(PDF) Η αισθητική της οχυρωματικής αρχιτεκτονικής

MNRAS 000, 1–15 (2016) Preprint 8 October 2018 Compiled using MNRAS LATEX style file v3.0 Herschel protocluster survey: A search for dusty star-forming galaxies in protoclusters at z = 2 − 3 Y. Kato,1,2⋆ Y. Matsuda,1,3 Ian Smail,4,5 A. M. Swinbank,4,5 B. Hatsukade,1 H. Umehata,6,7 I. Tanaka,8 T. Saito,1 D. Iono,1,3 Y. Tamura,7 K. Kohno,7,9 D. K. Erb,10 B. D. Lehmer,11 J. E. Geach,12 C. C. Steidel,13 D. M. Alexander,4 T. Yamada14 and arXiv:1605.07370v1 [astro-ph.GA] 24 May 2016 T. Hayashino15 1 National Astronomical Observatory of Japan, 2-21-1 Osawa, Mitaka, Tokyo, 181-8588, Japan 2 Department of Astronomy, Graduate school of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 133-0033, Japan 3 Department of Astronomy, School of Science, The Graduate University for Advanced Studies (SOKENDAI), Osawa, Mitaka, Tokyo 181-8588, Japan 4 Centre for Extragalactic Astronomy, Department of Physics, Durham University, South Road, Durham, DH1 3LE, UK 5 Institute for Computational Cosmology, Durham University, South Road, Durham DH1 3LE, UK 6 European Southern Observatory, Karl-Schwarzschild-Str. 2, D-85748 Garching, Germany 7 Institute of Astronomy, School of Science, The University of Tokyo, 2-21-1 Osawa, Mitaka, Tokyo 181-0015, Japan 8 Subaru Telescope, National Astronomical Observatory of Japan, 650 North AâĂŹohoku Place, Hilo, HI 96720, USA 9 Research Center for the Early Universe, The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113-0033 10 Center for Gravitation, Cosmology and Astrophysics, Department of Physics, University of Wisconsin Milwaukee 3135 North Maryland Avenue, Milwaukee, Wisconsin 53211, USA 11 Department of Physics, University of Arkansas, 226 Physics Building, 835 West Dickson Street, Fayetteville, AR 72701, USA 12 Centre for Astrophysics Research, Science & Technology Research Institute, University of Hertfordshire, Hatfield, AL10 9AB, UK 13 California Institute of Technology, MS 249-17, Pasadena, CA 91125, USA 14 Institute of Space Astronautical Science, Japan Aerospace Exploration Agency, Sagamihara, Kanagawa 252-5210, Japan 15 Research Center for Neutrino Science, Tohoku University, Sendai, Miyagi, 980-8578, Japan Accepted 2016 May 19. Received 2016 May 16; in original form 2016 March 18 ABSTRACT We present a Herschel /SPIRE survey of three protoclusters at z = 2 − 3 (2QZCluster, HS1700, SSA22). Based on the SPIRE colours (S350 /S250 and S500 /S350 ) of 250 µm sources, we selected high redshift dusty star-forming galaxies potentially associated with the protoclusters. In the 2QZCluster field, we found a 4σ overdensity of six SPIRE sources around 4.5′ (∼ 2.2 Mpc) from a density peak of Hα emitters at z = 2.2. In the HS1700 field, we found a 5σ overdensity of eight SPIRE sources around 2.1′ (∼ 1.0 Mpc) from a density peak of LBGs at z = 2.3. We did not find any significant overdensities in SSA22 field, but we found three 500 µm sources are concentrated 3′ (∼1.4 Mpc) east to the LAEs overdensity. If all the SPIRE sources in these three over- densities are associated with protoclusters, the inferred star-formation rate densities are 103 − 104 times higher than the average value at the same redshifts. This sug- gests that dusty star-formation activity could be very strongly enhanced in z ∼ 2 − 3 protoclusters. Further observations are needed to confirm the redshifts of the SPIRE sources and to investigate what processes enhance the dusty star-formation activity in z ∼ 2 − 3 protoclusters. Key words: galaxies: clusters: individual – galaxies: individual – galaxies: formation – galaxies: high-redshift – submillimetre: galaxies – infrared: galaxies 1 INTRODUCTION tions are old, with inferred formation redshifts of z & 2 (e.g., Ellis et al. 1997). High-redshift dusty star-forming galaxies The central regions of local clusters are dominated by passive (DSFGs) are strongly star-forming galaxies (SFR & 100 − early-type ellipticals and spheroidals, their stellar popula- 1000 M⊙ yr−1 ) and have been proposed to the precursors of present-day ellipticals in local clusters (e.g., Lilly et al. 1998; Smail et al. 1998; Lutz et al. 2001; Ivison et al. 2013). ⋆ E-mail:

