ± 0.21  >27.6  >27.3  >26.9  >26.5  >25.8  HSC J021638−043228  24.74 ± 0.23  >27.6  >27.3  >26.9  26.17 ± 0.60  >25.8  HSC J021609−050236  24.90 ± 0.26  27.53 ± 0.72  27.29 ± 0.75  >26.9  26.32 ± 0.67  >25.8  UD-COSMOS (NB816)  HSC J100243+024551  23.69 ± 0.08  >27.6  >27.3  26.49 ± 0.53  >26.6  >25.8  HSC J100239+022806  24.14 ± 0.13  >27.6  >27.3  26.76 ± 0.64  26.12 ± 0.54  >25.8  HSC J100243+015931  24.63 ± 0.19  >27.6  >27.3  >27.0  >26.6  >25.8  HSC J095936+014108  25.02 ± 0.26  >27.6  >27.3  >27.0  >26.6  >25.8  HSC J100245+021536  25.15 ± 0.29  >27.6  >27.3  >27.0  >26.6  >25.8  *(1) Object ID. (2)–(7) Total magnitude of NB, g, r, i, z, and y bands. The 2σ limits of the total magnitudes for the undetected bands. (The complete machine-readable catalogs will be available on our project web page.)2 View Large As shown in table 1, the number of z ≃ 5.7 LAEs in D-DEEP2-3 appears to be large compared to that of the other z ≃ 5.7 fields. This may be because the seeing of the NB816 images of D-DEEP2-3 is better than that of the other z ≃ 5.7 fields. Similarly, the small number of z ≃ 6.6 LAEs in UD-SXDS may be affected by the seeing size. The number density of LAEs is discussed in the next section. Note that edge regions of UD-COSMOS is overlapped with a flanking field, D-COSMOS (Aihara et al. 2018a). We find that 30(7) LAEs in UD-COSMOS are also selected in the HSC LAE ALL(forced) sample of D-COSMOS. To analyze the D field independently in the following sections, we include the overlapped LAEs in the D-COSMOS sample. Figure 4 shows the color–magnitude diagrams for the LAE candidates. The solid curves in the color–magnitude diagrams indicate the 3σ errors of BB − NB color as a function of the NB flux, fNB, given by   \begin{equation} \pm 3 \sigma _{\rm BB-NB} = -2.5 \log _{10} \left( 1\, \mp \, 3\frac{\sqrt{f^2_{\rm 1\sigma NB} + f^2_{\rm 1\sigma BB}}}{f_{\rm NB}} \right), \end{equation} (3)where f1σNB and f1σBB are the 1σ flux error in the z and NB921 (i and NB816) bands for z ≃ 6.6 (z ≃ 5.7), respectively. As shown in figure 4, the LAE candidates have a significant NB magnitude excess. Fig. 4. View largeDownload slide (Top) Color of z − NB921 as a function of NB921 magnitude for the LAEs at z ≃ 6.6 in the UD (left) and D (right) fields. The filled red and open magenta circles denote the forced and unforced LAEs, respectively. For the LAEs undetected in the z-band images, the z-band magnitudes are replaced with the 2σ limiting magnitudes. The x-axis denotes the forced (unforced) z − NB921 colors for the forced (unforced) LAEs. The horizontal dashed and dotted lines shows the color criteria of z − NB921 >1.0 and z − NB921 >1.8, respectively. The gray dots represent objects detected in the NB921 images. The solid curves show the 3σ error tracks of z − NB921 color for each field. The 3σ error tracks are derived by equation (3). (Bottom) Color of i − NB816 as a function of NB816 magnitude for the LAEs at z ≃ 5.7. The definitions of symbols, curves, and lines are the same as those of the top panels. (Color online) Fig. 4. View largeDownload slide (Top) Color of z − NB921 as a function of NB921 magnitude for the LAEs at z ≃ 6.6 in the UD (left) and D (right) fields. The filled red and open magenta circles denote the forced and unforced LAEs, respectively. For the LAEs undetected in the z-band images, the z-band magnitudes are replaced with the 2σ limiting magnitudes. The x-axis denotes the forced (unforced) z − NB921 colors for the forced (unforced) LAEs. The horizontal dashed and dotted lines shows the color criteria of z − NB921 >1.0 and z − NB921 >1.8, respectively. The gray dots represent objects detected in the NB921 images. The solid curves show the 3σ error tracks of z − NB921 color for each field. The 3σ error tracks are derived by equation (3). (Bottom) Color of i − NB816 as a function of NB816 magnitude for the LAEs at z ≃ 5.7. The definitions of symbols, curves, and lines are the same as those of the top panels. (Color online) 4 Checking the reliability of our LAE selection Here we check the reliability of our LAE selection. 4.1 Spectroscopic confirmations We have conducted optical spectroscopic observations with Subaru/FOCAS and Magellan/LDSS3 for 18 bright LAE candidates with NB ≲ 24 mag. From these observations, we have confirmed 13 LAEs. By investigating our spectroscopic catalog of Magellan/IMACS, we also spectroscopically identify eight LAEs with $$\mathit {NB}\lesssim 24\:$$mag. In addition, we find that there are 75 LAEs spectroscopically confirmed in literature (Murayama et al. 2007; Ouchi et al. 2008, 2010; Taniguchi et al. 2009; Mallery et al. 2012; Sobral et al. 2015; R. Higuchi et al. in preparation). In total, 96 LAEs have been confirmed in our spectroscopy and previous studies. Using the spectroscopic sample that has a known number of observed LAEs, we estimate the contamination rate to be ≃0%–30%. The details of the spectroscopic observations and contamination rates are given by Shibuya et al. (2018). 4.2 LAE surface number density Figure 5 shows the surface number density (SND) of our LAE candidates and LAEs identified in previous Subaru/SCam NB surveys, SCam LAEs (e.g., Ouchi et al. 2008, 2010). We find that the SNDs of the forced LAEs are comparable to those of SCam LAEs. On the other hand, the SNDs of unforced LAEs at z ≃ 6.6 are higher than that of SCam LAEs. The high SND of the unforced LAEs is mainly caused by the color criterion for the HSC LAE ALL catalog of $$z- \mathit {NB921} >1.0$$ that is less stringent than $$z- \mathit {NB921} >1.8$$ (see section 3). We also identify SND humps of our forced LAEs at z ≃ 6.6 at the bright-end of NB ≃ 23 mag in UD-COSMOS. The presence of such an SND hump has been reported by z ≃ 6.6 LAE studies (e.g., Matthee et al. 2015). The significance of the bright-end hump’s existence in Lyα LFs is ≃3σ, which are discussed in Konno et al. (2018). The slight decline in SNDs at a faint NB magnitude of NB ≳ 24.5 mag would originate from the incompleteness of the LAE detection and selection. Konno et al. (2018) present the SND corrected for the incompleteness. Fig. 5. View largeDownload slide Surface number density (SND) of the HSC LAEs at z ≃ 6.6 (five panels on the left) and ≃5.7 (four panels on the right) in each UD and D field. The filled red and open magenta circles indicate the LAEs in the forced and ALL catalog, respectively. The error bars are given by Poisson statistics from the number of LAEs. The gray crosses represent the LAEs in Ouchi et al. (2010) for z ≃ 6.6 and Ouchi et al. (2008) for z ≃ 5.7. The data points of the gray crosses are identical in all the fields for each redshift. The SND slight declines in the HSC LAEs at NB ≳ 24.5 mag would be originated from the incompleteness of the LAE detection and selection. The completeness-corrected SNDs are presented in Konno et al. (2018). The data points of the HSC LAEs are slightly shifted along the x-axis for clarity. (Color online) Fig. 5. View largeDownload slide Surface number density (SND) of the HSC LAEs at z ≃ 6.6 (five panels on the left) and ≃5.7 (four panels on the right) in each UD and D field. The filled red and open magenta circles indicate the LAEs in the forced and ALL catalog, respectively. The error bars are given by Poisson statistics from the number of LAEs. The gray crosses represent the LAEs in Ouchi et al. (2010) for z ≃ 6.6 and Ouchi et al. (2008) for z ≃ 5.7. The data points of the gray crosses are identical in all the fields for each redshift. The SND slight declines in the HSC LAEs at NB ≳ 24.5 mag would be originated from the incompleteness of the LAE detection and selection. The completeness-corrected SNDs are presented in Konno et al. (2018). The data points of the HSC LAEs are slightly shifted along the x-axis for clarity. (Color online) Figure 6 compiles the SNDs of all the HSC UD and D fields. We find that our SNDs show a small field-to-field variation, but typically follow those of the SCam LAEs. Fig. 6. View largeDownload slide SND as a function of NB magnitude for the LAEs at z ∼ 6.6 (left) and ∼5.7 (right) in the HSC LAE forced catalog. The colored symbols denote the LAEs in each UD and D field (green circles: UD-SXDS; magenta squares: UD-COSMOS; cyan triangles: D-ELAIS-N1; pink inverse-triangles: D-DEEP2-3; orange diamonds: D-COSMOS). The error bars are given by Poisson statistics from the number of LAEs. The gray crosses represent the LAEs in Ouchi et al. (2010) for z ≃ 6.6 and Ouchi et al. (2008) for z ≃ 5.7. The slight declinein SND in the HSC LAEs at NB ≳ 24.5 mag originates from the incompleteness of the LAE selection. The completeness-corrected SNDs are presented in Konno et al. (2018). The data points of the HSC LAEs are slightly shifted along the x-axis for clarity. (Color online) Fig. 6. View largeDownload slide SND as a function of NB magnitude for the LAEs at z ∼ 6.6 (left) and ∼5.7 (right) in the HSC LAE forced catalog. The colored symbols denote the LAEs in each UD and D field (green circles: UD-SXDS; magenta squares: UD-COSMOS; cyan triangles: D-ELAIS-N1; pink inverse-triangles: D-DEEP2-3; orange diamonds: D-COSMOS). The error bars are given by Poisson statistics from the number of LAEs. The gray crosses represent the LAEs in Ouchi et al. (2010) for z ≃ 6.6 and Ouchi et al. (2008) for z ≃ 5.7. The slight declinein SND in the HSC LAEs at NB ≳ 24.5 mag originates from the incompleteness of the LAE selection. The completeness-corrected SNDs are presented in Konno et al. (2018). The data points of the HSC LAEs are slightly shifted along the x-axis for clarity. (Color online) 4.3 Matching rate of HSC LAEs and SCam LAEs The UD-SXDS field has been observed previously by SCam equipped with the NB921 and NB816 filters (Ouchi et al. 2008, 2010). We compare the catalogs of our selected HSC LAE candidates and SCam LAEs, and calculate the object-matching rates as a function of NB magnitudes. The object-matching radius is 1″. The object-matching rate between the HSC LAEs and SCam LAEs is ≃90% at bright NB magnitudes of ≲ 24 mag. The high object-matching rate indicates that we adequately identify LAEs in our selection processes. However, the matching rate decreases to ≃70% at the faint magnitude of ≃24.5 mag. This is due to the shallow depth of the HSC NB fields compared to the SCam ones. Konno et al. (2018) discuss the detection completeness of faint LAEs. 5 Results Here we present the Lyα EW distributions (subsection 5.1) and LABs selected with the HSC data (subsection 5.2). For consistency with previous LAE studies, we use the forced LAE sample in the following analyses, unless otherwise specified. 5.1 Lyα EW distribution We present the Lyα EW distributions for LAEs at z ≃ 5.7–6.6. Using the method described in the Appendix, we calculate the rest-frame Lyα EW, EW0,Lyα, for the LAEs. The y(z)-band magnitudes are used for the rest-frame UV continuum emission of z ≃ 6.6(z ≃ 5.7) LAEs. Figure 7 shows the observed Lyα EW distributions at z ≃ 5.7–6.6 in the UD and D fields. To quantify these Lyα EW distributions we perform Monte Carlo (MC) simulations. The procedure of the MC simulations is similar to that of e.g., Shimasaku et al. (2006), Ouchi et al. (2008), and Zheng et al. (2014). First, we generate artificial LAEs in a Lyα luminosity range of log LLyα/erg s−1 = 42–44 according to z ≃ 5.7–6.6 Lyα LFs of Konno et al. (2018). Next, we assign Lyα EW and BB magnitudes to each LAE by assuming that the Lyα EW distributions are the exponential and Gaussian functions (e.g., Gronwall et al. 2007; Kashikawa et al. 2011; Oyarzún et al. 2016):   \begin{equation} \frac{{\rm d}N}{{\rm d} EW} = N \exp \left(-\frac{EW}{W_{\rm e}}\right), \end{equation} (4)and   \begin{equation} \frac{{\rm d}N}{{\rm d} EW} = N \frac{1}{\sqrt{2\pi \sigma _{\rm g}^2}} \exp \left(-\frac{EW^2}{2\sigma _{\rm g}^2} \right), \end{equation} (5)where N is the galaxy number, and We and σg are the Lyα EW scale lengths of the exponential and Gaussian functions, respectively. By changing the intrinsic We and σg values, we make samples of artificial Lyα EW distributions. We then select LAEs based on NB and BB limiting magnitudes and BB − NB colors corresponding to Lyα EW limits which are the same as those of our LAE selection criteria (section 3). Finally, the best-fitting Lyα EW scale lengths are obtained by fitting the artificial Lyα EW distribution to the observed ones. Fig. 7. View largeDownload slide Lyα EW distribution for the HSC LAEs at z ≃ 6.6 (left) and z ≃ 5.7 (right). The top and bottom panels show the UD and D fields, respectively. The thin gray histograms with error bars denote the Lyα EW distributions for the forced LAEs. The error bars are given by Poisson statistics from the number of sample LAEs. The red solid and blue dashed lines present the best-fitting exponential and Gaussian functions of equations (4) and (5), respectively, which are obtained from MC simulations with the EW0,Lyα uncertainties (see subsection 5.1 for more details). (Color online) Fig. 7. View largeDownload slide Lyα EW distribution for the HSC LAEs at z ≃ 6.6 (left) and z ≃ 5.7 (right). The top and bottom panels show the UD and D fields, respectively. The thin gray histograms with error bars denote the Lyα EW distributions for the forced LAEs. The error bars are given by Poisson statistics from the number of sample LAEs. The red solid and blue dashed lines present the best-fitting exponential and Gaussian functions of equations (4) and (5), respectively, which are obtained from MC simulations with the EW0,Lyα uncertainties (see subsection 5.1 for more details). (Color online) Figure 7 presents the Lyα EW distributions obtained in the MC simulations. As shown in figure 7, we find that the Lyα EW distributions are reasonably explained by the exponential and Gaussian profiles. The best-fitting scale lengths are summarized in table 4. The best-fitting exponential (Gaussian) Lyα scale lengths are, on average, 153 ± 18 Å and 154 ± 15 Å (146 ± 24 Å and 139 ± 14 Å) at z ≃ 5.7 and z ≃ 6.6 for the UD and D fields, respectively. As shown in table 4, there is no large difference in the Lyα EW scale lengths for the UD and D fields. This lack of large EW0,Lyα difference indicates that the results of our best-fitting Lyα EW scale lengths do not highly depend on the image depths or the detection incompleteness. In subsection 6.1, we discuss the redshift evolution of the Lyα EW scale lengths. We investigate LEW LAEs that have intrinsic Lyα EW values, $$EW_{\rm 0,\, Ly\alpha }^{\rm int}$$, that exceed 240 Å (e.g., Malhotra & Rhoads 2002; Dawson et al. 2004). To obtain $$EW_{\rm 0,\, Ly\alpha }^{\rm int}$$, we correct for the IGM attenuation for Lyα using the prescriptions of Madau (1995). In the HSC LAE ALL sample, we find that 45 and 230 LAEs have a LEW of $$EW_{\rm 0,\, Ly\alpha }^{\rm int} > 240\,$$Å, for z ≃ 6.6 and z ≃ 5.7 LAEs, respectively. These LEW LAEs are candidates of young metal-poor galaxies and AGNs. The fraction of the LEW LAEs in the sample is 21% for z ≃ 5.7 LAEs. The fraction of LEW LAEs at z ≃ 5.7 is comparable to that of previous studies on z ≃ 5.7 LAEs (e.g., ≃25% at z ≃ 5.7 in Ouchi et al. 2008; ≃30%–40% at z ≃ 5.7 in Shimasaku et al. 2006). In contrast, the fraction of LEW LAEs at z ≃ 6.6 is 4%, which is lower than that at z ≃ 5.7. This low fraction at z ≃ 6.6 might be due to the neutral hydrogen IGM absorbing the Lyα emission. Of the LEW LAEs, 32 and 150 LAEs at z ≃ 6.6 and z ≃ 5.7 exceed $$EW_{\rm 0,\, Ly\alpha }^{\rm int} = 240$$ beyond the 1σ uncertainty of $$EW_{\rm 0,\, Ly\alpha }^{\rm int}$$, respectively. 5.2 LABs at z ≃ 5.7–6.6 We search for LABs with spatially extended Lyα emission. To identify LABs, we measure the NB isophotal areas, Aiso, of the forced LAEs. In this process, we include an unforced LAE, Himiko, which is an LAB that was identified in a previous SCam NB survey (Ouchi et al. 2009). First, we estimate the sky background level of the NB cutout images. Next, we run the SExtractor with the sky background level, and obtain the Aiso values as pixels with fluxes brighter than the 2σ sky fluctuation. Note that the NB magnitudes include both fluxes of Lyα and the rest-frame UV continuum emission. Instead of creating Lyα images by subtracting the flux contribution of the rest-frame UV continuum emission, we here simply use the NB images for consistency with previous studies (e.g., Ouchi et al. 2009). Using Aiso and NB magnitude diagrams, we select LABs which are significantly extended compared to point sources. This selection is similar to that of Yang et al. (2010). Figure 8 presents Aiso as a function of total NB magnitude. We also plot star-like point sources which are randomly selected in HSC NB fields. The Aiso and NB magnitude selection window is defined by a 2.5σ deviation from the Aiso–NB magnitude distribution for the star-like point sources. The value of 2.5σ is applied for fair comparison with previous studies of e.g., Yang et al. (2009, 2010) who have used ≃2–4σ. We perform visual inspections for the NB cutout images to remove unreliable LABs which are significantly affected by, e.g., diffuse halos of nearby bright stars. Fig. 8. View largeDownload slide Isophotal area, Aiso, as a function of NB magnitude to select LABs at z ≃ 6.6 (left) and z ≃ 5.7 (right). The top and bottom panels show the UD and D fields, respectively. The green diamonds denote the LABs. The filled red circles indicate the LAEs in the forced catalog. The gray dots represent star-like point sources selected in the HSC NB images. The diagonal and vertical lines denote the LAB selection criteria of Aiso and NB magnitude. The diagonal lines are defined by the 2.5σ deviation from the Aiso–NB magnitude distribution for the star-like point sources. The filled red circles with crosses indicate unreliable LAB candidates which are affected by e.g., diffuse halos of nearby bright stars. The z ≃ 6.6 LABs in the UD fields are CR7 (Sobral et al. 2015) and Himiko (Ouchi et al. 2009). (Color online) Fig. 8. View largeDownload slide Isophotal area, Aiso, as a function of NB magnitude to select LABs at z ≃ 6.6 (left) and z ≃ 5.7 (right). The top and bottom panels show the UD and D fields, respectively. The green diamonds denote the LABs. The filled red circles indicate the LAEs in the forced catalog. The gray dots represent star-like point sources selected in the HSC NB images. The diagonal and vertical lines denote the LAB selection criteria of Aiso and NB magnitude. The diagonal lines are defined by the 2.5σ deviation from the Aiso–NB magnitude distribution for the star-like point sources. The filled red circles with crosses indicate unreliable LAB candidates which are affected by e.g., diffuse halos of nearby bright stars. The z ≃ 6.6 LABs in the UD fields are CR7 (Sobral et al. 2015) and Himiko (Ouchi et al. 2009). (Color online) In total, we identify 11 LABs at z ≃ 5.7–6.6. Figure 9 and table 5 present multi-band cutout images and properties for the LABs, respectively. As shown in figure 9, these LABs are spatially extended in NB. Our HSC LAB selection confirms that CR7 and Himiko have a spatially extended Lyα emission. Six out of our 11 LABs have been confirmed by our spectroscopic follow-up observations (Shibuya et al. 2018) and previous studies (Ouchi et al. 2009; Mallery et al. 2012; Sobral et al. 2015). In subsection 6.2, we discuss the redshift evolution of the LAB number density. Fig. 9. View largeDownload slide Postage stamps of the LABs selected with the HSC NB data. The yellow contours indicate isophotal apertures with a threshold of 2σ sky background noise level. The size of the cutout images is 4″ × 4″. (Color online) Fig. 9. View largeDownload slide Postage stamps of the LABs selected with the HSC NB data. The yellow contours indicate isophotal apertures with a threshold of 2σ sky background noise level. The size of the cutout images is 4″ × 4″. (Color online) Table 5. Properties of the LABs selected in the HSC NB data.