c 2016 The Authors 2 Y. Kato et al. Large hydro-dynamical simulations and galaxy formation Matsuda). We summarize the observations in Table 1. The models predict intense star formation could be detectable observations were executed in Large Map mode with a scan as concentrations of DSFGs in z & 2 protoclusters (e.g., rate of 30′′ s−1 , repeated 14 times for each field (Nrep =14). Granato et al. 2015). Indeed, there have been reports of a The dates of observations are 22 June 2012 (2QZCluster), reversal of the SFR-density relation (e.g., Elbaz et al. 2007; 4 March 2012 (HS1700), and 10 May 2012 (SSA22). The Tran et al. 2010), which is increasing SFR with increasing coverage of the maps are ∼ 23′ × 23′ (2QZCluster), ∼ local density at z & 1. Protoclusters at z & 2 are thus unique 22′ ×22′ (HS1700), and ∼ 33′ ×33′ (SSA22) corresponding to laboratories to explore bursting star-formation in a critical ∼ 40−60 comoving Mpc at the protocluster redshifts, which epoch of galaxy formation (Casey 2016). are sufficient to search for concentration of DSFGs around A numbers of studies have confirmed the pres- the density peak of protocluster members. The integration ence of DSFGs in z < 2 galaxy clusters (e.g., times are 1.8, 1.5 and 3.7 hours for 2QZCluster, HS1700 and Brodwin et al. 2013; Smail et al. 2014; Ma et al. 2015; SSA22, respectively. Maps were produced with the Herschel Webb et al. 2015), while studies confirming DSFGs at z & Interactive Processing Environment (HIPE, v11.0.0), follow- 2 protoclusters are also progressing (e.g., Tamura et al. ing the standard data processing and map-making steps with 2009; Clements et al. 2014; Umehata et al. 2014, 2015; destriping. The full width at half maximum (FWHM) of Casey et al. 2015). Some surveys to search for DSFGs the SPIRE beam is 18.1′′ , 24.9′′ and 36.6′′ at 250, 350 and at far-infrared wavelengths have focused on radio galaxy 500 µm, respectively (Swinyard et al. 2010). The final maps fields (e.g., Stevens et al. 2003; Valtchanov et al. 2013; have pixel sizes of 6′′ , 10′′ , and 14′′ at 250, 350, and 500 µm. Rigby et al. 2014; Dannerbauer et al. 2014). Radio galax- We measured the 1σ confusion noise (σconf ) as map vari- ies are thought to be tracers of large scale structures, and ance of flux density within effective area (Table 1), which are some fraction of z & 2 protoclusters around radio galax- slightly higher than the values of blank fields (Nguyen et al. ies indeed appear to have experienced bursting dusty star- 2010). The instrumental noise measured in our three proto- formation related to DSFGs (e.g., Ivison et al. 2000). How- clusters are about one-third of the confusion noise (Table 1). ever, such regions may be biased by the end to host a cur- rently accreting Super Massive Black Hole (SMBH) and so we also need to explore DSFGs large-scale structures se- 2.2 PROTOCLUSTER TARGETS lected by other techniques at z & 2 to investigate bursting 2QZCluster; This protocluster was originally identified as star-formation more generally. Clements et al. (2014) inves- a concentration of five Quasi Stellar Objects (QSOs) in a ∼ tigated ∼ 90 deg2 sky observed as part of the HerMES sur- 1 degree region at z = 2.23 from the 2dF Quasar Redshift vey with Planck and Herschel to search for clusters under- survey (Croom et al. 2001, 