* Object ID  α  δ  $$\mathit {NB}_{\rm tot}$$  UVtot  log LLyα  $$\mathit {EW}_{\rm 0,\, Ly\alpha }$$  zspec    (J2000.0)  (J2000.0)  (mag)  (mag)  (erg s−1)  (Å)    (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  NB921 (z ≃ 6.6)  HSC J100058+014815†  10h00m58$${^{\rm s}_{.}}$$00  +01°48΄15$${^{\prime\prime}_{.}}$$14  23.25  24.48  43.9‖  211 ± 20‖  6.604†  HSC J021757−050844‡  02h17m57$${^{\rm s}_{.}}$$58  −05°08΄44$${^{\prime\prime}_{.}}$$64  23.50  25.40  43.4‖  $$78^{+8}_{-6}{^{\Vert }}$$  6.595‡  HSC J100334+024546§  10h03m34$${^{\rm s}_{.}}$$66  +02°45΄46$${^{\prime\prime}_{.}}$$56  23.61  24.97  43.5‖  61 ± 20‖  6.575§  NB816 (z ≃ 5.7)  HSC J100129+014929  10h01m29$${^{\rm s}_{.}}$$07  +01°49΄29$${^{\prime\prime}_{.}}$$81  23.47  25.87  43.4  $$95^{+40}_{-19}$$  5.707♯  HSC J100109+021513  10h01m09$${^{\rm s}_{.}}$$72  +02°15΄13$${^{\prime\prime}_{.}}$$45  23.13  25.77  43.6  $$257^{+172}_{-76}$$  5.712♯  HSC J100123+015600  10h01m23$${^{\rm s}_{.}}$$84  +01°56΄00$${^{\prime\prime}_{.}}$$46  23.94  26.43  43.3  $$106^{+70}_{-27}$$  5.726♯  HSC J095946+013208  09h59m46$${^{\rm s}_{.}}$$73  +01°32΄08$${^{\prime\prime}_{.}}$$45  24.16  26.12  43.1  $$52^{+25}_{-13}$$  —  HSC J100139+015428  10h01m39$${^{\rm s}_{.}}$$94  +01°54΄28$${^{\prime\prime}_{.}}$$34  24.11  26.58  43.2  $$100^{+66}_{-30}$$  —  HSC J161927+551144  16h19m27$${^{\rm s}_{.}}$$73  +55°11΄44$${^{\prime\prime}_{.}}$$70  22.88  24.86  43.7  $$89^{+33}_{-20}$$  —  HSC J161403+535701  16h14m03$${^{\rm s}_{.}}$$82  +53°57΄01$${^{\prime\prime}_{.}}$$25  23.53  25.32  43.4  $$51^{+23}_{-12}$$  —  HSC J232924+003600  23h29m24$${^{\rm s}_{.}}$$85  +00°36΄00$${^{\prime\prime}_{.}}$$34  23.62  26.48  43.4  $$55^{+45}_{-14}$$  —  Object ID  α  δ  $$\mathit {NB}_{\rm tot}$$  UVtot  log LLyα  $$\mathit {EW}_{\rm 0,\, Ly\alpha }$$  zspec    (J2000.0)  (J2000.0)  (mag)  (mag)  (erg s−1)  (Å)    (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  NB921 (z ≃ 6.6)  HSC J100058+014815†  10h00m58$${^{\rm s}_{.}}$$00  +01°48΄15$${^{\prime\prime}_{.}}$$14  23.25  24.48  43.9‖  211 ± 20‖  6.604†  HSC J021757−050844‡  02h17m57$${^{\rm s}_{.}}$$58  −05°08΄44$${^{\prime\prime}_{.}}$$64  23.50  25.40  43.4‖  $$78^{+8}_{-6}{^{\Vert }}$$  6.595‡  HSC J100334+024546§  10h03m34$${^{\rm s}_{.}}$$66  +02°45΄46$${^{\prime\prime}_{.}}$$56  23.61  24.97  43.5‖  61 ± 20‖  6.575§  NB816 (z ≃ 5.7)  HSC J100129+014929  10h01m29$${^{\rm s}_{.}}$$07  +01°49΄29$${^{\prime\prime}_{.}}$$81  23.47  25.87  43.4  $$95^{+40}_{-19}$$  5.707♯  HSC J100109+021513  10h01m09$${^{\rm s}_{.}}$$72  +02°15΄13$${^{\prime\prime}_{.}}$$45  23.13  25.77  43.6  $$257^{+172}_{-76}$$  5.712♯  HSC J100123+015600  10h01m23$${^{\rm s}_{.}}$$84  +01°56΄00$${^{\prime\prime}_{.}}$$46  23.94  26.43  43.3  $$106^{+70}_{-27}$$  5.726♯  HSC J095946+013208  09h59m46$${^{\rm s}_{.}}$$73  +01°32΄08$${^{\prime\prime}_{.}}$$45  24.16  26.12  43.1  $$52^{+25}_{-13}$$  —  HSC J100139+015428  10h01m39$${^{\rm s}_{.}}$$94  +01°54΄28$${^{\prime\prime}_{.}}$$34  24.11  26.58  43.2  $$100^{+66}_{-30}$$  —  HSC J161927+551144  16h19m27$${^{\rm s}_{.}}$$73  +55°11΄44$${^{\prime\prime}_{.}}$$70  22.88  24.86  43.7  $$89^{+33}_{-20}$$  —  HSC J161403+535701  16h14m03$${^{\rm s}_{.}}$$82  +53°57΄01$${^{\prime\prime}_{.}}$$25  23.53  25.32  43.4  $$51^{+23}_{-12}$$  —  HSC J232924+003600  23h29m24$${^{\rm s}_{.}}$$85  +00°36΄00$${^{\prime\prime}_{.}}$$34  23.62  26.48  43.4  $$55^{+45}_{-14}$$  —  *(1) Object ID. (2) Right ascension. (3) Declination. (4) Total magnitudes of NB921 for z ≃ 6.6 and NB816 bands for z ≃ 5.7. (5) Total magnitudes of y band for z ≃ 6.6 and z bands for z ≃ 5.7. (6) Lyα luminosity. (7) Rest-frame equivalent width of Lyα emission line. (8) Spectroscopic redshift. †CR7 in Sobral et al. (2015). ‡Himiko in Ouchi et al. (2009). §Spectroscopically confirmed in Shibuya et al. (2018). ‖Spectroscopic measurements from the literature. ♯Spectroscopically confirmed in Mallery et al. (2012). View Large Table 4. Best-fitting Lyα EW scale lengths.* Redshift    We    σg      (Å)    (Å)  (1)    (2)    (3)  6.6 (UD)    $$168^{+4}_{-4}$$    $$124^{+8}_{-8}$$  5.7 (UD)    $$138^{+6}_{-4}$$    $$164^{+2}_{-2}$$  6.6 (D)    $$140^{+14}_{-16}$$    $$154^{+4}_{-24}$$  5.7 (D)    $$168^{+18}_{-18}$$    $$128^{+36}_{-12}$$  Redshift    We    σg      (Å)    (Å)  (1)    (2)    (3)  6.6 (UD)    $$168^{+4}_{-4}$$    $$124^{+8}_{-8}$$  5.7 (UD)    $$138^{+6}_{-4}$$    $$164^{+2}_{-2}$$  6.6 (D)    $$140^{+14}_{-16}$$    $$154^{+4}_{-24}$$  5.7 (D)    $$168^{+18}_{-18}$$    $$128^{+36}_{-12}$$  *(1) Redshift of the LAE sample. The parenthesis indicates the UD or D fields. (2) Best-fitting Lyα EW scale length of the exponential form. (3) Best-fitting Lyα EW scale length of the Gaussian form. View Large 6 Discussion 6.1 Redshift evolution of Lyα EW distribution We discuss the redshift evolution of the Lyα EW scale lengths through a compilation of the results from literature (Zheng et al. 2014; Ouchi et al. 2008; Nilsson et al. 2009; Hu et al. 2010; Kashikawa et al. 2011; Cowie et al. 2011; Ciardullo et al. 2012). Figure 10 shows the redshift evolution of the Lyα EW scale lengths at z ≃ 0–7. Our best-fitting Lyα scale lengths are comparable to those of Kashikawa et al. (2011) and/or Zheng et al. (2014) at z ≃ 5.7–6.6. The high Lyα EW scale lengths at high z would indicate that metal-poor and/or less dusty galaxies with a strong Lyα emission are more abundant at higher z (e.g., Stark et al. 2011). In addition, Zheng et al. (2014) have found that the Lyα EW scale length increases towards high z following a (1 + z)-form. Our We and σg values for z ≃ 5.7–6.6 are also roughly comparable to Zheng et al.’s (1 + z)-form evolution. However, no significant evolution in the Lyα EW scale lengths from z ≃ 5.7 to z ≃ 6.6 is identified in our HSC LAE data, although a possible decline in σg in the UD fields is found. A slight decrease both in We and σg from z ≃ 5.7 to z ≃ 6.6 has been found by Kashikawa et al. (2011). This sudden decline in the Lyα scale lengths at z ≃ 6.6 may be caused by the increasing hydrogen neutral fraction in the epoch of the cosmic reionization at z ≳ 7. Note that the Lyα EW scale length measurements would largely depend on BB and NB depths and Lyα EW cuts. Using deeper NB and BB images from the future HSC data release, we will examine the redshift evolution of Lyα scale lengths accurately. Fig. 10. View largeDownload slide Redshift evolution of the best-fitting Lyα EW scale lengths of the exponential (top) and Gaussian (bottom) functions. The red squares and circles indicate our HSC LAEs in the UD and D fields, respectively. The black symbols are taken from the data points in literature, as compiled in Zheng et al. (2014) (crosses: Cowie et al. 2011; asterisks: Ciardullo et al. 2012; filled triangle: Nilsson et al. 2009; filled inverse triangles: Ouchi et al. 2008; filled diamonds: Kashikawa et al. 2011; open circles: results of Monte Carlo simulations using data of Zheng et al. 2014 and Hu et al. 2010). The gray curves indicate the best-fitting (1 + z)-form functions obtained in Zheng et al. (2014). (Color online) Fig. 10. View largeDownload slide Redshift evolution of the best-fitting Lyα EW scale lengths of the exponential (top) and Gaussian (bottom) functions. The red squares and circles indicate our HSC LAEs in the UD and D fields, respectively. The black symbols are taken from the data points in literature, as compiled in Zheng et al. (2014) (crosses: Cowie et al. 2011; asterisks: Ciardullo et al. 2012; filled triangle: Nilsson et al. 2009; filled inverse triangles: Ouchi et al. 2008; filled diamonds: Kashikawa et al. 2011; open circles: results of Monte Carlo simulations using data of Zheng et al. 2014 and Hu et al. 2010). The gray curves indicate the best-fitting (1 + z)-form functions obtained in Zheng et al. (2014). (Color online) 6.2 Redshift evolution of LAB number density We discuss the redshift evolution of the LAB number density, NLAB. Figure 11 shows NLAB at z ≃ 0–7 measured by this study and from the literature (Saito et al. 2006; Keel et al. 2009; Matsuda et al. 2009; Yang et al. 2009, 2010). For the plot of the NLAB, Yang et al. (2010) have compiled NLAB measurements down to an NB surface brightness (SB) limit of 5 × 10−18 erg s−1 cm−2 arcsec−2. The SB limits of our HSC NB data are ≃5 × 10−18 and ≃8 × 10−18 erg s−1 cm−2 for the UD and D fields, respectively. Our HSC NB images at least for the UD fields are comparably deep, allowing for fair comparisons with Yang et al.’s NLAB plot. Our NLAB values are 1.4 × 10−6 and 2.9 × 10−7 Mpc−3 (2.6 × 10−7 and 1.1 × 10−7 Mpc−3) at z ≃ 5.7 and z ≃ 6.6 in the UD (D) fields, respectively. The number density at z ≃ 6–7 is ≃10–100 times lower than those claimed for LABs at z ≃ 2–3 (e.g., Matsuda et al. 2004; Yang et al. 2009, 2010). As shown in figure 11, there is an evolutional trend that NLAB increases from z ≃ 7 to ≃3 and subsequently decreases from z ≃ 3 to ≃0. This trend of the LAB number density evolution is similar to the Madau–Lilly plot of the cosmic star formation rate density (SFRD) evolution (e.g., Madau et al. 1996; Lilly et al. 1996). Similar to Shibuya et al. (2016), we fit the Madau–Lilly plot-type formula,   \begin{equation} N_{\rm LAB}(z) = a \frac{(1+z)^b}{1 + [(1+z)/c]^d}, \end{equation} (6)where a, b, c, and d are free parameters (Madau & Dickinson 2014), to our NLAB evolution. For the fitting, we exclude Matsuda et al.’s (2009) data point, which was obtained in an overdense region, SSA22. The best-fitting parameters are a = 9.1 × 10−8, b = 2.9, c = 5.0, and d = 11.7. Fig. 11. View largeDownload slide Redshift evolution of the LAB number density. The filled red squares and filled red circles denote the LABs selected in the HSC UD and D fields, respectively. The error bars are given by Poisson statistics from the LAB number counts. The black symbols show LABs found in the literature (filled diamond: Keel et al. 2009; filled circle: Yang et al. 2009; open circle: Yang et al. 2010; filled inverse-triangle: Matsuda et al. 2004; pentagon: Saito et al. 2006). All the measurements are based on LABs identified down to the surface brightness limit of ≃5 × 10−18 erg s−1 cm−2 arcsec−2. The gray solid curve represents the best-fitting formula of equation (6) to the data points expect for the measurement in the SSA22 proto-cluster region. (Color online) Fig. 11. View largeDownload slide Redshift evolution of the LAB number density. The filled red squares and filled red circles denote the LABs selected in the HSC UD and D fields, respectively. The error bars are given by Poisson statistics from the LAB number counts. The black symbols show LABs found in the literature (filled diamond: Keel et al. 2009; filled circle: Yang et al. 2009; open circle: Yang et al. 2010; filled inverse-triangle: Matsuda et al. 2004; pentagon: Saito et al. 2006). All the measurements are based on LABs identified down to the surface brightness limit of ≃5 × 10−18 erg s−1 cm−2 arcsec−2. The gray solid curve represents the best-fitting formula of equation (6) to the data points expect for the measurement in the SSA22 proto-cluster region. (Color online) The similarity of the cosmic SFRD and LAB evolution might indicate that the origin of LABs are related to the star formation activity. As described in section 1, LABs are thought to be formed via physical mechanisms that are connected with star formation, e.g., cold gas accretion and galactic superwinds. The cold gas accretion could produce the extended Lyα emission powered by the gravitational energy (e.g., Mas-Ribas & Dijkstra 2016; Momose et al. 2016; Mas-Ribas et al. 2017). On the other hand, the superwinds induced by the starbursts in the central galaxies would blow out the surrounding neutral gas, and form extended Lyα nebulae (e.g., Mori & Umemura 2006). The cold gas accretion rate and the strength of galactic superwinds are predicted to evolve with physical quantities related to the cosmic SFRD (e.g., Dekel et al. 2009; Kereš et al. 2009). The comparisons of the cosmic SFRD and LAB evolutions would provide useful hints that LABs are formed in these scenarios. However, it should be noted that the LAB selection method is not homogeneous in our comparison of NLAB at z ≃ 0–7. There is a possibility that the NLAB evolution from z ≃ 7 to z ≃ 3 is caused by the cosmological surface brightness dimming effect at high z. The cosmological surface brightness dimming would significantly affect the detection and selection completeness for LABs at high z. To confirm the NLAB evolution and quantitatively compare it with the cosmic SFRD, we need to homogenize the selection method for LABs at z ≃ 2–7 in any future HSC NB data. 7 Summary and conclusions We develop an unprecedentedly large catalog consisting of LAEs at z = 5.7 and 6.6 that are identified by the SILVERRUSH program using the first NB imaging data of the Subaru/HSC survey. The NB imaging data is about an order of magnitude larger than any other surveys for z ≃ 6–7 LAEs conducted to date. Our findings are as follows: We identify 2230 ≳ L* LAEs at z = 5.7 and 6.6 on the 13.8 and 21.2 deg2 sky, respectively. We confirm that the LAE catalog is reliable on the basis of 96 LAEs whose spectroscopic redshifts are already determined by this program (Shibuya et al. 2018) and previous studies (e.g., Mallery et al. 2012). The LAE catalog is introduced in this work, and published online. Using the large LAE catalog we derive the rest-frame Lyα EW distributions of LAEs at z ≃ 5.7 and ≃6.6 that are reasonably explained by the exponential profile. The best-fitting exponential (Gaussian) Lyα scale lengths are, on average, 153 ± 18 Å and 154 ± 15 Å (146 ± 24 Å and 139 ± 14 Å) at z ≃ 5.7 and z ≃ 6.6, for the Ultradeep and Deep fields, respectively, showing no significant evolution from z ≃ 5.7 to z ≃ 6.6. We find 45 and 230 LAEs at z ≃ 6.6 and z ≃ 5.7 with an LEW of $$EW_{\rm 0,\, Ly\alpha }^{\rm int}> 240\,$$Å, corrected for the IGM attenuation for Lyα. The fraction of LEW LAEs to all LAEs is ≃4% at z ≃ 6.6 and ≃21% at z ≃ 5.7. These LEW LAEs are candidates of young metal-poor galaxies and AGNs. We search for LABs that are LAEs with spatially extended Lyα emission that have profiles clearly distinguished from those of stellar objects at the ≳ 3σ level. In the search, we identify 11 LABs in the HSC NB images down to a surface brightness limit of ≃5–8 × 10−18 erg s−1 cm−2, which is as deep as data of previous studies. The number density of the LABs at z ≃ 6–7 is ∼10−7–10−6 Mpc−3, which is ∼10–100 times lower than those claimed for LABs at z ≃ 2–3, suggestive of disappearing LABs at z ≳ 6, although the selection methods are different in the low- and high-z LABs. It should be noted that Lyα EW scale length derivation methods and the LAB selections are not homogeneous in a redshift range of z ≃ 0–7. Using the future z ≃ 2.2, 5.7, 6.6, and 7.3 HSC NB data, we will systematically investigate the redshift evolution of Lyα EW scale lengths and NLAB at z ≃ 2–7 using homogeneous methods. Acknowledgements We would like to thank James Bosch, Richard S. Ellis, Masao Hayashi, Robert H. Lupton, and Michael A. Strauss for useful discussion and comments. We thank the anonymous referee for constructive comments and suggestions. This work is based on observations taken by the Subaru Telescope and the Keck telescope which are operated by the National Observatory of Japan. This work was supported by World Premier International Research Center Initiative (WPI Initiative), MEXT, Japan, KAKENHI (23244025) and (21244013) Grant-in-Aid for Scientific Research (A) through Japan Society for the Promotion of Science (JSPS), and an Advanced Leading Graduate Course for Photon Science grant. The NB816 filter was supported by Ehime University (PI: Y. Taniguchi). The NB921 filter was supported by KAKENHI (23244025) Grant-in-Aid for Scientific Research (A) through the Japan Society for the Promotion of Science (PI: M. Ouchi). NK is supported by the JSPS grant 15H03645. SY is supported by Faculty of Science, Mahidol University, Thailand and the Thailand Research Fund (TRF) through a research grant for new scholar (MRG5980153). The Hyper Suprime-Cam (HSC) collaboration includes the astronomical communities of Japan and Taiwan, and Princeton