2004). Four out of the five QSOs going dusty star-formation. They found four candidate clus- are even more strongly clustered in a 30 × 30 comoving Mpc ters, and for all four cases they found evidence of galaxy patch. An Hα narrow-band imaging revealed a filamentary clusters with red-sequence based on optical/NIR data. The large-scale structure of an over density of 22 HAEs connect- star-formation rate density of these at z & 2 are four order of ing the QSOs (Matsuda et al. 2011). Chandra/ACIS-I 100 ks magnitudes higher than the cosmic averaged values. But is observations of this structure also showed evidence that the this also true for dusty star-formation in known optical/UV- Active Galactic Nucleus (AGNs) fraction is a factor of ∼ 3.5 selected galaxy clusters at z & 2? higher than blank fields (Lehmer et al. 2013). In this paper, we report a result of observations with the Spectral and Photometric Imaging Receiver (SPIRE; HS1700; This protocluster was originally discovered as Griffin et al. 2010) on the Herschel Space Observatory a ∼ 7× density contrast redshift spike of UV/optical- selected star forming galaxies (BX/BM) within a ∼ (HSO; Pilbratt et al. 2010) for three protoclusters at z = 2 − 3 (2QZCluster, HS1700, and SSA22 at z = 2.2, 2.3, 25 comoving Mpc region at z = 2.30 (Steidel et al. and 3.1, respectively). The three protoclusters have filamen- 2005). A Lyα narrowband imaging survey revealed a filamentary large-scale structure of six giant Lyα tary, large scale structures of rest-frame UV to optical se- lected galaxies. The structure of this paper is following. In Blobs (LABs) (Erb, Bogosavljević, & Steidel 2011). Chan- §2 we introduce the Herschel /SPIRE observations, the data dra/ACIS-I 200 ks observations of this structure also showed tentative evidence of an enhancement of AGN fraction com- processing, and our targeted fields. In §3 we present source detection methods, number counts, and SPIRE colour se- pared to the field environment (Digby-North et al. 2010). lection. In §4 and §5 we present our results, discussion, and SSA22; This protocluster was originally discovered as a then summarize our main findings. We use the following cos- ∼ 4 − 6× density contrast in redshift distribution of Lyman mological parameters: Ωm = 0.3, ΩΛ = 0.7, h = 0.7. In this Break Galaxies (LBGs) and Lyα emitters (LAEs) within a cosmology, the Universe is 2.9, 2.8, and 2.0 Gyr old and ∼ 20 comoving Mpc region at z = 3.09 (Steidel et al. 1998, 1.0′′ corresponds to 8.3, 8.2 and 7.6 kpc in physical length 2000). A wide-field Lyα narrowband imaging survey with at z = 2.2, 2.3 and 3.1, respectively. Subaru Telescope revealed a filamentary large-scale struc- ture of 283 LAEs, extended to at least ∼ 60 comoving Mpc (Hayashino et al. 2004). This sample of LAEs includes 35 LABs with sizes of 30–150 kpc scale (Steidel et al. 2000; 2 OBSERVATIONS & TARGET FIELDS Matsuda et al. 2004). Chandra/ACIS-I 400 ks observations of this structure showed that the AGN fraction of protoclus- 2.1 SPIRE OBSERVATIONS ter members is ∼ 3× higher than that in the field environ- Our Herschel /SPIRE observations were performed as part ment (Lehmer et al. 2009a,b). of the second Open Time (OT2) Herschel programs (PI: Y. Blank field (COSMOS); We have chosen well-studied ex- MNRAS 000, 1–15 (2016) Herschel protocluster survey 3 Table 1. Summary of our Herschel /SPIRE observations. Target za R.A.b Decc Aread tint e σconf f σinst g S250 S350 S500 S250 S350 S500 (J2000) (J2000) (arcmin2 ) (hours) (mJy) (mJy) (mJy) (mJy) (mJy) (mJy) 2QZCluster 2.230 ± 0.016 10h03m51s +00d15m09s 515 1.8 6.7 7.0 7.1 2.0–3.9 1.6–2.6 2.0–3.3 HS1700 2.300 ± 0.015 17h01m15s +64d14m03s 497 1.5 7.3 7.4 7.5 2.0–3.7 1.6–2.7 2.0–3.2 SSA22 3.09 ± 0.03 22h17m34s +00d17m01s 1076 3.7 7.7 7.8 8.0 1.9–3.2 1.6–2.4 1.9–3.0 COSMOS - 10h00m37s +02d11m26s 3422 50.1 7.1 7.7 7.8 2.1–2.7 1.7–2.0 2.1–2.8 Notes. (a): Redshift range of the member galaxies, based on HAEs for 2QZCluster (Matsuda et al. 2011) and LBGs for HS1700 (Steidel et al. 2005) and SSA22 (Steidel et al. 1998, 2000). (b), (c): The coordinates of the field centre of Herschel /SPIRE observations. (d): The area where the integration time is greater than 30%. For COSMOS field, we used centre of 3422 arcmin2 area. We detected sources within this area. (see also Figure 3). (e): The total integration time. For the survey design of COSMOS field, please see Oliver et al. (2012). (f ), (g): The confusion noise and instrumental noise. Figure 1. Number counts at 250, 350 and 500 µm in the three protoclusters and COSMOS field. We detected S/N > 2 sources in the 250 µm maps and measured the 350 and 500 µm fluxes at the positions of the 250 µm sources. We fitted the data points of COSMOS filed with a bézier curve. The number counts of the SPIRE sources in the protoclusters averaged over their full fields are roughly consistent with those of COSMOS field. The number counts of the SPIRE sources in COSMOS field are also consistent with those of HerMES blank fields (Béthermin et al. 2012), suggesting that COSMOS field can be used as a control field. tragalactic field Cosmic Evolution Survey (COSMOS) to use region where the integration time is greater than 30% of as a blank field, which is observed as a part of Herschel the deepest parts (515 arcmin2 for 2QZCluster, 497 arcmin2 Multi-tiered Extra-galactic survey, HerMES (Oliver et al. for HS1700, 1076 arcmin2 for SSA22, and 3422 arcmin2 for 2012). The SPIRE map in COSMOS field is larger and COSMOS field). We measured 350 µm and 500 µm fluxes deeper than our observations, so we have reprocessed them at the positions of sources detected in the 250 µm maps, by limiting to the same depth using Nrep = 14. Subse- then listed these 250 µm sources only if the flux density is quently, we apply the same map making, and source de- above 12 mJy in at least one of the SPIRE bands (See also tection procedure in order to make sure that we match the Table 8–10, and the full tables are avaivable in online). This depth of our observations. The confusion and instrumental flux density limit corresponds to ∼ 4σ of the instrumental noise values are given in Table 1. noise and ∼ 2σ of the confusion noise in all three bands. For SSA22, we cut out the region shown in Figure 3 due to the high background fluxes from the Galactic cirrus. 3 ANALYSIS We compared these SPIRE number counts with COS- MOS field. We show the raw (i.e., not corrected for the com- 3.1 DETECTION & NUMBER COUNTS pleteness) number counts in Figure 1 (See also Table 5– 7). The source detection was conducted on the 250 µm maps, The raw number counts are roughly consistent with COS- because of the better spatial resolution compared with MOS field data at > 20 mJy bin at 250 and 350 µm. longer wavelength bands. We cation that 500 µm detection A moderate excess of number counts at S350 > 50 mJy causes a source blending (even 250 µm can deblend these) and S500 > 40 mJy were found (by a factor of 2–3) in and large spatial uncertainty. We used the sussextractor HS1700 and 2QZCluster, although they are within the er- (Smith et al. 2012) for the source detection and photome- ror bars based on Poisson noise. We also compared number try. We detected S/N > 2 sources in the maps within the counts in COSMOS field with wider HerMES survey data MNRAS 000, 1–15 (2016) 4 Y. Kato et al. Table 2. Summary of Herschel /SPIRE sources in the three protoclusters. Field Areaa Nb Nc Nd Ne Σf Σg δh σi (arcmin−2 ) (prior) (catalogued) (selected) (arcmin−2 ) (arcmin−2 ) 2QZCluster 515 643 399 12 6 0.023 ± 0.007 0.132 ± 0.054 3.0 ± 1.7 3.9 ± 2.1 HS1700 497 579 383 26 8 0.052 ± 0.010 0.186 ± 0.066 3.7 ± 1.7 5.0 ± 2.2 SSA22 1076 1253 772 55 5 0.051 ± 0.007 - - - COSMOS (matched to 2QZCluster) 3422 4923 2777 111 7 0.032 ± 0.003 - - - COSMOS (matched to HS1700) 3422 4923 2777 140 7 0.041 ± 0.004 - - - COSMOS (matched to SSA22) 3422 4923 2777 262 10 0.077 ± 0.005 - - - Notes. (a): The area where the integration time is greater than 30%. We detected sources within this area. (b): The numbers of S/N (250 µm) > 2 prior sources. (c): The numbers of catalogued sources which at least one SPIRE band flux is above 12 mJy. (d): The numbers of colour-selected bright SPIRE sources. For SSA22, we excluded the sources in high background fluxes (see Figure 3). (e): The numbers of colour-selected bright SPIRE sources within the overdensities. (f ): The surface density of colour-selected bright SPIRE sources. The errors assume Poisson statistics. (g): The surface density of colour-selected bright SPIRE sources in the overdensities, assume an area of the overdensities (r = 3.8′ for 2QZCluster, r = 3.7′ for HS1700). The errors assume Poisson statistics. (h) − (i): The δ and σ are calculated from δpc = (npc − nave )/nave and σpc = (npc − nave )/σave . The errors assume Poisson statistics. Figure 2. S500 /S350 vs S350 /S250 colour-colour diagram of the 250 µm sources. From left to right, we show the plots for 2QZCluster, HS1700, and SSA22. We selected candidates of DSFGs possibly associated with the protoclusters whose colours are consistent with a single gray body SEDs including a photometric error of ± 20% (shaded regions). We assume the protocluster redshifts, dust temperatures of Td = 30 − 40 K and dust emissivity β = 1.5. The grey solid and dashed lines show tracks of single grey body SEDs for different β. We plot sources with fluxes above 12mJy in least one SPIRE bands as small grey points. The coloured larger/smaller symbols show sources with LFIR larger/smaller than 5.0 × 1012 L⊙ . The error bars show the average errors of colour-selected bright SPIRE sources. We have selected 2%, 5%, and 4% colour-selected bright SPIRE sources to search for overdensities of DSFGs in 2QZCluster, HS1700, and SSA22. We plot the expected colours for Arp 220 as a function of redshifts (Polletta et al. 2007) for comparison. We plot open circles on track of Arp220 every z = 0.5 from z = 0 to z = 5.5. (Béthermin et al. 2012). The number counts in COSMOS (1983), and typically β = 1.0 − 2.0 for DSFGs. Different val- field agree with that of HerMES survey data, suggesting ues of β do not significantly affect our conclusion, yielding that COSMOS field is suitable as a control field. almost the same temperature and SPIRE colours. We in- cluded flux errors of ± 20% in estimating the uncertainties in the SPIRE colours. Figure 2 shows S500 /S350 vs. S350 /S250 3.2 SPIRE COLOUR SELECTION colour-colour diagram of the SPIRE sources in the three pro- toclusters. In order to search for DSFGs possibly associated with the protoclusters, we applied a colour constraint to the SPIRE We derived LFIR (8 − 1000 µm) by fitting single grey detected sources. The S350 /S250 and S500 /S350 colours were body SEDs with Td a free parameter. From the sources se- used to select the sources with single grey body SEDs at lected based on the SPIRE colours, we further applied a the protocluster’s redshift ranges, with dust temperatures of FIR luminosity cut of LFIR ≥ 5.0 × 1012 L⊙ for conser- Td = 30 − 40 K, and a dust emissivity β = 1.5. Casey et al. vative searching for DSFGs. Thus the lowest flux densities (2012) measured the dust temperatures of SPIRE-selected in our samples are (36.0, 30.0, 19.4) mJy for 2QZCluster, DSFGs