University. The HSC instrumentation and software were developed by the National Astronomical Observatory of Japan (NAOJ), the Kavli Institute for the Physics and Mathematics of the Universe (Kavli IPMU), the University of Tokyo, the High Energy Accelerator Research Organization (KEK), the Academia Sinica Institute for Astronomy and Astrophysics in Taiwan (ASIAA), and Princeton University. Funding was contributed by the FIRST program from Japanese Cabinet Office, the Ministry of Education, Culture, Sports, Science and Technology (MEXT), the Japan Society for the Promotion of Science (JSPS), Japan Science and Technology Agency (JST), the Toray Science Foundation, NAOJ, Kavli IPMU, KEK, ASIAA, and Princeton University. This paper makes use of software developed for the Large Synoptic Survey Telescope. We thank the LSST Project for making their code available as free software at ⟨http://dm.lsst.org⟩. The Pan-STARRS1 Surveys (PS1) have been made possible through contributions of the Institute for Astronomy, the University of Hawaii, the Pan-STARRS Project Office, the Max-Planck Society and its participating institutes, the Max Planck Institute for Astronomy, Heidelberg and the Max Planck Institute for Extraterrestrial Physics, Garching, The Johns Hopkins University, Durham University, the University of Edinburgh, Queen’s University Belfast, the Harvard-Smithsonian Center for Astrophysics, the Las Cumbres Observatory Global Telescope Network Incorporated, the National Central University of Taiwan, the Space Telescope Science Institute, the National Aeronautics and Space Administration under Grant No. NNX08AR22G issued through the Planetary Science Division of the NASA Science Mission Directorate, the National Science Foundation under Grant No. AST-1238877, the University of Maryland, and Eotvos Lorand University (ELTE) and the Los Alamos National Laboratory. This paper is based on data collected at the Subaru Telescope and retrieved from the HSC data archive system, which is operated by the Subaru Telescope and Astronomy Data Center at the National Astronomical Observatory of Japan. Appendix. Calculation of Lyα EW In this section, we describe the method used to calculate the EW0,Lyα values. The procedures and assumptions of this method are similar to those of e.g., Malhotra and Rhoads (2002), Dawson et al. (2004), Gronwall et al. (2007), and Kashikawa et al. (2011). For the calculation of EW0,Lyα, we assume that LAEs have a δ-function shaped Lyα line and a flat rest-frame UV continuum emission (i.e., βν = 0, where βν is the UV spectral slope per unit frequency). In such an LAE spectrum, the magnitude, m, of a waveband filter with a transmission curve, Tν, is described as follows:   \begin{eqnarray} 48.6 + m = -2.5 \log _{10} \frac{\int _0^\infty [f_{\rm c} + f_{\rm l} \delta (\nu - \nu _\alpha )] T_\nu {\rm d}\nu }{\int _0^\infty T_\nu {\rm d}\nu } , \end{eqnarray} (A1)where fl, fc, δ(ν), and να are the Lyα line flux, the flux density of the rest-frame UV continuum emission, the δ function, and the observed frequency of Lyα, respectively. Here we also assume that the Lyα line is located at 9215 Å (8177 Å), which is the central wavelength of the NB921 (NB816) filter, for z ≃ 6.6(z ≃ 5.7) LAEs. In this study, we do not take into account the IGM transmission for Lyα, if not specified. This is because the IGM transmission for Lyα highly depends on the Lyα line velocity offset from the systemic redshift (e.g., Hashimoto et al. 2013; Shibuya et al. 2014b). The numerator of the logarithm in equation (A1) corresponds to   \begin{eqnarray} && f_{\rm c} \int ^\infty _{\nu_{\rm c}} \exp {(-\tau _{\rm eff})} T_\nu {\rm d}\nu + f_{\rm c} \int _0^{\nu_{\rm c}} T_\nu {\rm d}\nu + f_{\rm l} T_\nu (\nu _\alpha ) \nonumber \\ && \quad = f_{\rm c} B + f_{\rm c} R + f_{\rm l} T_\nu (\nu _\alpha ). \end{eqnarray} (A2)In equation (A2), we use B and R that are defined by   \begin{eqnarray} B \equiv \int^\infty _{\nu_{\rm c}} \exp {(-\tau _{\rm eff})} T_\nu {\rm d}\nu , \end{eqnarray} (A3)  \begin{eqnarray} R \equiv \int _0^{\nu_{\rm c}} T_\nu {\rm d}\nu , \end{eqnarray} (A4)where τeff is the IGM optical depth calculated from the analytical models of Madau (1995). Using equations (A1) and (A2), we derive the flux density of the NB and BB filters, $$\overline{f^{\vphantom{i}}_{\rm NB}}$$ and $$\overline{f^{\vphantom{i}}_{\rm BB}}$$, as follows:   \begin{eqnarray} \overline{f^{\vphantom{i}}_{\rm NB}} = 10^{-0.4(m_{\rm NB} + 48.6)} = \frac{f_{\rm c} (B_{\rm NB} + R_{\rm NB}) + f_{\rm l} T_{\rm NB} (\nu _\alpha ) }{A_{\rm NB}}, \end{eqnarray} (A5)  \begin{eqnarray} \overline{f^{\vphantom{i}}_{\rm BB}} = 10^{-0.4(m_{\rm BB} + 48.6)} = \frac{f_{\rm c} (B_{\rm BB} + R_{\rm BB}) }{A_{\rm BB}}. \end{eqnarray} (A6)Here the following definition is used for the denominator of equations (A5) and (A6).   \begin{eqnarray} A \equiv \int_0^\infty T_\nu {\rm d}\nu , \end{eqnarray} (A7)The B, R, and A values with the subscripts of NB(BB) are calculated with the transmission curves of the NB(BB) filters, TNB(TBB). In this study, we use magnitudes of the y- and z-band filters which do not cover the wavelength of Lyα for z ≃ 6.6 and z ≃ 5.7 LAEs, respectively, indicating TBB(να) = 0. In the case that mBB is fainter than the 1σ limit, we use the 1σ limiting magnitude for the EW0,Lyα calculation. By combining the equations of $$\overline{f^{\vphantom{i}}_{\rm NB}}$$ and $$\overline{f^{\vphantom{i}}_{\rm BB}}$$, we obtain fc and fl,   \begin{eqnarray} f_{\rm c} = \frac{A_{\rm BB} \overline{f^{\vphantom{i}}_{\rm BB}}}{B_{\rm BB} + R_{\rm BB}} = \overline{f^{\vphantom{i}}_{\rm BB}}, \end{eqnarray} (A8)  \begin{eqnarray*} f_{\rm l} = \frac{A_{\rm NB}(B_{\rm BB} + R_{\rm BB})\overline{f^{\vphantom{i}}_{\rm NB}} - A_{\rm BB}(B_{\rm NB} + R_{\rm NB})\overline{f^{\vphantom{i}}_{\rm BB}}}{ (B_{\rm BB} + R_{\rm BB})T_{\rm NB} (\nu _\alpha ) } \end{eqnarray*}   \begin{eqnarray*} \hphantom{f_{\rm l}} = \frac{A_{\rm NB}\overline{f^{\vphantom{i}}_{\rm NB}} - (B_{\rm NB} + R_{\rm NB})\overline{f^{\vphantom{i}}_{\rm BB}}}{T_{\rm NB} (\nu_\alpha)} \end{eqnarray*}   \begin{eqnarray} \hphantom{f_{\rm l}} = a \overline{f^{\vphantom{i}}_{\rm NB}} - b \overline{f^{\vphantom{i}}_{\rm BB}}. \end{eqnarray} (A9)Note that BBB + RBB = ABB due to the negligible IGM absorption at the wavelengths of the BB filters. Here we define a and b as   \begin{eqnarray} a \equiv \frac{A_{\rm NB}}{T_{\rm NB} (\nu _\alpha )}, \end{eqnarray} (A10)  \begin{eqnarray} b \equiv \frac{B_{\rm NB} + R_{\rm NB}}{T_{\rm NB} (\nu _\alpha )}. \end{eqnarray} (A11)For the HSC NB921 and NB816 filters, the sets of the values are calculated to be (a, b) ≃ (4.7, 2.3) × 1012 and (a, b) ≃ (5.2, 2.7) × 1012, respectively. Using fc and fl, we calculate the EW0,Lyα values via   \begin{equation} EW_{\rm 0,\, Ly\alpha } = \frac{f_{\rm l}}{f_{\rm c}} \frac{c}{\nu ^2} \frac{1}{1+z}. \end{equation} (A12)To obtain the median values and uncertainties for EW0,Lyα, we perform MC simulations in a method similar to that of e.g., Shimasaku et al. (2006). In the simulation, we randomly generate a flux density value, $$\overline{f^{\vphantom{i}}_{\rm MC}}$$, following a Gaussian probability distribution with an average of $$\overline{f^{\vphantom{i}}}$$ and a dispersion of the 1σ sky background noise, $$\overline{f^{\vphantom{i}}_{\rm 1\sigma }}$$, for the NB and BB bands. Here we also randomize βν and να in Gaussian probability distributions with 1σ dispersions of Δβ = 0.2 and Δνα = FWHMNB/2.35, respectively, where FWHMNB is the FWHM of the NB filters. The dispersion of Δβ = 0.2 is typical for high-z galaxies (Bouwens et al. 2014). We calculate an EW0,Lyα value using $$\overline{f^{\vphantom{i}}_{\rm MC}}$$ for NB and BB via equation (A12). In this process, negative values of fc, fl, and EW0,Lyα are forced to be zero. Such a process is performed 1000 times for each object. During the iteration, a simulated EW0,Lyα value is discarded in the case that a BB − NB color does not meet the selection criteria of equations (1) and (2). Using the set of EW0,Lyα values obtained from the MC simulations, we calculate the median values and the 16- and 84-percentile errors for EW0,Lyα. Footnotes * Based on data obtained with the Subaru Telescope. The Subaru Telescope is operated by the National Astronomical Observatory of Japan. 1 ⟨https://www.naoj.org/Observing/Instruments/HSC/sensitivity.html⟩. 2 ⟨http://cos.icrr.u-tokyo.ac.jp/rush.html⟩. References Aihara H. et al.   2018a, PASJ , 70, S4 Aihara H. et al.   2018b, PASJ , 70, S8 Ajiki M. et al.   2002, ApJ , 576, L25 CrossRef Search ADS   Arrigoni Battaia F., Hennawi J. F., Prochaska J. X., Cantalupo S. 2015a, ApJ , 809, 163 CrossRef Search ADS   Arrigoni Battaia F., Yang Y., Hennawi J. F., Prochaska J. X., Matsuda Y., Yamada T., Hayashino T. 2015b, ApJ , 804, 26 CrossRef Search ADS   Axelrod T., Kantor J., Lupton R. H., Pierfederici F. 2010, in Proc. SPIE, 7740, Software and Cyberinfrastructure for Astronomy , ed. Radziwill N. M., Bridger A. ( Bellingham, WA: SPIE), 774015 Bertin E., Arnouts S. 1996, A&AS , 117, 393 CrossRef Search ADS   Blanc G. A. et al.   2011, ApJ , 736, 31 CrossRef Search ADS   Bond N. A., Gawiser E., Guaita L., Padilla N., Gronwall C., Ciardullo R., Lai K. 2012, ApJ , 753, 95 CrossRef Search ADS   Bosch J. et al.   2018, PASJ , 70, S5 Bouwens R. J. et al.   2014, ApJ , 793, 115 CrossRef Search ADS   Cai Z. et al.   2017, ApJ , 837, 71 CrossRef Search ADS   Cantalupo S., Arrigoni-Battaia F., Prochaska J. X., Hennawi J. F., Madau P. 2014, Nature , 506, 63 CrossRef Search ADS PubMed  Cantalupo S., Porciani C., Lilly S. J., Miniati F. 2005, ApJ , 628, 61 CrossRef Search ADS   Ciardullo R. et al.   2012, ApJ , 744, 110 CrossRef Search ADS   Cowie L. L., Barger A. J., Hu E. M. 2010, ApJ , 711, 928 CrossRef Search ADS   Cowie L. L., Hu E. M., Songaila A. 2011, ApJ , 735, L38 CrossRef Search ADS   Dawson S. et al.   2004, ApJ , 617, 707 CrossRef Search ADS   Dekel A. et al.   2009, Nature , 457, 451 CrossRef Search ADS PubMed  Dressler A., Martin C. L., Henry A., Sawicki M., McCarthy P. 2011, ApJ , 740, 71 CrossRef Search ADS   Duval F., Schaerer D., Östlin G., Laursen P. 2014, A&A , 562, A52 CrossRef Search ADS   Erb D. K., Bogosavljević M., Steidel C. C. 2011, ApJ , 740, L31 CrossRef Search ADS   Finkelstein S. L., Rhoads J. E., Malhotra S., Grogin N., Wang J. 2008, ApJ , 678, 655 CrossRef Search ADS   Finkelstein S. L., Rhoads J. E., Malhotra S., Pirzkal N., Wang J. 2007, ApJ , 660, 1023 CrossRef Search ADS   Gawiser E. et al.   2007, ApJ , 671, 278 CrossRef Search ADS   Gronwall C. et al.   2007, ApJ , 667, 79 CrossRef Search ADS   Guaita L. et al.   2011, ApJ , 733, 114 CrossRef Search ADS   Haiman Z., Abel T., Rees M. J. 2000, ApJ , 534, 11 CrossRef Search ADS   Hansen M., Oh S. P. 2006, MNRAS , 367, 979 CrossRef Search ADS   Harikane Y. et al.   2018, PASJ , 70, S11 Hashimoto T., Ouchi M., Shimasaku K., Ono Y., Nakajima K., Rauch M., Lee J., Okamura S. 2013, ApJ , 765, 70 CrossRef Search ADS   Hashimoto T., Ouchi M., Shimasaku K., Schaerer D., Nakajima K., Shibuya T., Ono Y., Rauch M., Goto R. 2017, MNRAS , 465, 1543 CrossRef Search ADS   Hayes M., Scarlata C., Siana B. 2011, Nature , 476, 304 CrossRef Search ADS PubMed  Hennawi J. F., Prochaska J. X., Cantalupo S., Arrigoni-Battaia F. 2015, Science , 348, 779 CrossRef Search ADS PubMed  Hu E. M., Cowie L. L., Barger A. J., Capak P., Kakazu Y., Trouille L. 2010, ApJ , 725, 394 CrossRef Search ADS   Ivezic Z. et al.   2008, arXiv:0805.2366 Jurić M. et al.   2015, arXiv:1512.07914 Kashikawa N. et al.   2006, ApJ , 648, 7 CrossRef Search ADS   Kashikawa N. et al.   2011, ApJ , 734, 119 CrossRef Search ADS   Kashikawa N. et al.   2012, ApJ , 761, 85 CrossRef Search ADS   Kawanomoto S. et al.   2017, PASJ , submitted Keel W. C., White R. E. III, Chapman S., Windhorst R. A. 2009, AJ , 138, 986 CrossRef Search ADS   Kereš D., Katz N., Fardal M., Davé R., Weinberg D. H. 2009, MNRAS , 395, 160 CrossRef Search ADS   Kodaira K. et al.   2003, PASJ , 55, L17 CrossRef Search ADS   Kojima T., Ouchi M., Nakajima K., Shibuya T., Harikane Y., Ono Y. 2017, PASJ , 69, 44 CrossRef Search ADS   Konno A. et al.   2014, ApJ , 797, 16 CrossRef Search ADS   Konno A. et al.   2018, PASJ , 70, S16 Kusakabe H., Shimasaku K., Nakajima K., Ouchi M. 2015, ApJ , 800, L29 CrossRef Search ADS   Laursen P., Duval F., Östlin G. 2013, ApJ , 766, 124 CrossRef Search ADS   Laursen P., Sommer-Larsen J. 2007, ApJ , 657, L69 CrossRef Search ADS   Laursen P., Sommer-Larsen J., Andersen A. C. 2009, ApJ , 704, 1640 CrossRef Search ADS   Lilly S. J., Le Fevre O., Hammer F., Crampton D. 1996, ApJ , 460, L1 CrossRef Search ADS   Madau P. 1995, ApJ , 441, 18 CrossRef Search ADS   Madau P., Dickinson M. 2014, ARA&A , 52, 415 CrossRef Search ADS   Madau P., Ferguson H. C., Dickinson M. E., Giavalisco M., Steidel C. C., Fruchter A. 1996, MNRAS , 283, 1388 CrossRef Search ADS   Magnier E. A. et al.   2013, ApJS , 205, 20 CrossRef Search ADS   Malhotra S., Rhoads J. E. 2002, ApJ , 565, L71 CrossRef Search ADS   Malhotra S., Rhoads J. E. 2004, ApJ , 617, L5 CrossRef Search ADS   Malhotra S., Rhoads J. E. 2006, ApJ , 647, L95 CrossRef Search ADS   Mallery R. P. et al.   2012, ApJ , 760, 128 CrossRef Search ADS   Mas-Ribas L., Dijkstra M. 2016, ApJ , 822, 84 CrossRef Search ADS   Mas-Ribas L., Dijkstra M., Hennawi J. F., Trenti M., Momose R., Ouchi M. 2017, ApJ , 841, 19 CrossRef Search ADS   Matsuda Y. et al.   2004, AJ , 128, 569 CrossRef Search ADS   Matsuda Y. et al.   2009, MNRAS , 400, L66 CrossRef Search ADS   Matsuda Y. et al.   2011, MNRAS , 410, L13 CrossRef Search ADS   Matthee J., Sobral D., Santos S., Röttgering H., Darvish B., Mobasher B. 2015, MNRAS , 451, 400 CrossRef Search ADS   Momose R., Ouchi M., Nakajima K., Ono Y., Shibuya T., Shimasaku K., Yuma S., Mori M., Umemura M. 2016, MNRAS , 457, 2318 CrossRef Search ADS   Mori M., Umemura M. 2006, Nature , 440, 644 CrossRef Search ADS PubMed  Murayama T. et al.   2007, ApJS , 172, 523 CrossRef Search ADS   Nagao T. et al.   2008, ApJ , 680, 100 CrossRef Search ADS   Nakajima K., Ouchi M. 2014, MNRAS , 442, 900 CrossRef Search ADS   Nakajima K., Ouchi M., Shimasaku K., Hashimoto T., Ono Y., Lee J. C. 2013, ApJ , 769, 3 CrossRef Search ADS   Nakajima K. et al.   2012, ApJ , 745, 12 CrossRef Search ADS   Neufeld D. A. 1991, ApJ , 370, L85 CrossRef Search ADS   Nilsson K. K., Tapken C., Møller P., Freudling W., Fynbo J. P. U., Meisenheimer K., Laursen P., Östlin G. 2009, A&A , 498, 13 CrossRef Search ADS   Oke J. B., Gunn J. E. 1983, ApJ , 266, 713 CrossRef Search ADS   Ono Y. et al.   2010, MNRAS , 402, 1580 CrossRef Search ADS   Ono Y. et al.   2018, PASJ , 70, S10 Ota K. et al.   2017, ApJ , 844, 85 CrossRef Search ADS   Ouchi M. et al.   2008, ApJS , 176, 301 CrossRef Search ADS   Ouchi M. et al.   2009, ApJ , 696, 1164 CrossRef Search ADS   Ouchi M. et al.   2010, ApJ , 723, 869 CrossRef Search ADS   Ouchi M. et al.   2018, PASJ , 70, S13 Oyarzún G. A., Blanc G. A., González V., Mateo M., Bailey J. I., III, Finkelstein S. L., Lira P., Crane J. D., Olszewski E. W. 2016, ApJ , 821, L14 CrossRef Search ADS   Planck Collaboration 2016, A&A , 594, A13 CrossRef Search ADS   Prescott M. K. M. et al.   2012b, ApJ , 752, 86 CrossRef Search ADS   Prescott M. K. M., Dey A., Jannuzi B. T. 2009, ApJ , 702, 554 CrossRef Search ADS   Prescott M. K. M., Dey A., Jannuzi B. T. 2012a, ApJ , 748, 125 CrossRef Search ADS   Prescott M. K. M., Dey A., Jannuzi B. T. 2013, ApJ , 762, 38 CrossRef Search ADS   Prescott M. K. M., Martin C. L., Dey A. 2015, ApJ , 799, 62 CrossRef Search ADS   Rhoads J. E., Malhotra S. 2001, ApJ , 563, L5 CrossRef Search ADS   Saito T., Shimasaku K., Okamura S., Ouchi M., Akiyama M., Yoshida M. 2006, ApJ , 648, 54 CrossRef Search ADS   Salpeter E. E. 1955, ApJ , 121, 161 CrossRef Search ADS   Schlafly E. F. et al.   2012, ApJ , 756, 158 CrossRef Search ADS   Shibuya T., Kashikawa N., Ota K., Iye M., Ouchi M., Furusawa H., Shimasaku K., Hattori T. 2012, ApJ , 752, 114 CrossRef Search ADS   Shibuya T., Ouchi M., Kubo M., Harikane Y. 2016, ApJ , 821, 72 CrossRef Search ADS   Shibuya T., Ouchi M., Nakajima K., Yuma S., Hashimoto T., Shimasaku K., Mori M., Umemura M. 2014a, ApJ , 785, 64 CrossRef Search ADS   Shibuya T. et al.   2014b, ApJ , 788, 74 CrossRef Search ADS   Shibuya T. et al.   2018, PASJ , 70, S15 Shimasaku K. et al.   2006, PASJ , 58, 313 CrossRef Search ADS   Sobral D., Matthee J., Darvish B., Schaerer D., Mobasher B., Röttgering H. J. A., Santos S., Hemmati S. 2015, ApJ , 808, 139 CrossRef Search ADS   Stark D. P., Ellis R. S., Ouchi M. 2011, ApJ , 728, L2 CrossRef Search ADS   Steidel C. C., Adelberger K. L., Shapley A. E., Pettini M., Dickinson M., Giavalisco M. 2000, ApJ , 532, 170 CrossRef Search ADS   Taniguchi Y. et al.   2005, PASJ , 57, 165 CrossRef Search ADS   Taniguchi Y. et al.   2009, ApJ , 701, 915 CrossRef Search ADS   Taniguchi Y., Shioya Y. 2000, ApJ , 532, L13 CrossRef Search ADS   Tonry J. L. et al.   2012, ApJ , 750, 99 CrossRef Search ADS   Toshikawa J. et al.   2018, PASJ , 70, S12 Yajima H., Li Y., Zhu Q. 2013, ApJ , 773, 151 CrossRef Search ADS   Yajima H., Li Y., Zhu Q., Abel T., Gronwall C., Ciardullo R. 2014, MNRAS , 440, 776 CrossRef Search ADS   Yang Y., Zabludoff A., Eisenstein D., Davé R. 2010, ApJ , 719, 1654 CrossRef Search ADS   Yang Y., Zabludoff A., Tremonti C., Eisenstein D., Davé R. 2009, ApJ , 693, 1579 CrossRef Search ADS   Zheng Z., Cen R., Trac H., Miralda-Escudé J. 2010, ApJ , 716, 574 CrossRef Search ADS   Zheng Z., Wallace J. 2014, ApJ , 794, 116 CrossRef Search ADS   Zheng Z.-Y., Wang J.-X., Malhotra S., Rhoads J. E., Finkelstein S. L., Finkelstein K. 2014, MNRAS , 439, 1101 CrossRef Search ADS   Zheng Z.-Y. et al.   2017, ApJ , 842, L22 CrossRef Search ADS   © The Author(s) 2017. Published by Oxford University Press on behalf of the Astronomical Society of Japan. All rights reserved. For Permissions, please email: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Publications of the Astronomical Society of Japan Oxford University Press

Loading next page...
 
/lp/ou_press/silverrush-ii-first-catalogs-and-properties-of-2000-ly-emitters-and-CDZsE2MYNn
Publisher
Oxford University Press
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of the Astronomical Society of Japan. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
0004-6264
eISSN
2053-051X
D.O.I.
10.1093/pasj/psx122
Publisher site
See Article on Publisher Site

Abstract

Abstract We present an unprecedentedly large catalog consisting of 2230 ≳ L* Lyα emitters (LAEs) at z = 5.7 and 6.6 on the 13.8 and 21.2 deg2 sky, respectively, that are identified by the SILVERRUSH program with the first narrow-band imaging data of the Hyper Suprime-Cam (HSC) survey. We confirm that the LAE catalog is reliable on the basis of 96 LAEs whose spectroscopic redshifts are already determined by this program and previous studies. This catalogue is also available online. Based on this catalogue, we derive the rest-frame Lyα equivalent-width distributions of LAEs at z ≃ 5.7–6.6 that are reasonably explained by the exponential profiles with scale lengths of ≃ 120–170 Å, showing no significant evolution from z ≃ 5.7 to z ≃ 6.6. We find that 275 LAEs with large equivalent widths (LEWs) of >240 Å are candidates of young metal poor galaxies and AGNs. We also find that the fraction of LEW LAEs to all LAEs is 4% and 21% at z ≃ 5.7 and z ≃ 6.6, respectively. Our LAE catalog includes 11 Lyα blobs (LABs) that are LAEs with spatially extended Lyα emission with a profile that is clearly distinguished from those of stellar objects at the ≳3σ level. The number density of the LABs at z = 6–7 is ∼10−7–10−6 Mpc−3, being ∼10–100 times lower than those claimed for LABs at z ≃ 2–3, suggestive of disappearing LABs at z ≳ 6, albeit with the different selection methods and criteria for the low and high-z LABs. 1 Introduction Lyα emitters (LAEs) are one of the important populations of high-z star-forming galaxies in the paradigm of galaxy formation and evolution. Such galaxies are thought to be typically young (an order of 100 Myr; e.g., Finkelstein et al. 2007; Gawiser et al. 2007), compact (an effective radius of <1 kpc; e.g., Taniguchi et al. 2009; Bond et al. 2012), less massive (a stellar mass of 108–109 M⊙; e.g., Ono et al. 2010; Guaita et al. 2011), metal-poor (≃0.1 of the solar metallicity; e.g., Nakajima et al. 2012, 2013; Nakajima & Ouchi 2014; Kojima et al. 2017), less dusty than Lyman break galaxies (e.g., Blanc et al. 2011; Kusakabe et al. 2015), and possible progenitors of Milky Way mass galaxies (e.g., Dressler et al. 2011). In addition, LAEs are used to probe the cosmic reionizaiton, because ionizing photons that escape from the large number of massive stars formed in LAEs contribute to the ionization of the intergalactic medium (IGM; e.g., Rhoads & Malhotra 2001; Malhotra & Rhoads 2006; Shimasaku et al. 2006; Kashikawa et al. 2006, 2011; Ouchi et al. 2008, 2010; Cowie et al. 2010; Hu et al. 2010; Shibuya et al. 2012; Konno et al. 2014; Matthee et al. 2015; Ota et al. 2017; Zheng et al. 2017). LAEs have been surveyed by imaging observations with dedicated narrow-band (NB) filters for a prominent redshifted Lyα emission (e.g., Ajiki et al. 2002; Malhotra & Rhoads 2004; Kodaira et al. 2003; Taniguchi et al. 2005; Gronwall et al. 2007; Erb et al. 2011; Ciardullo et al. 2012). In a large LAE sample constructed by the NB observations, two rare Lyα-emitting populations have been identified: large equivalent width (LEW) LAEs, and spatially extended Lyα LAEs, Lyα blobs (LABs). LEW LAEs are objects with a large Lyα equivalent width (EW) of ≳240 Å which are not reproduced with the normal Salpeter (1955) stellar initial mass function (e.g., Malhotra & Rhoads 2002). Such an LEW is expected to have originated from complicated physical processes such as (i) photoionization by young and/or low-metallicity star-formation, (ii) photoionization by active galactic nucleus (AGN), (iii) photoionization by external UV sources (QSO fluorescence), (iv) collisional excitation due to strong outflows (shock heating), (v) collisional excitation due to gas inflows (gravitational cooling), and (vi) clumpy ISM (see e.g., Hashimoto et al. 2017). The highly complex radiative transfer of Lyα in the interstellar medium (ISM) makes it difficult to understand the Lyα emitting mechanism (Neufeld 1991; Hansen & Oh 2006; Finkelstein et al. 2008; Laursen et al. 2009, 2013; Laursen & Sommer-Larsen 2007; Zheng et al. 2010; Yajima et al. 2014; Duval et al. 2014; Zheng & Wallace 2014). LABs are spatially extended Lyα gaseous nebulae in the high-z universe (e.g., Steidel et al. 2000; Matsuda et al. 2004, 2009, 2011; Prescott et al. 2009, 2012a, 2012b, 2013, 2015; Cantalupo et al. 2014; Arrigoni Battaia et al. 2015a, 2015b; Hennawi et al. 2015; Cai et al. 2017). The origins of LABs (LAEs with a diameter ≃20–400 kpc) are also explained by several mechanisms: (1) resonant scattering of Lyα photons emitted from central sources in dense and extended neutral hydrogen clouds (e.g., Hayes et al. 2011), (2) cooling radiation from gravitationally heated gas in collapsed halos (e.g., Haiman et al. 2000), (3) shock heating by galactic superwind originating from starbursts and/or AGN activity (e.g., Taniguchi & Shioya 2000), (4) galaxy major mergers (e.g., Yajima et al. 2013), and (5) photoionization by external UV sources (QSO fluorescence; e.g., Cantalupo et al. 2005). Moreover, LABs have often been discovered in over-dense regions at z ≃ 2–3 (e.g., Yang et al. 2009, 2010; Matsuda et al. 2011). Thus, such LABs could be closely related to the galaxy environments, and might be linkd to the formation mechanisms of central massive galaxies in galaxy protoclusters. During the previous decades, Suprime-Cam (SCam) on the Subaru telescope has led the world on identifying such rare Lyα-emitting populations at z ≳ 6 (LEW LAEs; e.g., Nagao et al. 2008; Kashikawa et al. 2012; LABs; e.g., Ouchi et al. 2009; Sobral et al. 2015). However, the formation mechanisms of these rare Lyα-emitting populations are still controversial due to the small statistics. While LEW LAEs and LABs at z ≃ 2–5 have been studied intensively with a sample of ≳ 100 sources, only a few sources have been found so far at z ≳ 6. Large-area NB data are required to carry out a statistical study on LEW LAEs and LABs at z ≳ 6. In 2014 March, the Subaru telescope started a large-area NB survey using a new wide field of view (FoV) camera, the Hyper Suprime-Cam (HSC) as part of a Subaru strategic program (SSP: Aihara et al. 2018a). In the five-year project, HSC, equipped with four NB filters of NB387, NB816, NB921, and NB101, will survey for LAEs at z ≃ 2.2, 5.7, 6.6, and 7.3, respectively. The HSC-SSP NB survey data consist of two layers; Ultradeep (UD), covering two fields (UD-COSMOS, UD-SXDS), and Deep (D), covering four fields (D-COSMOS, D-SXDS, D-DEEP2-3, D-ELAIS-N1). The NB816, NB921, and NB101 images will be taken for the UD fields. The NB387, NB816, and NB921 observations will be conducted in 15 HSC-pointing D fields. Using the large HSC NB data complemented by optical and near-infrared (NIR) spectroscopic observations, we launch a research project for Lyα-emitting objects: the Systematic Identification of LAEs for Visible Exploration and Reionization Research Using Subaru HSC (SILVERRUSH). The large LAE samples provided by SILVERRUSH enable us to investigate, e.g., LAE clustering (Ouchi et al. 2018), LEW LAEs and LABs (this work), the spectroscopic properties of bright LAEs (Shibuya et al. 2018), Lyα luminosity functions (Konno et al. 2018), and LAE overdensity (R. Higuchi et al. in preparation). The LAE survey strategy is given by Ouchi et al. (2018). This program is one of the twin programs. Another program is the study for dropouts, the Great Optically Luminous Dropout Research Using Subaru HSC (GOLDRUSH), which is detailed in Ono et al. (2018), Harikane et al. (2018), and Toshikawa et al. (2018). This is the second paper in the SILVERRUSH project. In this paper, we present LAE selection processes and machine-readable catalogs of the LAE candidates at z ≃ 5.7–6.6. Using the large LAE sample obtained from the first HSC NB data, we examine the redshift evolutions of Lyα EW distributions and LAB number density. This paper has the following structure. In section 2, we describe the details of the SSP HSC data. Section 3 presents the LAE selection processes. In section 4, we check the reliability of our LAE selection. Section 5 presents Lyα EW distributions and LABs at z ≃ 6–7. In section 6, we discuss the physical origins of LEW LAEs and LABs. We summarize our findings in section 7. Throughout this page, we adopt the concordance cosmology with (Ωm, ΩΛ, h) = (0.3, 0.7, 0.7) (Planck Collaboration 2016). All magnitudes are given in the AB system (Oke & Gunn 1983). 2 HSC-SSP imaging data We use the HSC-SSP S16A data products of g, r, i, z, and y broad-band (BB: Kawanomoto 2017), NB921, and NB816 (Ouchi et al. 2018) images that were obtained between 2014–2016. It should be noted that this HSC-SSP S16A data set is significantly larger than that of the first-data release in Aihara et al. (2018b). The NB921 (NB816) filter has a central wavelength of λc = 9215 Å (8177 Å) and an FWHM of Δλ = 135 Å (113 Å), all of which are the area-weighted mean values. The NB921 and NB816 filters trace the redshifted Lyα emission lines at z = 6.580 ± 0.056 and z = 5.726 ± 0.046, respectively. The NB filter transmission curves are shown in figure 1. The central wavelength, FWHM, and the bandpass shape for these NB filters are almost uniform over the HSC FoV. The deviation of the λc and FWHM values are typically within ≃0.3% and ≃10%, respectively. Thus, we use the area-weighted mean transmission curves in this study. The detailed specifications of these NB filters are given in Ouchi et al. (2018). Fig. 1. View largeDownload slide Filter transmission curves of the NB and BB filters. The red and blue curves represent the NB921 and NB816 filters, respectively. The red and blue ticks show the NB central wavelengths with the same color coding as for the NB filter transmission curves. The black solid curves indicate the i-, z-, and y-band filters, from left to right. The gray line denotes the OH sky lines. The bandpass of these NB and BB filters corresponds to the area-weighted mean transmission curves.1 The transmission curves are derived by taking into account (1) the quantum efficiency of CCD, the transmittance of (2) the dewar window and (3) the HSC primary focus unit (POpt2), (4) the reflectivity of the primary mirror, and (5) the sky transparency (see Aihara et al. 2018b). The upper x-axis corresponds to the redshift of Lyα. (Color online) Fig. 1. View largeDownload slide Filter transmission curves of the NB and BB filters. The red and blue curves represent the NB921 and NB816 filters, respectively. The red and blue ticks show the NB central wavelengths with the same color coding as for the NB filter transmission curves. The black solid curves indicate the i-, z-, and y-band filters, from left to right. The gray line denotes the OH sky lines. The bandpass of these NB and BB filters corresponds to the area-weighted mean transmission curves.1 The transmission curves are derived by taking into account (1) the quantum efficiency of CCD, the transmittance of (2) the dewar window and (3) the HSC primary focus unit (POpt2), (4) the reflectivity of the primary mirror, and (5) the sky transparency (see Aihara et al. 2018b). The upper x-axis corresponds to the redshift of Lyα. (Color online) Table 1 summarizes the survey areas, exposure time, and depth of the HSC-SSP S16A NB data. The current HSC-SSP S16A NB data covers UD-COSMOS, UD-SXDS, D-COSMOS, D-DEEP2-3, and D-ELAIS-N1 for z ≃ 6.6, and UD-COSMOS, UD-SXDS, D-DEEP2-3, and D-ELAIS-N1 for z ≃ 5.7. The effective survey areas of the NB921 and NB816 images are 21.2 and 13.8 arcmin2, corresponding to survey volumes of ≃1.9 × 107 and ≃1.2 × 107 Mpc3, respectively. The area of these HSC NB fields are covered by the observations of all the BB filters. The typical limiting magnitudes of BB filters are g ≃ 26.9, r ≃ 26.5, r ≃ 26.3, z ≃ 25.7, and y ≃ 25.0 (g ≃ 26.6, r ≃ 26.1, r ≃ 25.9, z ≃ 25.2, and y ≃ 24.4) in a 1$${^{\prime\prime}_{.}}$$5 aperture at 5σ for the UD(D) fields. The FWHM size of the point spread function in the HSC images is typically ≃0$${^{\prime\prime}_{.}}$$8 (Aihara et al. 2018b). Table 1. Properties of the HSC-SSP S16A NB data.* Field  RA  Dec  Area  Texp  mlim(5σ, 1$${^{\prime\prime}_{.}}$$5ϕ)  NLAE,ALL  NLAE,F    (J2000.0)  (J2000.0)  (deg2)  (hr)  (mag)      (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  NB921 (z ≃ 6.6)  UD-COSMOS  10h00m28s  +02°12΄21″  2.05  11.25  25.6  338  116  UD-SXDS  02h18m00s  −05°00΄00″  2.02  7.25  25.5  58  23  D-COSMOS  10h00m60s  +02°13΄53″  5.31  2.75  25.3  244†  47†  D-DEEP2-3  23h30m22s  −00°44΄38″  5.76  1.00  24.9  164  35  D-ELAIS-N1  16h10m00s  +54°17΄51″  6.08  1.75  25.3  349  48  Total  —  —  21.2  24.00  —  1153  269  NB816 (z ≃ 5.7)  UD-COSMOS  10h00m28s  +02°12΄21″  1.97  5.50  25.7  201  176  UD-SXDS  02h18m00s  −05°00΄00″  1.93  3.75  25.5  224  188  D-DEEP2-3  23h30m22s  −00°44΄38″  4.37  1.00  25.2  423  282  D-ELAIS-N1  16h10m00s  +54°17΄51″  5.56  1.00  25.3  229  130  Total  —  —  13.8  11.25  —  1077  776  Field  RA  Dec  Area  Texp  mlim(5σ, 1$${^{\prime\prime}_{.}}$$5ϕ)  NLAE,ALL  NLAE,F    (J2000.0)  (J2000.0)  (deg2)  (hr)  (mag)      (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  NB921 (z ≃ 6.6)  UD-COSMOS  10h00m28s  +02°12΄21″  2.05  11.25  25.6  338  116  UD-SXDS  02h18m00s  −05°00΄00″  2.02  7.25  25.5  58  23  D-COSMOS  10h00m60s  +02°13΄53″  5.31  2.75  25.3  244†  47†  D-DEEP2-3  23h30m22s  −00°44΄38″  5.76  1.00  24.9  164  35  D-ELAIS-N1  16h10m00s  +54°17΄51″  6.08  1.75  25.3  349  48  Total  —  —  21.2  24.00  —  1153  269  NB816 (z ≃ 5.7)  UD-COSMOS  10h00m28s  +02°12΄21″  1.97  5.50  25.7  201  176  UD-SXDS  02h18m00s  −05°00΄00″  1.93  3.75  25.5  224  188  D-DEEP2-3  23h30m22s  −00°44΄38″  4.37  1.00  25.2  423  282  D-ELAIS-N1  16h10m00s  +54°17΄51″  5.56  1.00  25.3  229  130  Total  —  —  13.8  11.25  —  1077  776  *(1) Field. (2) Right ascension. (3) Declination. (4) Survey area with the HSC SQL parameters in table 2. (5) Total exposure time of the NB imaging observation. (6) Limiting magnitude of the NB image defined by a 5σ sky noise in a 1$${^{\prime\prime}_{.}}$$5 diameter circular aperture. (7) Number of the LAE candidates in the ALL (unforced + forced) catalog. (8) Number of the LAE candidates in the forced catalog. †This value of NLAE,ALL (NLAE,F) includes 30 (7) LAEs selected in UD-COSMOS. View Large The HSC images were reduced with the HSC pipeline, hscPipe 4.0.2 (Bosch et al. 2018), which is a code from the Large Synoptic Survey Telescope (LSST) software pipeline (Ivezic et al. 2008; Axelrod et al. 2010; Jurić et al. 2015). The HSC pipeline performs CCD-by-CCD reduction, calibration for astrometry, and photometric zero-point determination. The pipeline then conducts mosaic-stacking that combines reduced CCD images into a large coadd image, and creates source catalogs by detecting and measuring sources on the coadd images. The photometric calibration is carried out with the PanSTARRS1 processing version 2 imaging survey data (Magnier et al. 2013; Schlafly et al. 2012; Tonry et al. 2012). The details of the HSC-SSP survey, data reduction, and source detection and photometric catalog construction are provided in Aihara et al. (2018a, 2018b) and Bosch et al. (2018). In the HSC images, source detection and photometry were carried out via two methods: unforced and forced. The unforced photometry is a method that performs measurements of coordinates, shapes, and fluxes individually in each band image for an object. The forced photometry is a method that carries out photometry by fixing centroid and shape determined in a reference band and applying them to all the other bands. The algorithm of the forced detection and photometry is similar to the double-image mode of SExtractor (Bertin & Arnouts 1996) which is used in most of the previous studies for high-z galaxies. According to which of them depends on magnitudes, S/N, positions, and profiles for detected sources, either the BB or the NB filter is regarded as a reference band. For merging the catalogs of each band, the object-matching radius is not a specific value; it depends on the area of a region with a >5σ sky noise level. We refer the detailed algorithm to choose the reference filter and filter priority to Bosch et al. (2018). In the hscPipe detection and photometry, an NB filter is basically chosen as a reference band for the NB-bright and BB-faint sources such as LAEs. However, a BB filter is used as a reference band in the case that sources are bright in the BB image. The current version of hscPipe has not implemented the NB-reference forced photometry for BB-bright sources. In this specification, there is a possibility that we miss BB-bright sources with a spatial offset between centroids of BB and NB by using only the forced photometry. Thus, we combine the unforced or forced photometry for BB − NB colors to identify such BB-bright objects with a spatial offset between centroids of BB and NB (e.g., Shibuya et al. 2014a). See section 3 for details of the LAE selection criteria. We use cmodel magnitudes for estimating total magnitudes of sources. The cmodel magnitude is a weighted combination of exponential and de Vaucouleurs fits to the light profiles of each object. The detailed algorithm of the cmodel photometry are presented in Bosch et al. (2018). To measure the S/N values for source detections, we use 1$${^{\prime\prime}_{.