and showed that they are in the range of Td = 20 − (28.8, 26.3, 15.2) mJy for HS1700, and (15.6, 19.2, 12.9) mJy 60 K. The dust emissivity index is determined by Hildebrand for SSA22 in the SPIRE bands (250, 350, 500) µm, respec- MNRAS 000, 1–15 (2016) Herschel protocluster survey 5 Figure 3. Sky distributions of colour-selected bright 250 µm sources (red circles) over-plotted on the SPIRE 250 µm maps. The areas outlined with magenta contour corresponds to the 30% depth coverage. For SSA22, we do not use below the dashed magenta line due to the high background fluxes. The dashed black contours show the density of 19 HAEs and 3 QSOs (Matsuda et al. 2011) for the 2QZCluster, 45 LBGs for HS1700 (Rudie et al. 2012; Steidel et al. 2014), and 742 LAEs for SSA22 (Yamada et al. 2012) respectively. The steps show 1–4σ, 1–4σ, and 3–6σ for 2QZCluster, HS1700, and SSA22 (smoothed with a Gaussian kernel with a FWHM of 6 comoving Mpc). The red contours show 3–4σ, 4–5σ for colour-selected bright SPIRE sources in 2QZCluster and HS1700 with a same Gaussian kernel (1σ is standard deviation of surface density of colour-selected bright SPIRE sources measured in COSMOS field). The large blue circles show the overdensity of colour-selected bright 250 µm sources for 2QZCluster and HS1700, and colour-selected bright 500 µm sources for SSA22 using filtering a radius of 6 comoving Mpc. We find a 4σ and 5σ overdensity in the 2QZCluster and HS1700 fields. We do not find any significant overdensities of 250 µm sources in SSA22, but we found six colour-selected bright 500 µm sources (black squares), and three sources are concentrated 3′ (∼1.4 Mpc) east to the LAEs overdensity. The dashed cyan and orange boxes in 2QZCluster show the UKIRT/WFCAM (for HAEs) and Subaru/MOIRCS (for HAEs) coverage. The solid cyan, dashed purple, and orange boxes in HS1700 show the Keck/LRIS (for LAEs and LBGs, respectively) and Palomar/WIRC (for HAEs) coverage. The dashed cyan large circle in SSA22 shows ASTE/AzTEC coverage. MNRAS 000, 1–15 (2016) 6 Y. Kato et al. within each aperture. For SSA22, we excluded a high back- ground region (see in Figure 3). We found 3.9σ and 5.0σ overdensities in the 2QZClus- ter and HS1700 fields respectively, but did not find any sig- nificant (> 3σ) overdensities in the SSA22 field. We calcu- lated the significance of overdensity, σpc = (npc −nave )/σave , where npc is the number of colour-selected bright SPIRE sources in overdense region (6 comoving Mpc search radius) of the protocluster fields, nave is the average number within a 6 comoving Mpc search radius for COSMOS field, and σave is the standard deviation of nave . We also searched for overdensities of colour-selected bright SPIRE sources in case for 10% and 30% flux error boundaries in Figure 2. The re- sult does not significantly change for 20% and 30% bound- aries. We found 30% boundary shows same position of over- densities of colour-selected bright SPIRE sources compared with that of 20% boundary in 2QZCluster and HS1700. We found 10% boundary is not suitable because the number of colour-selected bright SPIRE sources are too small to search for overdensities. In Figure 3, we show the sky distribution of the colour-selected bright SPIRE sources. We also show the density of 19 HAEs and 3 QSOs (Matsuda et al. 2011) for 2QZCluster, 45 LBGs for HS1700 (Rudie et al. 2012; Steidel et al. 2014), and 742 LAEs for SSA22 (Yamada et al. 2012) respectively. Figure 4. (Upper ). The density distributions of the colour- We searched for overdensities of the colour-selected selected bright SPIRE sources in 2QZCluster (left) and HS1700 bright SPIRE sources in the COSMOS field in an identical (right) within an aperture radius of 6 comoving Mpc. 2QZClus- manner as for