}}$$5-diameter aperture magnitudes. 3 LAE selection Using the HSC data, we perform a selection for LAEs at z ≃ 6.6 and ≃ 5.7. Basically, we select objects showing a significant flux excess in the NB images and a spectral break at the wavelength of redshifted Lyα emission. In this study, we create two LAE catalogs: an HSC LAE ALL (forced+unforced) catalog and an HSC LAE forced catalog. The HSC LAE ALL catalog is constructed using a combination of the forced and unforced photometry. We use this HSC LAE ALL catalog for identifying objects with a spatial offset between centroids of BB and NB (see section 2). On the other hand, the HSC LAE forced catalog consists of LAEs that meet only the selection criteria of the forced photometry. We use this HSC LAE forced catalog for statistical studies of LAEs [e.g., Lyα luminosity functions (LFs)]. The HSC LAE forced catalog is a subsample of the ALL one. Figure 2 shows the flow chart of the LAE selection process. We carry out the following processes: (1) SQL selection, (2) visual inspections of the object images, (3) rejections of variable and moving objects with the multi-epoch images, and (4) forced selection. The details are described as below. SQLselection: We retrieve detection and photometric catalogs from postgreSQL database tables. Using SQL scripts, we select objects meeting the following criteria of (i) magnitude and color selections and (ii) hscPipe parameters and flags. Magnitude and color selection: To identify objects with an NB magnitude excess in the HSC catalog, we apply magnitude and color selection criteria that are similar to e.g., Ouchi et al. (2008, 2010):   \begin{eqnarray} &&{\mathit {NB921}^\mathrm{ap}_\mathrm{frc} < \mathit {NB921}_{5\sigma }} \nonumber \\ && \&\&\ \left({\it g}_\mathrm{frc} > {\it g}_{3\sigma } \,{\|}\, {\it g}^\mathrm{ap}_\mathrm{frc} > {\it g}_{3\sigma } \right) \nonumber \\ && \&\&\ \left({\it r}_\mathrm{frc} > {\it r}_{3\sigma } \,{\|}\, {\it r}^\mathrm{ap}_\mathrm{frc} > {\it r}_{3\sigma } \right) \nonumber \\ && \&\&\ \left( {\it z}_\mathrm{frc} \!-\! {\it NB921}_\mathrm{frc} \!>\! 1.0 \,{\|}\, {\it z}_\mathrm{unf} \!-\! {\it NB921}_\mathrm{unf} \!>\! 1.0 \right) \nonumber \\ && \&\&\ \left\lbrace [ ({\it z}_\mathrm{frc} < {\it z}_{3\sigma } \,{\|}\, {\it z}^\mathrm{ap}_\mathrm{frc} < {\it z}_{3\sigma }) \right. \nonumber \\ &&\qquad\quad \&\&\ ({\it i}_\mathrm{frc} - {\it z}_\mathrm{frc} > 1.3 \,{\|}\, {\it i}_\mathrm{unf} - {\it z}_\mathrm{unf} > 1.3) ] \nonumber\\ &&\qquad\quad \left. {\|}\, ({\it z}_\mathrm{frc} > {\it z}_{3\sigma } \,{\|}\, {\it z}^\mathrm{ap}_\mathrm{frc} > {\it z}_{3\sigma }) \right\rbrace \!, \end{eqnarray} (1)for z ≃ 6.6, and   \begin{eqnarray} &&{\mathit {NB816}^\mathrm{ap}_\mathrm{frc} < {\it NB816}_{5\sigma }} \nonumber \\ && \&\&\ \left({\it g}_\mathrm{frc} > {\it g}_{3\sigma } \,{\|}\, {\it g}^\mathrm{ap}_\mathrm{frc} > {\it g}_{3\sigma } \right) \nonumber \\ && \&\&\ \left( {\it i}_\mathrm{frc} \!-\! {\it NB816}_\mathrm{frc} > 1.2 \,{\|}\, {\it i}_\mathrm{unf} \!-\! {\it NB816}_\mathrm{unf} \!>\! 1.2 \right) \nonumber \\ && \&\&\ \left\lbrace [ ({\it r}_\mathrm{frc} < {\it r}_{3\sigma } \,{\|}\, {\it r}^\mathrm{ap}_\mathrm{frc} < {\it r}_{3\sigma }) \right. \nonumber \\ &&\qquad\quad \&\&\ ({\it r}_\mathrm{frc} - {\it i}_\mathrm{frc} > 1.0 \,{\|}\, {\it r}_\mathrm{unf} - {\it i}_\mathrm{unf} > 1.0) ] \nonumber \\ &&\qquad\quad \left. {\|}\, ({\it r}_\mathrm{frc} > {\it r}_{3\sigma } \,{\|}\, {\it r}^\mathrm{ap}_\mathrm{frc} > {\it r}_{3\sigma }) \right\rbrace \!, \end{eqnarray} (2)for z ≃ 5.7, where the indices of “frc” and “unf” represent the forced and unforced photometry, respectively. The subscript of 5σ(3σ) indicates the 5σ(3σ) limiting magnitude for a given filter. The values with and without a superscript of “ap” indicate the aperture and total magnitudes, respectively. These magnitudes are derived with the hscPipe software (see section 2; Bosch et al. 2018). The limits of the $$i-\mathit {NB816}$$ and $$z- \mathit {NB921}$$ colors are the same as those of Ouchi et al. (2008, 2010), respectively. To exploit the survey capability of HSC in identifying rare objects, we use the 3σ g and r limiting magnitude (instead of the value of 2σ used in Ouchi et al. 2008) for the criteria of Lyman break off-band non-detection. In process (4), we replace 3σ with 2σ for the g- and r-magnitude criteria for consistency with the previous studies. Note that we do not apply the flags_pixel_bright_object_[center/any] masking to the LAE ALL catalog in order to maximize LAE targets for future follow-up observations (Aihara et al. 2018b). These object masking flags are used in process (4). Parameters and flags: Similar to Ono et al. (2018), we set several hscPipe parameters and flags in the HSC catalog to exclude e.g., blended sources, objects affected by saturated pixels, and nearby bright source halos. We also mask regions where exposure times are relatively short by using the countinputs parameter, Nc, which denotes the number of exposures at a source position for a given filter. Table 2 summarizes the values and provides brief explanations of the hscPipe parameters and flags used for our LAE selection. The full details of these parameters and flags are presented in Aihara et al. (2018b). To search for LAEs in large areas of the HSC fields, we do not apply the countinputs parameter to the BB images. The number of objects selected in this process is nSQL ≃ 121000. Visual inspections for object images: To exclude cosmic rays, cross-talks, compact stellar objects, and artificial diffuse objects, we perform visual inspections for the BB and NB images of all the objects selected in the process (1). Most spurious sources are diffuse components near bright stars and extended nearby galaxies. The hscPipe software conducts the cmodel fit to broad light profiles of such diffuse sources in the NB images, which enhances the BB − NB colors. For this reason, the samples constructed in the current SQL selection are contaminated by many diffuse components. Due to the clear difference of the appearance between LAE candidates and diffuse components, such spurious sources can be easily excluded through the visual inspections. The number of objects selected in this process is nvis ≃ 10900. The visual inspection processes have mainly been conducted by one of the authors. As a reliability check, four authors in this paper have individually carried out such visual inspections for ≃5300 objects in the UD-COSMOS NB816 fields, and compared the results of the LAE selection. The difference in the number of selected LAEs is within  ± 5 objects. Thus, we do not find a large difference in our visual inspection results. Rejection of variable and moving objects with multi-epoch images: We exclude variable and moving objects such as supernovae, AGNs, satellite trails, and asteroids using multi-epoch NB images. The NB images were typically taken a few months to years after the BB imaging observations. For this reason, there is a possibility that sources with an NB flux excess are variable or moving objects which happened to enhance the luminosities during the NB imaging observations. The NB images are created by coadding ≃10–20 and ≃3–5 frames of 15 min exposures for the current HSC UD and D data, respectively. Using the multi-epoch images, we automatically remove the variable and moving objects as follows. First, we measure the flux for individual epoch images, f1epoch, for each object. Next, we obtain an average, fave, and a standard deviation, σepoch, from a set of the f1epoch values after a 2σ flux clipping. Finally, we discard any object having at least a multi-epoch image with a significantly large f1epoch value of f1epoch ≥ fave + Aepoch × σepoch. Here we tune the Aepoch factor based on the depth of the NB fields. The Aepoch value is typically ≃2.0–2.5. Figure 3 shows examples of the spurious sources. We also perform visual inspections for multi-epoch images to remove contaminants which are not excluded in the automatic rejection above. We refer to the objects remaining after this process as the LAE ALL catalog. forcedselection: In the selection criteria of equations (1) and (2), the HSC LAE ALL catalog is obtained in the combination of the forced and unforced colors. In this process, we select LAEs only with the forced color excess to create the forced LAE subsamples from the HSC LAE ALL catalog. In addition, the 3σ limit is replaced with 2σ for the criteria of g- and r-band non-detections. Here we also adopt a new stringent color criterion of $$z- \mathit {NB921} >1.8$$ for z ≃ 6.6 LAEs. Due to the difference of the z-band transmission curves between SCam and HSC, the criterion of $$z- \mathit {NB921} >1.0$$ in equation (1) does not allow us to select LAEs whose EW0,Lyα is similar to those of previous SCam studies. The BB − NB color criteria in in the forced selection correspond to the rest-frame Lyα EW of EW0,Lyα > 14 Å and >10 Å for z ≃ 6.6 and z ≃ 5.7 LAEs, respectively. These EW0,Lyα limits are comparable to those of the previous SCam studies (e.g., Ouchi et al. 2010). The relation between EW0,Lyα and BB − NB colors is described in Konno et al. (2018) in detail. Moreover, we remove the objects in masked regions defined by the flags_pixel_bright_object_[center/any] parameters (Aihara et al. 2018b). We refer to the set of objects remaining after this process as the forced LAE catalog. This forced LAE catalog is used for studies of LAE statistics, such as measurements of Lyα EW scale lengths. The LAE candidates selected in this forced selection are referred to as the forced LAEs. On the other hand, we refer to the remaining LAE candidates in the HSC LAE ALL catalog as the unforced LAEs. The examples of forced and unforced LAEs are shown in figure 3. As shown in the top right-hand panels of figure 3, the unforced LAEs have a ≃0$${^{\prime\prime}_{.}}$$2–0$${^{\prime\prime}_{.}}$$3 spatial offset between centroids in NB and BB. Fig. 2. View largeDownload slide Flow chart of the HSC LAE selection process. See section 3 for more details. (Color online) Fig. 2. View largeDownload slide Flow chart of the HSC LAE selection process. See section 3 for more details. (Color online) Fig. 3. View largeDownload slide Multi-band cutout images of our example LAEs and spurious sources. (a) LAEs at z ≃ 6.6 (top) and z ≃ 5.7 (bottom) in the forced LAE catalog. (b) LAEs at z ≃ 6.6 (top) and z ≃ 5.7 (bottom) in the unforced catalog. In the rightmost cutout images, the yellow solid and cyan dashed circles represent the central positions of the unforced LAEs in the NB and BB images, respectively. The diameters of the yellow solid and dashed circles in the cutout images of the unforced LAEs are 1″ and 0$${^{\prime\prime}_{.}}$$5, respectively. (c) Spurious sources with an NB magnitude-excess similar to that of LAE candidates (four panel sets at the top): 1, variable (e.g., supernova); 2, cosmic ray; 3, cross-talk artifact; 4, moving object (e.g., asteroids) and corresponding multi-epoch images (four panel sets at the bottom). The image size is 4″ × 4″ for the LAEs and spurious sources. (Color online) Fig. 3. View largeDownload slide Multi-band cutout images of our example LAEs and spurious sources. (a) LAEs at z ≃ 6.6 (top) and z ≃ 5.7 (bottom) in the forced LAE catalog. (b) LAEs at z ≃ 6.6 (top) and z ≃ 5.7 (bottom) in the unforced catalog. In the rightmost cutout images, the yellow solid and cyan dashed circles represent the central positions of the unforced LAEs in the NB and BB images, respectively. The diameters of the yellow solid and dashed circles in the cutout images of the unforced LAEs are 1″ and 0$${^{\prime\prime}_{.}}$$5, respectively. (c) Spurious sources with an NB magnitude-excess similar to that of LAE candidates (four panel sets at the top): 1, variable (e.g., supernova); 2, cosmic ray; 3, cross-talk artifact; 4, moving object (e.g., asteroids) and corresponding multi-epoch images (four panel sets at the bottom). The image size is 4″ × 4″ for the LAEs and spurious sources. (Color online) Table 2. HSC SQL parameters and flags for our LAE selection. Parameter or flag  Value  Band  Comment  detect_is_tract_inner  True  —  Object is in an inner region of a tract and not in the overlapping region with adjacent tracts  detect_is_patch_inner  True  —  Object is in an inner region of a patch and not in the overlapping region with adjacent patches  countinputs  >=3  NB  Number of visits at a source position for a given filter  flags_pixel_edge  False  grizy, NB  Locate within images  flags_pixel_interpolated_center  False  grizy, NB  None of the central 3 × 3 pixels of an object is interpolated  flags_pixel_saturated_center  False  grizy, NB  None of the central 3 × 3 pixels of an object is saturated  flags_pixel_cr_center  False  grizy, NB  None of the central 3 × 3 pixels of an object is masked as cosmic ray  flags_pixel_bad  False  grizy, NB  None of the pixels in the footprint of an object is labelled as bad  Parameter or flag  Value  Band  Comment  detect_is_tract_inner  True  —  Object is in an inner region of a tract and not in the overlapping region with adjacent tracts  detect_is_patch_inner  True  —  Object is in an inner region of a patch and not in the overlapping region with adjacent patches  countinputs  >=3  NB  Number of visits at a source position for a given filter  flags_pixel_edge  False  grizy, NB  Locate within images  flags_pixel_interpolated_center  False  grizy, NB  None of the central 3 × 3 pixels of an object is interpolated  flags_pixel_saturated_center  False  grizy, NB  None of the central 3 × 3 pixels of an object is saturated  flags_pixel_cr_center  False  grizy, NB  None of the central 3 × 3 pixels of an object is masked as cosmic ray  flags_pixel_bad  False  grizy, NB  None of the pixels in the footprint of an object is labelled as bad  View Large In total, we identify 2230 and 1045 LAE candidates in the HSC LAE ALL and forced catalogs, respectively. Table 1 presents the numbers of LAE candidates in each field. The machine-readable catalogs of all the LAE candidates will be provided on our project website.2 The photometric properties of example LAE candidates are shown in table 3. Table 3. Photometric properties of example LAE candidates.* Object ID  NB  g  r  i  z  y    (mag)  (mag)  (mag)  (mag)  (mag)  (mag)  (1)  (2)  (3)  (4)  (5)  (6)  (7)  UD-SXDS (NB921)  HSC J021601−041442  23.85 ± 0.10  26.89 ± 0.45  27.03 ± 0.62  26.65 ± 0.63  25.28 ± 0.31  25.29 ± 0.53  HSC J021754−051454  24.01 ± 0.12  >27.6  >27.3  >26.9  26.09 ± 0.57  25.21 ± 0.50  HSC J021702−050604  24.64 ± 0.21  >27.6  >27.3  >26.9  >26.5  >25.8  HSC J021638−043228  24.74 ± 0.23  >27.6  >27.3  >26.9  26.17 ± 0.60  >25.8  HSC J021609−050236  24.90 ± 0.26  27.53 ± 0.72  27.29 ± 0.75  >26.9  26.32 ± 0.67  >25.8  UD-COSMOS (NB816)  HSC J100243+024551  23.69 ± 0.08  >27.6  >27.3  26.49 ± 0.53  >26.6  >25.8  HSC J100239+022806  24.14 ± 0.13  >27.6  >27.3  26.76 ± 0.64  26.12 ± 0.54  >25.8  HSC J100243+015931  24.63 ± 0.19  >27.6  >27.3  >27.0  >26.6  >25.8  HSC J095936+014108  25.02 ± 0.26  >27.6  >27.3  >27.0  >26.6  >25.8  HSC J100245+021536  25.15 ± 0.29  >27.6  >27.3  >27.0  >26.6  >25.8  Object ID  NB  g  r  i  z  y    (mag)  (mag)  (mag)  (mag)  (mag)  (mag)  (1)  (2)  (3)  (4)  (5)  (6)  (7)  UD-SXDS (NB921)  HSC J021601−041442  23.85 ± 0.10  26.89 ± 0.45  27.03 ± 0.62  26.65 ± 0.63  25.28 ± 0.31  25.29 ± 0.53  HSC J021754−051454  24.01 ± 0.12  >27.6  >27.3  >26.9  26.09 ± 0.57  25.21 ± 0.50  HSC J021702−050604  24.64 ± 0.21  >27.6  >27.3  >26.9  >26.5  >25.8  HSC J021638−043228  24.74 ± 0.23  >27.6  >27.3  >26.9  26.17 ± 0.60  >25.8  HSC J021609−050236  24.90 ± 0.26  27.53 ± 0.72  27.29 ± 0.75  >26.9  26.32 ± 0.67  >25.8  UD-COSMOS (NB816)  HSC J100243+024551  23.69 ± 0.08  >27.6  >27.3  26.49 ± 0.53  >26.6  >25.8  HSC J100239+022806  24.14 ± 0.13  >27.6  >27.3  26.76 ± 0.64  26.12 ± 0.54  >25.8  HSC J100243+015931  24.63 ± 0.19  >27.6  >27.3  >27.0  >26.6  >25.8  HSC J095936+014108  25.02 ± 0.26  >27.6  >27.3  >27.0  >26.6  >25.8  HSC J100245+021536  25.15 ± 0.29  >27.6  >27.3  >27.0  >26.6  >25.8  *(1) Object ID. (2)–(7) Total magnitude of NB, g, r, i, z, and y bands. The 2σ limits of the total magnitudes for the undetected bands. (The complete machine-readable catalogs will be available on our project web page.)2 View Large As shown in table 1, the number of z ≃ 5.7 LAEs in D-DEEP2-3 appears to be large compared to that of the other z ≃ 5.7 fields. This may be because the seeing of the NB816 images of D-DEEP2-3 is better than that of the other z ≃ 5.7 fields. Similarly, the small number of z ≃ 6.6 LAEs in UD-SXDS may be affected by the seeing size. The number density of LAEs is discussed in the next section. Note that edge regions of UD-COSMOS is overlapped with a flanking field, D-COSMOS (Aihara et al. 2018a). We find that 30(7) LAEs in UD-COSMOS are also selected in the HSC LAE ALL(forced) sample of D-COSMOS. To analyze the D field independently in the following sections, we include the overlapped LAEs in the D-COSMOS sample. Figure 4 shows the color–magnitude diagrams for the LAE candidates. The solid curves in the color–magnitude diagrams indicate the 3σ errors of BB − NB color as a function of the NB flux, fNB, given by   \begin{equation} \pm 3 \sigma _{\rm BB-NB} = -2.5 \log _{10} \left( 1\, \mp \, 3\frac{\sqrt{f^2_{\rm 1\sigma NB} + f^2_{\rm 1\sigma BB}}}{f_{\rm NB}} \right), \end{equation} (3)where f1σNB and f1σBB are the 1σ flux error in the z and NB921 (i and NB816) bands for z ≃ 6.6 (z ≃ 5.7), respectively. As shown in figure 4, the LAE candidates have a significant NB magnitude excess. Fig. 4. View largeDownload slide (Top) Color of z − NB921 as a function of NB921 magnitude for the LAEs at z ≃ 6.6 in the UD (left) and D (right) fields. The filled red and open magenta circles denote the forced and unforced LAEs, respectively. For the LAEs undetected in the z-band images, the z-band magnitudes are replaced with the 2σ limiting magnitudes. The x-axis denotes the forced (unforced) z − NB921 colors for the forced (unforced) LAEs. The horizontal dashed and dotted lines shows the color criteria of z − NB921 >1.0 and z − NB921 >1.8, respectively. The gray dots represent objects detected in the NB921 images. The solid curves show the 3σ error tracks of z − NB921 color for each field. The 3σ error tracks are derived by equation (3). (Bottom) Color of i − NB816 as a function of NB816 magnitude for the LAEs at z ≃ 5.7. The definitions of symbols, curves, and lines are the same as those of the top panels. (Color online) Fig. 4. View largeDownload slide (Top) Color of z − NB921 as a function of NB921 magnitude for the LAEs at z ≃ 6.6 in the UD (left) and D (right) fields. The filled red and open magenta circles denote the forced and unforced LAEs, respectively. For the LAEs undetected in the z-band images, the z-band magnitudes are replaced with the 2σ limiting magnitudes. The x-axis denotes the forced (unforced) z − NB921 colors for the forced (unforced) LAEs. The horizontal dashed and dotted lines shows the color criteria of z − NB921 >1.0 and z − NB921 >1.8, respectively. The gray dots represent objects detected in the NB921 images. The solid curves show the 3σ error tracks of z − NB921 color for each field. The 3σ error tracks are derived by equation (3). (Bottom) Color of i − NB816 as a function of NB816 magnitude for the LAEs at z ≃ 5.7. The definitions of symbols, curves, and lines are the same as those of the top panels. (Color online) 4 Checking the reliability of our LAE selection Here we check the reliability of our LAE selection. 4.1 Spectroscopic confirmations We have conducted optical spectroscopic observations with Subaru/FOCAS and Magellan/LDSS3 for 18 bright LAE candidates with NB ≲ 24 mag. From these observations, we have confirmed 13 LAEs. By investigating our spectroscopic catalog of Magellan/IMACS, we also spectroscopically identify eight LAEs with $$\mathit {NB}\lesssim 24\:$$mag. In addition, we find that there are 75 LAEs spectroscopically confirmed in literature (Murayama et al. 2007; Ouchi et al. 2008, 2010; Taniguchi et al. 2009; Mallery et al. 2012; Sobral et al. 2015; R. Higuchi et al. in preparation). In total, 96 LAEs have been confirmed in our spectroscopy and previous studies. Using the spectroscopic sample that has a known number of observed LAEs, we estimate the contamination rate to be ≃0%–30%. The details of the spectroscopic observations and contamination rates are given by Shibuya et al. (2018). 4.2 LAE surface number density Figure 5 shows the surface number density (SND) of our LAE candidates and LAEs identified in previous Subaru/SCam NB surveys, SCam LAEs (e.g., Ouchi et al. 2008, 2010). We find that the SNDs of the forced LAEs are comparable to those of SCam LAEs. On the other hand, the SNDs of unforced LAEs at z ≃ 6.6 are higher than that of SCam LAEs. The high SND of the unforced LAEs is mainly caused by the color criterion for the HSC LAE ALL catalog of $$z- \mathit {NB921} >1.0$$ that is less stringent than $$z- \mathit {NB921} >1.8$$ (see section 3). We also identify SND humps of our forced LAEs at z ≃ 6.6 at the bright-end of NB ≃ 23 mag in UD-COSMOS. The presence of such an SND hump has been reported by z ≃ 6.6 LAE studies (e.g., Matthee et al. 2015). The significance of the bright-end hump’s existence in Lyα LFs is ≃3σ, which are discussed in Konno et al. (2018). The slight decline in SNDs at a faint NB magnitude of NB ≳ 24.5 mag would originate from the incompleteness of the LAE detection and selection. Konno et al. (2018) present the SND corrected for the incompleteness. Fig. 5. View largeDownload slide Surface number density (SND) of the HSC LAEs at z ≃ 6.6 (five panels on the left) and ≃5.7 (four panels on the right) in each UD and D field. The filled red and open magenta circles indicate the LAEs in the forced and ALL catalog, respectively. The error bars are given by Poisson statistics from the number of LAEs. The gray crosses represent the LAEs in Ouchi et al. (2010) for z ≃ 6.6 and Ouchi et al. (2008) for z ≃ 5.7. The data points of the gray crosses are identical in all the fields for each redshift. The SND slight declines in the HSC LAEs at NB ≳ 24.5 mag would be originated from the incompleteness of the LAE detection and selection. The completeness-corrected SNDs are presented in Konno et al. (2018). The data points of the HSC LAEs are slightly shifted along the x-axis for clarity. (Color online) Fig. 5. View largeDownload slide Surface number density (SND) of the HSC LAEs at z ≃ 6.6 (five panels on the left) and ≃5.7 (four panels on the right) in each UD and D field. The filled red and open magenta circles indicate the LAEs in the forced and ALL catalog, respectively. The error bars are given by Poisson statistics from the number of LAEs. The gray crosses represent the LAEs in Ouchi et al. (2010) for z ≃ 6.6 and Ouchi et al. (2008) for z ≃ 5.7. The data points of the gray crosses are identical in all the fields for each redshift. The SND slight declines in the HSC LAEs at NB ≳ 24.5 mag would be originated from the incompleteness of the LAE detection and selection. The completeness-corrected SNDs are presented in Konno et al. (2018). The data points of the HSC LAEs are slightly shifted along the x-axis for clarity. (Color online) Figure 6 compiles the SNDs of all the HSC UD and D fields. We find that our SNDs show a small field-to-field variation, but typically follow those of the SCam LAEs. Fig. 6. View largeDownload slide SND as a function of NB magnitude for the LAEs at z ∼ 6.6 (left) and ∼5.7 (right) in the HSC LAE forced catalog. The colored symbols denote the LAEs in each UD and D field (green circles: UD-SXDS; magenta squares: UD-COSMOS; cyan triangles: D-ELAIS-N1; pink inverse-triangles: D-DEEP2-3; orange diamonds: D-COSMOS). The error bars are given by Poisson statistics from the number of LAEs. The gray crosses represent the LAEs in Ouchi et al. (2010) for z ≃ 6.6 and Ouchi et al. (2008) for z ≃ 5.7. The slight declinein SND in the HSC LAEs at NB ≳ 24.5 mag originates from the incompleteness of the LAE selection. The completeness-corrected SNDs are presented in Konno et al. (2018). The data points of the HSC LAEs are slightly shifted along the x-axis for clarity. (Color online) Fig. 6. View largeDownload slide SND as a function of NB magnitude for the LAEs at z ∼ 6.6 (left) and ∼5.7 (right) in the HSC LAE forced catalog. The colored symbols denote the LAEs in each UD and D field (green circles: UD-SXDS; magenta squares: UD-COSMOS; cyan triangles: D-ELAIS-N1; pink inverse-triangles: D-DEEP2-3; orange diamonds: D-COSMOS). The error bars are given by Poisson statistics from the number of LAEs. The gray crosses represent the LAEs in Ouchi et al. (2010) for z ≃ 6.6 and Ouchi et al. (2008) for z ≃ 5.7. The slight declinein SND in the HSC LAEs at NB ≳ 24.5 mag originates from the incompleteness of the LAE selection. The completeness-corrected SNDs are presented in Konno et al. (2018). The data points of the HSC LAEs are slightly shifted along the x-axis for clarity. (Color online) 4.3 Matching rate of HSC LAEs and SCam LAEs The UD-SXDS field has been observed previously by SCam equipped with the NB921 and NB816 filters (Ouchi et al. 2008, 2010). We compare the catalogs of our selected HSC LAE candidates and SCam LAEs, and calculate the object-matching rates as a function of NB magnitudes. The object-matching radius is 1″. The object-matching rate between the HSC LAEs and SCam LAEs is ≃90% at bright NB magnitudes of ≲ 24 mag. The high object-matching rate indicates that we adequately identify LAEs in our selection processes. However, the matching rate decreases to ≃70% at the faint magnitude of ≃24.5 mag. This is due to the shallow depth of the HSC NB fields compared to the SCam ones. Konno et al. (2018) discuss the detection completeness of faint LAEs. 5 Results Here we present the Lyα EW distributions (subsection 5.1) and LABs selected with the HSC data (subsection 5.2). For consistency with previous LAE studies, we use the forced LAE sample in the following analyses, unless otherwise specified. 5.1 Lyα EW distribution We present the Lyα EW distributions for LAEs at z ≃ 5.7–6.6. Using the method described in the Appendix, we calculate the rest-frame Lyα EW, EW0,Lyα, for the LAEs. The y(z)-band magnitudes are used for the rest-frame UV continuum emission of z ≃ 6.6(z ≃ 5.7) LAEs. Figure 7 shows the observed Lyα EW distributions at z ≃ 5.7–6.6 in the UD and D fields. To quantify these Lyα EW distributions we perform Monte Carlo (MC) simulations. The procedure of the MC simulations is similar to that of e.g., Shimasaku et al. (2006), Ouchi et al. (2008), and Zheng et al. (2014). First, we generate artificial LAEs in a Lyα luminosity range of log LLyα/erg s−1 = 42–44 according to z ≃ 5.7–6.6 Lyα LFs of Konno et al. (2018). Next, we assign Lyα EW and BB magnitudes to each LAE by assuming that the Lyα EW distributions are the exponential and Gaussian functions (e.g., Gronwall et al. 2007; Kashikawa et al. 2011; Oyarzún et al. 2016):   \begin{equation} \frac{{\rm d}N}{{\rm d} EW} = N \exp \left(-\frac{EW}{W_{\rm e}}\right), \end{equation} (4)and   \begin{equation} \frac{{\rm d}N}{{\rm d} EW} = N \frac{1}{\sqrt{2\pi \sigma _{\rm g}^2}} \exp \left(-\frac{EW^2}{2\sigma _{\rm g}^2} \right), \end{equation} (5)where N is the galaxy number, and We and σg are the Lyα EW scale lengths of the exponential and Gaussian functions, respectively. By changing the intrinsic We and σg values, we make samples of artificial Lyα EW distributions. We then select LAEs based on NB and BB limiting magnitudes and BB − NB colors corresponding to Lyα EW limits which are the same as those of our LAE selection criteria (section 3). Finally, the best-fitting Lyα EW scale lengths are obtained by fitting the artificial Lyα EW distribution to the observed ones. Fig. 7. View largeDownload slide Lyα EW distribution for the HSC LAEs at z ≃ 6.6 (left) and z ≃ 5.7 (right). The top and bottom panels show the UD and D fields, respectively. The thin gray histograms with error bars denote the Lyα EW distributions for the forced LAEs. The error bars are given by Poisson statistics from the number of sample LAEs. The red solid and blue dashed lines present the best-fitting exponential and Gaussian functions of equations (4) and (5), respectively, which are obtained from MC simulations with the EW0,Lyα uncertainties (see subsection 5.1 for more details). (Color online) Fig. 7. View largeDownload slide Lyα EW distribution for the HSC LAEs at z ≃ 6.6 (left) and z ≃ 5.7 (right). The top and bottom panels show the UD and D fields, respectively. The thin gray histograms with error bars denote the Lyα EW distributions for the forced LAEs. The error bars are given by Poisson statistics from the number of sample LAEs. The red solid and blue dashed lines present the best-fitting exponential and Gaussian functions of equations (4) and (5), respectively, which are obtained from MC simulations with the EW0,Lyα uncertainties (see subsection 5.1 for more details). (Color online) Figure 7 presents the Lyα EW distributions obtained in the MC simulations. As shown in figure 7, we find that the Lyα EW distributions are reasonably explained by the exponential and Gaussian profiles. The best-fitting scale lengths are summarized in table 4. The best-fitting exponential (Gaussian) Lyα scale lengths are, on average, 153 ± 18 Å and 154 ± 15 Å (146 ± 24 Å and 139 ± 14 Å) at z ≃ 5.7 and z ≃ 6.6 for the UD and D fields, respectively. As shown in table 4, there is no large difference in the Lyα EW scale lengths for the UD and D fields. This lack of large EW0,Lyα difference indicates that the results of our best-fitting Lyα EW scale lengths do not highly depend on the image depths or the detection incompleteness. In subsection 6.1, we discuss the redshift evolution of the Lyα EW scale lengths. We investigate LEW LAEs that have intrinsic Lyα EW values, $$EW_{\rm 0,\, Ly\alpha }^{\rm int}$$, that exceed 240 Å (e.g., Malhotra & Rhoads 2002; Dawson et al. 2004). To obtain $$EW_{\rm 0,\, Ly\alpha }^{\rm int}$$, we correct for the IGM attenuation for Lyα using the prescriptions of Madau (1995). In the HSC LAE ALL sample, we find that 45 and 230 LAEs have a LEW of $$EW_{\rm 0,\, Ly\alpha }^{\rm int} > 240\,$$Å, for z ≃ 6.6 and z ≃ 5.7 LAEs, respectively. These LEW LAEs are candidates of young metal-poor galaxies and AGNs. The fraction of the LEW LAEs in the sample is 21% for z ≃ 5.7 LAEs. The fraction of LEW LAEs at z ≃ 5.7 is comparable to that of previous studies on z ≃ 5.7 LAEs (e.g., ≃25% at z ≃ 5.7 in Ouchi et al. 2008; ≃30%–40% at z ≃ 5.7 in Shimasaku et al. 2006). In contrast, the fraction of LEW LAEs at z ≃ 6.6 is 4%, which is lower than that at z ≃ 5.7. This low fraction at z ≃ 6.6 might be due to the neutral hydrogen IGM absorbing the Lyα emission. Of the LEW LAEs, 32 and 150 LAEs at z ≃ 6.6 and z ≃ 5.7 exceed $$EW_{\rm 0,\, Ly\alpha }^{\rm int} = 240$$ beyond the 1σ uncertainty of $$EW_{\rm 0,\, Ly\alpha }^{\rm int}$$, respectively. 5.2 LABs at z ≃ 5.7–6.6 We search for LABs with spatially extended Lyα emission. To identify LABs, we measure the NB isophotal areas, Aiso, of the forced LAEs. In this process, we include an unforced LAE, Himiko, which is an LAB that was identified in a previous SCam NB survey (Ouchi et al. 2009). First, we estimate the sky background level of the NB cutout images. Next, we run the SExtractor with the sky background level, and obtain the Aiso values as pixels with fluxes brighter than the 2σ sky fluctuation. Note that the NB magnitudes include both fluxes of Lyα and the rest-frame UV continuum emission. Instead of creating Lyα images by subtracting the flux contribution of the rest-frame UV continuum emission, we here simply use the NB images for consistency with previous studies (e.g., Ouchi et al. 2009). Using Aiso and NB magnitude diagrams, we select LABs which are significantly extended compared to point sources. This selection is similar to that of Yang et al. (2010). Figure 8 presents Aiso as a function of total NB magnitude. We also plot star-like point sources which are randomly selected in HSC NB fields. The Aiso and NB magnitude selection window is defined by a 2.5σ deviation from the Aiso–NB magnitude distribution for the star-like point sources. The value of 2.5σ is applied for fair comparison with previous studies of e.g., Yang et al. (2009, 2010) who have used ≃2–4σ. We perform visual inspections for the NB cutout images to remove unreliable LABs which are significantly affected by, e.g., diffuse halos of nearby bright stars. Fig. 8. View largeDownload slide Isophotal area, Aiso, as a function of NB magnitude to select LABs at z ≃ 6.6 (left) and z ≃ 5.7 (right). The top and bottom panels show the UD and D fields, respectively. The green diamonds denote the LABs. The filled red circles indicate the LAEs in the forced catalog. The gray dots represent star-like point sources selected in the HSC NB images. The diagonal and vertical lines denote the LAB selection criteria of Aiso and NB magnitude. The diagonal lines are defined by the 2.5σ deviation from the Aiso–NB magnitude distribution for the star-like point sources. The filled red circles with crosses indicate unreliable LAB candidates which are affected by e.g., diffuse halos of nearby bright stars. The z ≃ 6.6 LABs in the UD fields are CR7 (Sobral et al. 2015) and Himiko (Ouchi et al. 2009). (Color online) Fig. 8. View largeDownload slide Isophotal area, Aiso, as a function of NB magnitude to select LABs at z ≃ 6.6 (left) and z ≃ 5.7 (right). The top and bottom panels show the UD and D fields, respectively. The green diamonds denote the LABs. The filled red circles indicate the LAEs in the forced catalog. The gray dots represent star-like point sources selected in the HSC NB images. The diagonal and vertical lines denote the LAB selection criteria of Aiso and NB magnitude. The diagonal lines are defined by the 2.5σ deviation from the Aiso–NB magnitude distribution for the star-like point sources. The filled red circles with crosses indicate unreliable LAB candidates which are affected by e.g., diffuse halos of nearby bright stars. The z ≃ 6.6 LABs in the UD fields are CR7 (Sobral et al. 2015) and Himiko (Ouchi et al. 2009). (Color online) In total, we identify 11 LABs at z ≃ 5.7–6.6. Figure 9 and table 5 present multi-band cutout images and properties for the LABs, respectively. As shown in figure 9, these LABs are spatially extended in NB. Our HSC LAB selection confirms that CR7 and Himiko have a spatially extended Lyα emission. Six out of our 11 LABs have been confirmed by our spectroscopic follow-up observations (Shibuya et al. 2018) and previous studies (Ouchi et al. 2009; Mallery et al. 2012; Sobral et al. 2015). In subsection 6.2, we discuss the redshift evolution of the LAB number density. Fig. 9. View largeDownload slide Postage stamps of the LABs selected with the HSC NB data. The yellow contours indicate isophotal apertures with a threshold of 2σ sky background noise level. The size of the cutout images is 4″ × 4″. (Color online) Fig. 9. View largeDownload slide Postage stamps of the LABs selected with the HSC NB data. The yellow contours indicate isophotal apertures with a threshold of 2σ sky background noise level. The size of the cutout images is 4″ × 4″. (Color online) Table 5. Properties of the LABs selected in the HSC NB data.* Object ID  α  δ  $$\mathit {NB}_{\rm tot}$$  UVtot  log LLyα  $$\mathit {EW}_{\rm 0,\, Ly\alpha }$$  zspec    (J2000.0)  (J2000.0)  (mag)  (mag)  (erg s−1)  (Å)    (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  NB921 (z ≃ 6.6)  HSC J100058+014815†  10h00m58$${^{\rm s}_{.}}$$00  +01°48΄15$${^{\prime\prime}_{.}}$$14  23.25  24.48  43.9‖  211 ± 20‖  6.604†  HSC J021757−050844‡  02h17m57$${^{\rm s}_{.}}$$58  −05°08΄44$${^{\prime\prime}_{.}}$$64  23.50  25.40  43.4‖  $$78^{+8}_{-6}{^{\Vert }}$$  6.595‡  HSC J100334+024546§  10h03m34$${^{\rm s}_{.}}$$66  +02°45΄46$${^{\prime\prime}_{.}}$$56  23.61  24.97  43.5‖  61 ± 20‖  6.575§  NB816 (z ≃ 5.7)  HSC J100129+014929  10h01m29$${^{\rm s}_{.}}$$07  +01°49΄29$${^{\prime\prime}_{.}}$$81  23.47  25.87  43.4  $$95^{+40}_{-19}$$  5.707♯  HSC J100109+021513  10h01m09$${^{\rm s}_{.}}$$72  +02°15΄13$${^{\prime\prime}_{.}}$$45  23.13  25.77  43.6  $$257^{+172}_{-76}$$  5.712♯  HSC J100123+015600  10h01m23$${^{\rm s}_{.}}$$84  +01°56΄00$${^{\prime\prime}_{.}}$$46  23.94  26.43  43.3  $$106^{+70}_{-27}$$  5.726♯  HSC J095946+013208  09h59m46$${^{\rm s}_{.}}$$73  +01°32΄08$${^{\prime\prime}_{.}}$$45  24.16  26.12  43.1  $$52^{+25}_{-13}$$  —  HSC J100139+015428  10h01m39$${^{\rm s}_{.}}$$94  +01°54΄28$${^{\prime\prime}_{.}}$$34  24.11  26.58  43.2  $$100^{+66}_{-30}$$  —  HSC J161927+551144  16h19m27$${^{\rm s}_{.}}$$73  +55°11΄44$${^{\prime\prime}_{.}}$$70  22.88  24.86  43.7  $$89^{+33}_{-20}$$  —  HSC J161403+535701  16h14m03$${^{\rm s}_{.}}$$82  +53°57΄01$${^{\prime\prime}_{.}}$$25  23.53  25.32  43.4  $$51^{+23}_{-12}$$  —  HSC J232924+003600  23h29m24$${^{\rm s}_{.}}$$85  +00°36΄00$${^{\prime\prime}_{.}}$$34  23.62  26.48  43.4  $$55^{+45}_{-14}$$  —  Object ID  α  δ  $$\mathit {NB}_{\rm tot}$$  UVtot  log LLyα  $$\mathit {EW}_{\rm 0,\, Ly\alpha }$$  zspec    (J2000.0)  (J2000.0)  (mag)  (mag)  (erg s−1)  (Å)    (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  NB921 (z ≃ 6.6)  HSC J100058+014815†  10h00m58$${^{\rm s}_{.}}$$00  +01°48΄15$${^{\prime\prime}_{.}}$$14  23.25  24.48  43.9‖  211 ± 20‖  6.604†  HSC J021757−050844‡  02h17m57$${^{\rm s}_{.}}$$58  −05°08΄44$${^{\prime\prime}_{.}}$$64  23.50  25.40  43.4‖  $$78^{+8}_{-6}{^{\Vert }}$$  6.595‡  HSC J100334+024546§  10h03m34$${^{\rm s}_{.}}$$66  +02°45΄46$${^{\prime\prime}_{.}}$$56  23.61  24.97  43.5‖  61 ± 20‖  6.575§  NB816 (z ≃ 5.7)  HSC J100129+014929  10h01m29$${^{\rm s}_{.}}$$07  +01°49΄29$${^{\prime\prime}_{.}}$$81  23.47  25.87  43.4  $$95^{+40}_{-19}$$  5.707♯  HSC J100109+021513  10h01m09$${^{\rm s}_{.}}$$72  +02°15΄13$${^{\prime\prime}_{.}}$$45  23.13  25.77  43.6  $$257^{+172}_{-76}$$  5.712♯  HSC J100123+015600  10h01m23$${^{\rm s}_{.}}$$84  +01°56΄00$${^{\prime\prime}_{.}}$$46  23.94  26.43  43.3  $$106^{+70}_{-27}$$  5.726♯  HSC J095946+013208  09h59m46$${^{\rm s}_{.}}$$73  +01°32΄08$${^{\prime\prime}_{.}}$$45  24.16  26.12  43.1  $$52^{+25}_{-13}$$  —  HSC J100139+015428  10h01m39$${^{\rm s}_{.}}$$94  +01°54΄28$${^{\prime\prime}_{.}}$$34  24.11  26.58  43.2  $$100^{+66}_{-30}$$  —  HSC J161927+551144  16h19m27$${^{\rm s}_{.}}$$73  +55°11΄44$${^{\prime\prime}_{.}}$$70  22.88  24.86  43.7  $$89^{+33}_{-20}$$  —  HSC J161403+535701  16h14m03$${^{\rm s}_{.}}$$82  +53°57΄01$${^{\prime\prime}_{.}}$$25  23.53  25.32  43.4  $$51^{+23}_{-12}$$  —  HSC J232924+003600  23h29m24$${^{\rm s}_{.}}$$85  +00°36΄00$${^{\prime\prime}_{.}}$$34  23.62  26.48  43.4  $$55^{+45}_{-14}$$  —  *(1) Object ID. (2) Right ascension. (3) Declination. (4) Total magnitudes of NB921 for z ≃ 6.6 and NB816 bands for z ≃ 5.7. (5) Total magnitudes of y band for z ≃ 6.6 and z bands for z ≃ 5.7. (6) Lyα luminosity. (7) Rest-frame equivalent width of Lyα emission line. (8) Spectroscopic redshift. †CR7 in Sobral et al. (2015). ‡Himiko in Ouchi et al. (2009). §Spectroscopically confirmed in Shibuya et al. (2018). ‖Spectroscopic measurements from the literature. ♯Spectroscopically confirmed in Mallery et al. (2012). View Large Table 4. Best-fitting Lyα EW scale lengths.* Redshift    We    σg      (Å)    (Å)  (1)    (2)    (3)  6.6 (UD)    $$168^{+4}_{-4}$$    $$124^{+8}_{-8}$$  5.7 (UD)    $$138^{+6}_{-4}$$    $$164^{+2}_{-2}$$  6.6 (D)    $$140^{+14}_{-16}$$    $$154^{+4}_{-24}$$  5.7 (D)    $$168^{+18}_{-18}$$    $$128^{+36}_{-12}$$  Redshift    We    σg      (Å)    (Å)  (1)    (2)    (3)  6.6 (UD)    $$168^{+4}_{-4}$$    $$124^{+8}_{-8}$$  5.7 (UD)    $$138^{+6}_{-4}$$    $$164^{+2}_{-2}$$  6.6 (D)    $$140^{+14}_{-16}$$    $$154^{+4}_{-24}$$  5.7 (D)    $$168^{+18}_{-18}$$    $$128^{+36}_{-12}$$  *(1) Redshift of the LAE sample. The parenthesis indicates the UD or D fields. (2) Best-fitting Lyα EW scale length of the exponential form. (3) Best-fitting Lyα EW scale length of the Gaussian form. View Large 6 Discussion 6.1 Redshift evolution of Lyα EW distribution We discuss the redshift evolution of the Lyα EW scale lengths through a compilation of the results from literature (Zheng et al. 2014; Ouchi et al. 2008; Nilsson et al. 2009; Hu et al. 2010; Kashikawa et al. 2011; Cowie et al. 2011; Ciardullo et al. 2012). Figure 10 shows the redshift evolution of the Lyα EW scale lengths at z ≃ 0–7. Our best-fitting Lyα scale lengths are comparable to those of Kashikawa et al. (2011) and/or Zheng et al. (2014) at z ≃ 5.7–6.6. The high Lyα EW scale lengths at high z would indicate that metal-poor and/or less dusty galaxies with a strong Lyα emission are more abundant at higher z (e.g., Stark et al. 2011). In addition, Zheng et al. (2014) have found that the Lyα EW scale length increases towards high z following a (1 + z)-form. Our We and σg values for z ≃ 5.7–6.6 are also roughly comparable to Zheng et al.’s (1 + z)-form evolution. However, no significant evolution in the Lyα EW scale lengths from z ≃ 5.7 to z ≃ 6.6 is identified in our HSC LAE data, although a possible decline in σg in the UD fields is found. A slight decrease both in We and σg from z ≃ 5.7 to z ≃ 6.6 has been found by Kashikawa et al. (2011). This sudden decline in the Lyα scale lengths at z ≃ 6.6 may be caused by the increasing hydrogen neutral fraction in the epoch of the cosmic reionization at z ≳ 7. Note that the Lyα EW scale length measurements would largely depend on BB and NB depths and Lyα EW cuts. Using deeper NB and BB images from the future HSC data release, we will examine the redshift evolution of Lyα scale lengths accurately. Fig. 10. View largeDownload slide Redshift evolution of the best-fitting Lyα EW scale lengths of the exponential (top) and Gaussian (bottom) functions. The red squares and circles indicate our HSC LAEs in the UD and D fields, respectively. The black symbols are taken from the data points in literature, as compiled in Zheng et al. (2014) (crosses: Cowie et al. 2011; asterisks: Ciardullo et al. 2012; filled triangle: Nilsson et al. 2009; filled inverse triangles: Ouchi et al. 2008; filled diamonds: Kashikawa et al. 2011; open circles: results of Monte Carlo simulations using data of Zheng et al. 2014 and Hu et al. 2010). The gray curves indicate the best-fitting (1 + z)-form functions obtained in Zheng et al. (2014). (Color online) Fig. 10. View largeDownload slide Redshift evolution of the best-fitting Lyα EW scale lengths of the exponential (top) and Gaussian (bottom) functions. The red squares and circles indicate our HSC LAEs in the UD and D fields, respectively. The black symbols are taken from the data points in literature, as compiled in Zheng et al. (2014) (crosses: Cowie et al. 2011; asterisks: Ciardullo et al. 2012; filled triangle: Nilsson et al. 2009; filled inverse triangles: Ouchi et al. 2008; filled diamonds: Kashikawa et al. 2011; open circles: results of Monte Carlo simulations using data of Zheng et al. 2014 and Hu et al. 2010). The gray curves indicate the best-fitting (1 + z)-form functions obtained in Zheng et al. (2014). (Color online) 6.2 Redshift evolution of LAB number density We discuss the redshift evolution of the LAB number density, NLAB. Figure 11 shows NLAB at z ≃ 0–7 measured by this study and from the literature (Saito et al. 2006; Keel et al. 2009; Matsuda et al. 2009; Yang et al. 2009, 2010). For the plot of the NLAB, Yang et al. (2010) have compiled NLAB measurements down to an NB surface brightness (SB) limit of 5 × 10−18 erg s−1 cm−2 arcsec−2. The SB limits of our HSC NB data are ≃5 × 10−18 and ≃8 × 10−18 erg s−1 cm−2 for the UD and D fields, respectively. Our HSC NB images at least for the UD fields are comparably deep, allowing for fair comparisons with Yang et al.’s NLAB plot. Our NLAB values are 1.4 × 10−6 and 2.9 × 10−7 Mpc−3 (2.6 × 10−7 and 1.1 × 10−7 Mpc−3) at z ≃ 5.7 and z ≃ 6.6 in the UD (D) fields, respectively. The number density at z ≃ 6–7 is ≃10–100 times lower than those claimed for LABs at z ≃ 2–3 (e.g., Matsuda et al. 2004; Yang et al. 2009, 2010). As shown in figure 11, there is an evolutional trend that NLAB increases from z ≃ 7 to ≃3 and subsequently decreases from z ≃ 3 to ≃0. This trend of the LAB number density evolution is similar to the Madau–Lilly plot of the cosmic star formation rate density (SFRD) evolution (e.g., Madau et al. 1996; Lilly et al. 1996). Similar to Shibuya et al. (2016), we fit the Madau–Lilly plot-type formula,   \begin{equation} N_{\rm LAB}(z) = a \frac{(1+z)^b}{1 + [(1+z)/c]^d}, \end{equation} (6)where a, b, c, and d are free parameters (Madau & Dickinson 2014), to our NLAB evolution. For the fitting, we exclude Matsuda et al.’s (2009) data point, which was obtained in an overdense region, SSA22. The best-fitting parameters are a = 9.1 × 10−8, b = 2.9, c = 5.0, and d = 11.7. Fig. 11. View largeDownload slide Redshift evolution of the LAB number density. The filled red squares and filled red circles denote the LABs selected in the HSC UD and D fields, respectively. The error bars are given by Poisson statistics from the LAB number counts. The black symbols show LABs found in the literature (filled diamond: Keel et al. 2009; filled circle: Yang et al. 2009; open circle: Yang et al. 2010; filled inverse-triangle: Matsuda et al. 2004; pentagon: Saito et al. 2006). All the measurements are based on LABs identified down to the surface brightness limit of ≃5 × 10−18 erg s−1 cm−2 arcsec−2. The gray solid curve represents the best-fitting formula of equation (6) to the data points expect for the measurement in the SSA22 proto-cluster region. (Color online) Fig. 11. View largeDownload slide Redshift evolution of the LAB number density. The filled red squares and filled red circles denote the LABs selected in the HSC UD and D fields, respectively. The error bars are given by Poisson statistics from the LAB number counts. The black symbols show LABs found in the literature (filled diamond: Keel et al. 2009; filled circle: Yang et al. 2009; open circle: Yang et al. 2010; filled inverse-triangle: Matsuda et al. 2004; pentagon: Saito et al. 2006). All the measurements are based on LABs identified down to the surface brightness limit of ≃5 × 10−18 erg s−1 cm−2 arcsec−2. The gray solid curve represents the best-fitting formula of equation (6) to the data points expect for the measurement in the SSA22 proto-cluster region. (Color online) The similarity of the cosmic SFRD and LAB evolution might indicate that the origin of LABs are related to the star formation activity. As described in section 1, LABs are thought to be formed via physical mechanisms that are connected with star formation, e.g., cold gas accretion and galactic superwinds. The cold gas accretion could produce the extended Lyα emission powered by the gravitational energy (e.g., Mas-Ribas & Dijkstra 2016; Momose et al. 2016; Mas-Ribas et al. 2017). On the other hand, the superwinds induced by the starbursts in the central galaxies would blow out the surrounding neutral gas, and form extended Lyα nebulae (e.g., Mori & Umemura 2006). The cold gas accretion rate and the strength of galactic superwinds are predicted to evolve with physical quantities related to the cosmic SFRD (e.g., Dekel et al. 2009; Kereš et al. 2009). The comparisons of the cosmic SFRD and LAB evolutions would provide useful hints that LABs are formed in these scenarios. However, it should be noted that the LAB selection method is not homogeneous in our comparison of NLAB at z ≃ 0–7. There is a possibility that the NLAB evolution from z ≃ 7 to z ≃ 3 is caused by the cosmological surface brightness dimming effect at high z. The cosmological surface brightness dimming would significantly affect the detection and selection completeness for LABs at high z. To confirm the NLAB evolution and quantitatively compare it with the cosmic SFRD, we need to homogenize the selection method for LABs at z ≃ 2–7 in any future HSC NB data. 7 Summary and conclusions We develop an unprecedentedly large catalog consisting of LAEs at z = 5.7 and 6.6 that are identified by the SILVERRUSH program using the first NB imaging data of the Subaru/HSC survey. The NB imaging data is about an order of magnitude larger than any other surveys for z ≃ 6–7 LAEs conducted to date. Our findings are as follows: We identify 2230 ≳ L* LAEs at z = 5.7 and 6.6 on the 13.8 and 21.2 deg2 sky, respectively. We confirm that the LAE catalog is reliable on the basis of 96 LAEs whose spectroscopic redshifts are already determined by this program (Shibuya et al. 2018) and previous studies (e.g., Mallery et al. 2012). The LAE catalog is introduced in this work, and published online. Using the large LAE catalog we derive the rest-frame Lyα EW distributions of LAEs at z ≃ 5.7 and ≃6.6 that are reasonably explained by the exponential profile. The best-fitting exponential (Gaussian) Lyα scale lengths are, on average, 153 ± 18 Å and 154 ± 15 Å (146 ± 24 Å and 139 ± 14 Å) at z ≃ 5.7 and z ≃ 6.6, for the Ultradeep and Deep fields, respectively, showing no significant evolution from z ≃ 5.7 to z ≃ 6.6. We find 45 and 230 LAEs at z ≃ 6.6 and z ≃ 5.7 with an LEW of $$EW_{\rm 0,\, Ly\alpha }^{\rm int}> 240\,$$Å, corrected for the IGM attenuation for Lyα. The fraction of LEW LAEs to all LAEs is ≃4% at z ≃ 6.6 and ≃21% at z ≃ 5.7. These LEW LAEs are candidates of young metal-poor galaxies and AGNs. We search for LABs that are LAEs with spatially extended Lyα emission that have profiles clearly distinguished from those of stellar objects at the ≳ 3σ level. In the search, we identify 11 LABs in the HSC NB images down to a surface brightness limit of ≃5–8 × 10−18 erg s−1 cm−2, which is as deep as data of previous studies. The number density of the LABs at z ≃ 6–7 is ∼10−7–10−6 Mpc−3, which is ∼10–100 times lower than those claimed for LABs at z ≃ 2–3, suggestive of disappearing LABs at z ≳ 6, although the selection methods are different in the low- and high-z LABs. It should be noted that Lyα EW scale length derivation methods and the LAB selections are not homogeneous in a redshift range of z ≃ 0–7. Using the future z ≃ 2.2, 5.7, 6.6, and 7.3 HSC NB data, we will systematically investigate the redshift evolution of Lyα EW scale lengths and NLAB at z ≃ 2–7 using homogeneous methods. Acknowledgements We would like to thank James Bosch, Richard S. Ellis, Masao Hayashi, Robert H. Lupton, and Michael A. Strauss for useful discussion and comments. We thank the anonymous referee for constructive comments and suggestions. This work is based on observations taken by the Subaru Telescope and the Keck telescope which are operated by the National Observatory of Japan. This work was supported by World Premier International Research Center Initiative (WPI Initiative), MEXT, Japan, KAKENHI (23244025) and (21244013) Grant-in-Aid for Scientific Research (A) through Japan Society for the Promotion of Science (JSPS), and an Advanced Leading Graduate Course for Photon Science grant. The NB816 filter was supported by Ehime University (PI: Y. Taniguchi). The NB921 filter was supported by KAKENHI (23244025) Grant-in-Aid for Scientific Research (A) through the Japan Society for the Promotion of Science (PI: M. Ouchi). NK is supported by the JSPS grant 15H03645. SY is supported by Faculty of Science, Mahidol University, Thailand and the Thailand Research Fund (TRF) through a research grant for new scholar (MRG5980153). The Hyper Suprime-Cam (HSC) collaboration includes the astronomical communities of Japan and Taiwan, and Princeton University. The HSC instrumentation and software were developed by the National Astronomical Observatory of Japan (NAOJ), the Kavli Institute for the Physics and Mathematics of the Universe (Kavli IPMU), the University of Tokyo, the High Energy Accelerator Research Organization (KEK), the Academia Sinica Institute for Astronomy and Astrophysics in Taiwan (ASIAA), and Princeton University. Funding was contributed by the FIRST program from Japanese Cabinet Office, the Ministry of Education, Culture, Sports, Science and Technology (MEXT), the Japan Society for the Promotion of Science (JSPS), Japan Science and Technology Agency (JST), the Toray Science Foundation, NAOJ, Kavli IPMU, KEK, ASIAA, and Princeton University. This paper makes use of software developed for the Large Synoptic Survey Telescope. We thank the LSST Project for making their code available as free software at ⟨http://dm.lsst.org⟩. The Pan-STARRS1 Surveys (PS1) have been made possible through contributions of the Institute for Astronomy, the University of Hawaii, the Pan-STARRS Project Office, the Max-Planck Society and its participating institutes, the Max Planck Institute for Astronomy, Heidelberg and the Max Planck Institute for Extraterrestrial Physics, Garching, The Johns Hopkins University, Durham University, the University of Edinburgh, Queen’s University Belfast, the Harvard-Smithsonian Center for Astrophysics, the Las Cumbres Observatory Global Telescope Network Incorporated, the National Central University of Taiwan, the Space Telescope Science Institute, the National Aeronautics and Space Administration under Grant No. NNX08AR22G issued through the Planetary Science Division of the NASA Science Mission Directorate, the National Science Foundation under Grant No. AST-1238877, the University of Maryland, and Eotvos Lorand University (ELTE) and the Los Alamos National Laboratory. This paper is based on data collected at the Subaru Telescope and retrieved from the HSC data archive system, which is operated by the Subaru Telescope and Astronomy Data Center at the National Astronomical Observatory of Japan. Appendix. Calculation of Lyα EW In this section, we describe the method used to calculate the EW0,Lyα values. The procedures and assumptions of this method are similar to those of e.g., Malhotra and Rhoads (2002), Dawson et al. (2004), Gronwall et al. (2007), and Kashikawa et al. (2011). For the calculation of EW0,Lyα, we assume that LAEs have a δ-function shaped Lyα line and a flat rest-frame UV continuum emission (i.e., βν = 0, where βν is the UV spectral slope per unit frequency). In such an LAE spectrum, the magnitude, m, of a waveband filter with a transmission curve, Tν, is described as follows:   \begin{eqnarray} 48.6 + m = -2.5 \log _{10} \frac{\int _0^\infty [f_{\rm c} + f_{\rm l} \delta (\nu - \nu _\alpha )] T_\nu {\rm d}\nu }{\int _0^\infty T_\nu {\rm d}\nu } , \end{eqnarray} (A1)where fl, fc, δ(ν), and να are the Lyα line flux, the flux density of the rest-frame UV continuum emission, the δ function, and the observed frequency of Lyα, respectively. Here we also assume that the Lyα line is located at 9215 Å (8177 Å), which is the central wavelength of the NB921 (NB816) filter, for z ≃ 6.6(z ≃ 5.7) LAEs. In this study, we do not take into account the IGM transmission for Lyα, if not specified. This is because the IGM transmission for Lyα highly depends on the Lyα line velocity offset from the systemic redshift (e.g., Hashimoto et al. 2013; Shibuya et al. 2014b). The numerator of the logarithm in equation (A1) corresponds to   \begin{eqnarray} && f_{\rm c} \int ^\infty _{\nu_{\rm c}} \exp {(-\tau _{\rm eff})} T_\nu {\rm d}\nu + f_{\rm c} \int _0^{\nu_{\rm c}} T_\nu {\rm d}\nu + f_{\rm l} T_\nu (\nu _\alpha ) \nonumber \\ && \quad = f_{\rm c} B + f_{\rm c} R + f_{\rm l} T_\nu (\nu _\alpha ). \end{eqnarray} (A2)In equation (A2), we use B and R that are defined by   \begin{eqnarray} B \equiv \int^\infty _{\nu_{\rm c}} \exp {(-\tau _{\rm eff})} T_\nu {\rm d}\nu , \end{eqnarray} (A3)  \begin{eqnarray} R \equiv \int _0^{\nu_{\rm c}} T_\nu {\rm d}\nu , \end{eqnarray} (A4)where τeff is the IGM optical depth calculated from the analytical models of Madau (1995). Using equations (A1) and (A2), we derive the flux density of the NB and BB filters, $$\overline{f^{\vphantom{i}}_{\rm NB}}$$ and $$\overline{f^{\vphantom{i}}_{\rm BB}}$$, as follows:   \begin{eqnarray} \overline{f^{\vphantom{i}}_{\rm NB}} = 10^{-0.4(m_{\rm NB} + 48.6)} = \frac{f_{\rm c} (B_{\rm NB} + R_{\rm NB}) + f_{\rm l} T_{\rm NB} (\nu _\alpha ) }{A_{\rm NB}}, \end{eqnarray} (A5)  \begin{eqnarray} \overline{f^{\vphantom{i}}_{\rm BB}} = 10^{-0.4(m_{\rm BB} + 48.6)} = \frac{f_{\rm c} (B_{\rm BB} + R_{\rm BB}) }{A_{\rm BB}}. \end{eqnarray} (A6)Here the following definition is used for the denominator of equations (A5) and (A6).   \begin{eqnarray} A \equiv \int_0^\infty T_\nu {\rm d}\nu , \end{eqnarray} (A7)The B, R, and A values with the subscripts of NB(BB) are calculated with the transmission curves of the NB(BB) filters, TNB(TBB). In this study, we use magnitudes of the y- and z-band filters which do not cover the wavelength of Lyα for z ≃ 6.6 and z ≃ 5.7 LAEs, respectively, indicating TBB(να) = 0. In the case that mBB is fainter than the 1σ limit, we use the 1σ limiting magnitude for the EW0,Lyα calculation. By combining the equations of $$\overline{f^{\vphantom{i}}_{\rm NB}}$$ and $$\overline{f^{\vphantom{i}}_{\rm BB}}$$, we obtain fc and fl,   \begin{eqnarray} f_{\rm c} = \frac{A_{\rm BB} \overline{f^{\vphantom{i}}_{\rm BB}}}{B_{\rm BB} + R_{\rm BB}} = \overline{f^{\vphantom{i}}_{\rm BB}}, \end{eqnarray} (A8)  \begin{eqnarray*} f_{\rm l} = \frac{A_{\rm NB}(B_{\rm BB} + R_{\rm BB})\overline{f^{\vphantom{i}}_{\rm NB}} - A_{\rm BB}(B_{\rm NB} + R_{\rm NB})\overline{f^{\vphantom{i}}_{\rm BB}}}{ (B_{\rm BB} + R_{\rm BB})T_{\rm NB} (\nu _\alpha ) } \end{eqnarray*}   \begin{eqnarray*} \hphantom{f_{\rm l}} = \frac{A_{\rm NB}\overline{f^{\vphantom{i}}_{\rm NB}} - (B_{\rm NB} + R_{\rm NB})\overline{f^{\vphantom{i}}_{\rm BB}}}{T_{\rm NB} (\nu_\alpha)} \end{eqnarray*}   \begin{eqnarray} \hphantom{f_{\rm l}} = a \overline{f^{\vphantom{i}}_{\rm NB}} - b \overline{f^{\vphantom{i}}_{\rm BB}}. \end{eqnarray} (A9)Note that BBB + RBB = ABB due to the negligible IGM absorption at the wavelengths of the BB filters. Here we define a and b as   \begin{eqnarray} a \equiv \frac{A_{\rm NB}}{T_{\rm NB} (\nu _\alpha )}, \end{eqnarray} (A10)  \begin{eqnarray} b \equiv \frac{B_{\rm NB} + R_{\rm NB}}{T_{\rm NB} (\nu _\alpha )}. \end{eqnarray} (A11)For the HSC NB921 and NB816 filters, the sets of the values are calculated to be (a, b) ≃ (4.7, 2.3) × 1012 and (a, b) ≃ (5.2, 2.7) × 1012, respectively. Using fc and fl, we calculate the EW0,Lyα values via   \begin{equation} EW_{\rm 0,\, Ly\alpha } = \frac{f_{\rm l}}{f_{\rm c}} \frac{c}{\nu ^2} \frac{1}{1+z}. \end{equation} (A12)To obtain the median values and uncertainties for EW0,Lyα, we perform MC simulations in a method similar to that of e.g., Shimasaku et al. (2006). In the simulation, we randomly generate a flux density value, $$\overline{f^{\vphantom{i}}_{\rm MC}}$$, following a Gaussian probability distribution with an average of $$\overline{f^{\vphantom{i}}}$$ and a dispersion of the 1σ sky background noise, $$\overline{f^{\vphantom{i}}_{\rm 1\sigma }}$$, for the NB and BB bands. Here we also randomize βν and να in Gaussian probability distributions with 1σ dispersions of Δβ = 0.2 and Δνα = FWHMNB/2.35, respectively, where FWHMNB is the FWHM of the NB filters. The dispersion of Δβ = 0.2 is typical for high-z galaxies (Bouwens et al. 2014). We calculate an EW0,Lyα value using $$\overline{f^{\vphantom{i}}_{\rm MC}}$$ for NB and BB via equation (A12). In this process, negative values of fc, fl, and EW0,Lyα are forced to be zero. Such a process is performed 1000 times for each object. During the iteration, a simulated EW0,Lyα value is discarded in the case that a BB − NB color does not meet the selection criteria of equations (1) and (2). Using the set of EW0,Lyα values obtained from the MC simulations, we calculate the median values and the 16- and 84-percentile errors for EW0,Lyα. Footnotes * Based on data obtained with the Subaru Telescope. The Subaru Telescope is operated by the National Astronomical Observatory of Japan. 1 ⟨https://www.naoj.org/Observing/Instruments/HSC/sensitivity.html⟩. 2 ⟨http://cos.icrr.u-tokyo.ac.jp/rush.html⟩. References Aihara H. et al.   2018a, PASJ , 70, S4 Aihara H. et al.   2018b, PASJ , 70, S8 Ajiki M. et al.   2002, ApJ , 576, L25 CrossRef Search ADS   Arrigoni Battaia F., Hennawi J. F., Prochaska J. X., Cantalupo S. 2015a, ApJ , 809, 163 CrossRef Search ADS   Arrigoni Battaia F., Yang Y., Hennawi J. F., Prochaska J. X., Matsuda Y., Yamada T., Hayashino T. 2015b, ApJ , 804, 26 CrossRef Search ADS   Axelrod T., Kantor J., Lupton R. H., Pierfederici F. 2010, in Proc. SPIE, 7740, Software and Cyberinfrastructure for Astronomy , ed. Radziwill N. M., Bridger A. ( Bellingham, WA: SPIE), 774015 Bertin E., Arnouts S. 1996, A&AS , 117, 393 CrossRef Search ADS   Blanc G. A. et al.   2011, ApJ , 736, 31 CrossRef Search ADS   Bond N. A., Gawiser E., Guaita L., Padilla N., Gronwall C., Ciardullo R., Lai K. 2012, ApJ , 753, 95 CrossRef Search ADS   Bosch J. et al.   2018, PASJ , 70, S5 Bouwens R. J. et al.   2014, ApJ , 793, 115 CrossRef Search ADS   Cai Z. et al.   2017, ApJ , 837, 71 CrossRef Search ADS   Cantalupo S., Arrigoni-Battaia F., Prochaska J. X., Hennawi J. F., Madau P. 2014, Nature , 506, 63 CrossRef Search ADS PubMed  Cantalupo S., Porciani C., Lilly S. J., Miniati F. 2005, ApJ , 628, 61 CrossRef Search ADS   Ciardullo R. et al.   2012, ApJ , 744, 110 CrossRef Search ADS   Cowie L. L., Barger A. J., Hu E. M. 2010, ApJ , 711, 928 CrossRef Search ADS   Cowie L. L., Hu E. M., Songaila A. 2011, ApJ , 735, L38 CrossRef Search ADS   Dawson S. et al.   2004, ApJ , 617, 707 CrossRef Search ADS   Dekel A. et al.   2009, Nature , 457, 451 CrossRef Search ADS PubMed  Dressler A., Martin C. L., Henry A., Sawicki M., McCarthy P. 2011, ApJ , 740, 71 CrossRef Search ADS   Duval F., Schaerer D., Östlin G., Laursen P. 2014, A&A , 562, A52 CrossRef Search ADS   Erb D. K., Bogosavljević M., Steidel C. C. 2011, ApJ , 740, L31 CrossRef Search ADS   Finkelstein S. L., Rhoads J. E., Malhotra S., Grogin N., Wang J. 2008, ApJ , 678, 655 CrossRef Search ADS   Finkelstein S. L., Rhoads J. E., Malhotra S., Pirzkal N., Wang J. 2007, ApJ , 660, 1023 CrossRef Search ADS   Gawiser E. et al.   2007, ApJ , 671, 278 CrossRef Search ADS   Gronwall C. et al.   2007, ApJ , 667, 79 CrossRef Search ADS   Guaita L. et al.   2011, ApJ , 733, 114 CrossRef Search ADS   Haiman Z., Abel T., Rees M. J. 2000, ApJ , 534, 11 CrossRef Search ADS   Hansen M., Oh S. P. 2006, MNRAS , 367, 979 CrossRef Search ADS   Harikane Y. et al.   2018, PASJ , 70, S11 Hashimoto T., Ouchi M., Shimasaku K., Ono Y., Nakajima K., Rauch M., Lee J., Okamura S. 2013, ApJ , 765, 70 CrossRef Search ADS   Hashimoto T., Ouchi M., Shimasaku K., Schaerer D., Nakajima K., Shibuya T., Ono Y., Rauch M., Goto R. 2017, MNRAS , 465, 1543 CrossRef Search ADS   Hayes M., Scarlata C., Siana B. 2011, Nature , 476, 304 CrossRef Search ADS PubMed  Hennawi J. F., Prochaska J. X., Cantalupo S., Arrigoni-Battaia F. 2015, Science , 348, 779 CrossRef Search ADS PubMed  Hu E. M., Cowie L. L., Barger A. J., Capak P., Kakazu Y., Trouille L. 2010, ApJ , 725, 394 CrossRef Search ADS   Ivezic Z. et al.   2008, arXiv:0805.2366 Jurić M. et al.   2015, arXiv:1512.07914 Kashikawa N. et al.   2006, ApJ , 648, 7 CrossRef Search ADS   Kashikawa N. et al.   2011, ApJ , 734, 119 CrossRef Search ADS   Kashikawa N. et al.   2012, ApJ , 761, 85 CrossRef Search ADS   Kawanomoto S. et al.   2017, PASJ , submitted Keel W. C., White R. E. III, Chapman S., Windhorst R. A. 2009, AJ , 138, 986 CrossRef Search ADS   Kereš D., Katz N., Fardal M., Davé R., Weinberg D. H. 2009, MNRAS , 395, 160 CrossRef Search ADS   Kodaira K. et al.   2003, PASJ , 55, L17 CrossRef Search ADS   Kojima T., Ouchi M., Nakajima K., Shibuya T., Harikane Y., Ono Y. 2017, PASJ , 69, 44 CrossRef Search ADS   Konno A. et al.   2014, ApJ , 797, 16 CrossRef Search ADS   Konno A. et al.   2018, PASJ , 70, S16 Kusakabe H., Shimasaku K., Nakajima K., Ouchi M. 2015, ApJ , 800, L29 CrossRef Search ADS   Laursen P., Duval F., Östlin G. 2013, ApJ , 766, 124 CrossRef Search ADS   Laursen P., Sommer-Larsen J. 2007, ApJ , 657, L69 CrossRef Search ADS   Laursen P., Sommer-Larsen J., Andersen A. C. 2009, ApJ , 704, 1640 CrossRef Search ADS   Lilly S. J., Le Fevre O., Hammer F., Crampton D. 1996, ApJ , 460, L1 CrossRef Search ADS   Madau P. 1995, ApJ , 441, 18 CrossRef Search ADS   Madau P., Dickinson M. 2014, ARA&A , 52, 415 CrossRef Search ADS   Madau P., Ferguson H. C., Dickinson M. E., Giavalisco M., Steidel C. C., Fruchter A. 1996, MNRAS , 283, 1388 CrossRef Search ADS   Magnier E. A. et al.   2013, ApJS , 205, 20 CrossRef Search ADS   Malhotra S., Rhoads J. E. 2002, ApJ , 565, L71 CrossRef Search ADS   Malhotra S., Rhoads J. E. 2004, ApJ , 617, L5 CrossRef Search ADS   Malhotra S., Rhoads J. E. 2006, ApJ , 647, L95 CrossRef Search ADS   Mallery R. P. et al.   2012, ApJ , 760, 128 CrossRef Search ADS   Mas-Ribas L., Dijkstra M. 2016, ApJ , 822, 84 CrossRef Search ADS   Mas-Ribas L., Dijkstra M., Hennawi J. F., Trenti M., Momose R., Ouchi M. 2017, ApJ , 841, 19 CrossRef Search ADS   Matsuda Y. et al.   2004, AJ , 128, 569 CrossRef Search ADS   Matsuda Y. et al.   2009, MNRAS , 400, L66 CrossRef Search ADS   Matsuda Y. et al.   2011, MNRAS , 410, L13 CrossRef Search ADS   Matthee J., Sobral D., Santos S., Röttgering H., Darvish B., Mobasher B. 2015, MNRAS , 451, 400 CrossRef Search ADS   Momose R., Ouchi M., Nakajima K., Ono Y., Shibuya T., Shimasaku K., Yuma S., Mori M., Umemura M. 2016, MNRAS , 457, 2318 CrossRef Search ADS   Mori M., Umemura M. 2006, Nature , 440, 644 CrossRef Search ADS PubMed  Murayama T. et al.   2007, ApJS , 172, 523 CrossRef Search ADS   Nagao T. et al.   2008, ApJ , 680, 100 CrossRef Search ADS   Nakajima K., Ouchi M. 2014, MNRAS , 442, 900 CrossRef Search ADS   Nakajima K., Ouchi M., Shimasaku K., Hashimoto T., Ono Y., Lee J. C. 2013, ApJ , 769, 3 CrossRef Search ADS   Nakajima K. et al.   2012, ApJ , 745, 12 CrossRef Search ADS   Neufeld D. A. 1991, ApJ , 370, L85 CrossRef Search ADS   Nilsson K. K., Tapken C., Møller P., Freudling W., Fynbo J. P. U., Meisenheimer K., Laursen P., Östlin G. 2009, A&A , 498, 13 CrossRef Search ADS   Oke J. B., Gunn J. E. 1983, ApJ , 266, 713 CrossRef Search ADS   Ono Y. et al.   2010, MNRAS , 402, 1580 CrossRef Search ADS   Ono Y. et al.   2018, PASJ , 70, S10 Ota K. et al.   2017, ApJ , 844, 85 CrossRef Search ADS   Ouchi M. et al.   2008, ApJS , 176, 301 CrossRef Search ADS   Ouchi M. et al.   2009, ApJ , 696, 1164 CrossRef Search ADS   Ouchi M. et al.   2010, ApJ , 723, 869 CrossRef Search ADS   Ouchi M. et al.   2018, PASJ , 70, S13 Oyarzún G. A., Blanc G. A., González V., Mateo M., Bailey J. I., III, Finkelstein S. L., Lira P., Crane J. D., Olszewski E. W. 2016, ApJ , 821, L14 CrossRef Search ADS   Planck Collaboration 2016, A&A , 594, A13 CrossRef Search ADS   Prescott M. K. M. et al.   2012b, ApJ , 752, 86 CrossRef Search ADS   Prescott M. K. M., Dey A., Jannuzi B. T. 2009, ApJ , 702, 554 CrossRef Search ADS   Prescott M. K. M., Dey A., Jannuzi B. T. 2012a, ApJ , 748, 125 CrossRef Search ADS   Prescott M. K. M., Dey A., Jannuzi B. T. 2013, ApJ , 762, 38 CrossRef Search ADS   Prescott M. K. M., Martin C. L., Dey A. 2015, ApJ , 799, 62 CrossRef Search ADS   Rhoads J. E., Malhotra S. 2001, ApJ , 563, L5 CrossRef Search ADS   Saito T., Shimasaku K., Okamura S., Ouchi M., Akiyama M., Yoshida M. 2006, ApJ , 648, 54 CrossRef Search ADS   Salpeter E. E. 1955, ApJ , 121, 161 CrossRef Search ADS   Schlafly E. F. et al.   2012, ApJ , 756, 158 CrossRef Search ADS   Shibuya T., Kashikawa N., Ota K., Iye M., Ouchi M., Furusawa H., Shimasaku K., Hattori T. 2012, ApJ , 752, 114 CrossRef Search ADS   Shibuya T., Ouchi M., Kubo M., Harikane Y. 2016, ApJ , 821, 72 CrossRef Search ADS   Shibuya T., Ouchi M., Nakajima K., Yuma S., Hashimoto T., Shimasaku K., Mori M., Umemura M. 2014a, ApJ , 785, 64 CrossRef Search ADS   Shibuya T. et al.   2014b, ApJ , 788, 74 CrossRef Search ADS   Shibuya T. et al.   2018, PASJ , 70, S15 Shimasaku K. et al.   2006, PASJ , 58, 313 CrossRef Search ADS   Sobral D., Matthee J., Darvish B., Schaerer D., Mobasher B., Röttgering H. J. A., Santos S., Hemmati S. 2015, ApJ , 808, 139 CrossRef Search ADS   Stark D. P., Ellis R. S., Ouchi M. 2011, ApJ , 728, L2 CrossRef Search ADS   Steidel C. C., Adelberger K. L., Shapley A. E., Pettini M., Dickinson M., Giavalisco M. 2000, ApJ , 532, 170 CrossRef Search ADS   Taniguchi Y. et al.   2005, PASJ , 57, 165 CrossRef Search ADS   Taniguchi Y. et al.   2009, ApJ , 701, 915 CrossRef Search ADS   Taniguchi Y., Shioya Y. 2000, ApJ , 532, L13 CrossRef Search ADS   Tonry J. L. et al.   2012, ApJ , 750, 99 CrossRef Search ADS   Toshikawa J. et al.   2018, PASJ , 70, S12 Yajima H., Li Y., Zhu Q. 2013, ApJ , 773, 151 CrossRef Search ADS   Yajima H., Li Y., Zhu Q., Abel T., Gronwall C., Ciardullo R. 2014, MNRAS , 440, 776 CrossRef Search ADS   Yang Y., Zabludoff A., Eisenstein D., Davé R. 2010, ApJ , 719, 1654 CrossRef Search ADS   Yang Y., Zabludoff A., Tremonti C., Eisenstein D., Davé R. 2009, ApJ , 693, 1579 CrossRef Search ADS   Zheng Z., Cen R., Trac H., Miralda-Escudé J. 2010, ApJ , 716, 574 CrossRef Search ADS   Zheng Z., Wallace J. 2014, ApJ , 794, 116 CrossRef Search ADS   Zheng Z.-Y., Wang J.-X., Malhotra S., Rhoads J. E., Finkelstein S. L., Finkelstein K. 2014, MNRAS , 439, 1101 CrossRef Search ADS   Zheng Z.-Y. et al.   2017, ApJ , 842, L22 CrossRef Search ADS   © The Author(s) 2017. Published by Oxford University Press on behalf of the Astronomical Society of Japan. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Journal

Publications of the Astronomical Society of JapanOxford University Press

Published: Jan 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve Freelancer

DeepDyve Pro

Price
FREE
$49/month

$360/year
Save searches from Google Scholar, PubMed
Create lists to organize your research
Export lists, citations
Access to DeepDyve database
Abstract access only
Unlimited access to over
18 million full-text articles
Print
20 pages/month
PDF Discount
20% off