the protocluster fields. We plot the density dis- ter tends to have lower number of colour-selected bright SPIRE tribution of the colour-selected bright SPIRE sources for the sources compared with COSMOS field. This could be explained 2QZCluster, HS1700 and COSMOS fields in Figure 4. The if there are void-like structures within the survey area although histogram shows the distribution of the number of colour- there are also overdense regions. In contrast, HS1700 tends to selected bright SPIRE sources within a 6 comoving Mpc have higher number of sources. (Lower ). Cumulative number of the apertures with a radius of 6 comoving Mpc. We normalized aperture (normalized by the number of searched apertures). the cumulative number of COSMOS field to the protoclusters. We found that 2QZCluster hasa low number of sources. The fraction of the apertures which have no sources is about two times higher compared with COSMOS field. This suggests tively. We finally obtained a sample of colour-selected bright that there is a void like distribution while there are also SPIRE sources with colours consistent with the protoclus- overdensities in 2QZCluster. In contrast, HS1700 tends to ter redshifts by rejecting ∼ 95% of the 250 µm sources have larger number of colour-selected bright SPIRE sources (thus we select just 12/643 (2%), 26/579 (5%), and 55/1253 in the search apertures. (4%) colour-selected bright SPIRE sources in 2QZCluster, We also plot the cumulative number of the cricles in Fig- HS1700 and SSA22). For COSMOS field, we applied the ure 4. We normalized the cumulative number of COSMOS same colour selection as the three protoclusters to obtain a field to the protoclusters. We found that the normalized cu- control sample of field galaxies. We have selected 111/4923 mulative number with > 3.9σ overdensities in 2QZCluster (2%), 140/4923 (3%), and 262/4923 (5%) colour-selected field is about two times higher than that in the COSMOS bright SPIRE sources in COSMOS field for 2QZCluster, field. There are no overdensities in the COSMOS field which HS1700, and SSA22 (see also Table 2). contain eight colour-selected bright SPIRE sources (5.0σ) as HS1700. This means that such overdensities are preferen- tially located in the protoclusters. 4 RESULTS 4.2 2QZCLUSTER 4.1 SEARCH FOR OVERDENSITIES In the 2QZCluster field, an overdensity of colour-selected We searched for overdensities in an aperture with a radius of bright SPIRE sources were found 4.5′ (∼2.2 Mpc) west to 6 comoving Mpc (3.8′ , 3.7′ , and 3.2′ radius for z = 2.2, 2.3, the HAEs overdensity. and 3.1, respectively). This scale corresponds to physical We searched for the counterparts of SPIRE sources with scale of ∼ 1.5 − 2 Mpc radius for each protocluster redshift, HAEs and QSOs in 2QZCluster, which are summarized by matching the size of the overdensity of DSFGs around radio Matsuda et al. (2011), although the SPIRE overdense re- galaxies (e.g., Rigby et al. 2014; Dannerbauer et al. 2014). gion did not have Hα image with UKIRT/WFCAM (see We put down a grid of these apertures every 10′′ for the pro- Figure 3). In total, 19 HAEs and 3 QSOs are within the toclusters and COSMOS field, and counted the number of 174 arcmin2 overlap region. We define the following qual- colour-selected bright SPIRE sources (LFIR ≥ 5.0×1012 L⊙ ) ity criteria based on Downes et al. (1986) for assessing the MNRAS 000, 1–15 (2016) Herschel protocluster survey 7 robustness of identified candidate counterparts. We classify sify sources with p-value described above. We deduced the sources with p ≤ 0.05 as secure counterparts, and those with number density of 45 LBGs which have spectroscopically 0.05 < p ≤ 0.10 as tentative counterparts. We calculated the confirmed redshift in the protocluster (i.e., in the overden- p-value defined by p = 1 − exp(−πnθ2 ) where n is the source sity of z = 2.285 − 2.315) separately from remaining 2965 density, θ is the angular offset. We searched for counterparts LBGs outside of overdensity. We searched for counterparts within 11′′ radius from the centre of the SPIRE sources. This within 11′′ radius from the centre of the SPIRE sources. search radius corresponds to ∼40% beam response in 250 µm We found three colour-selected bright SPIRE sources band. that have secure counterparts (see Table 12). But one of For the QSOs, 1 optically luminous QSO, which is the LAEs BNB1, the spectrum of this object is extremely known as 2QZC-C1-HAE2 (Matsuda et al. 2011), coin- odd and that makes difficult to identify the redshift. We cides with one of our colour-selected bright SPIRE source identified it as a âĂIJLo-BALâĂİ QSO and it is more likely 2QZCluster-SPIRE10 (2.7′′ offset). The p-value is lower than to have z ∼ 2.00, so we do not treat as it is a protocluster 0.01, suggesting a secure counterpart, the probability of member galaxy. chance association of counterparts is lower than 1%. 2QZ- We also found one LAE and LBG are secure, and one C1-HAE3 also coincides with our SPIRE source 2QZCluster- DRG and LAE are tentative counterparts of colour-selected SPIRE124 (5.6′′ offset) as a secure counterpart. Such op- faint (i.e., LFIR < 5.0 × 1012 L⊙ ) SPIRE source. For the tically luminous FIR bright QSOs are thought to be in a SPIRE sources which are not selected with our colour selec- transient phase between DSFGs and QSOs (Simpson et al. tion, but have at least one SPIRE band above 12 mJy, we 2012). No other HAEs matched with colour-selected bright found two secure LAEs, three secure HAEs, and one ten- SPIRE sources. We then repeated this matching using all tative HAE. Thus, we conclude that seven SPIRE sources 250 µm sources, and found five secure HAE counterparts in the overdensity have secure counterparts of protocluster (see Table 11). galaxies (see Table 12). We recently conducted Subaru/MOIRCS follow-up NB In order to assess the success rate of the SPIRE colour imaging observations in May 5, 2015 to search for new HAEs selection, we compared the matching result for colour- as well as counterparts of the SPIRE sources in the SPIRE selected bright/faint SPIRE sources and all 250 µm sources overdense region (4′ × 7′ ; dashed orange box in Figure 3). lying within overdensity. We assumed that all HAEs and The total exposure time of Ks and NB data is each 1.5 and LAEs are associated with the protocluster except for BNB1, 2.4 ksec, respectively. The 5σ detection limit of NB < 19.8 although we should note that the HAEs without spectro- (Vega) is almost the same as the previous HAE search depth scopic follow-up are much less likely than spectroscopic (NB < 19.9 in Vega) by Matsuda et al. (2011). confirmed HAEs to be at the cluster redshift. For all se- We find no new HAE within the area. This may sug- cure counterparts, the fractions are 2/7(29%), 1/5(20%), gest that the area is actually a little under-dense in terms and 4/28(14%), respectively. This suggests that our SPIRE of bright HAE, because Matsuda et al. (2011) claim an av- colour selection can select possible protocluster members erage field count of 0.09 per arcmin2 , which corresponds to with three times higher probability compared to not colour- two or three HAEs per MOIRCS FOV. Such void structures selected sources. adjacent to protocluster has been reported previously (e,g., In the HS1700 field, there are several foreground galaxy Koyama et al. 2013; Saito et al. 2015). groups in the field. A z = 0.453 group is very close to the po- sition of HS1700-SPIRE24. Indeed, Peter et al. (2007) found that BX913 is lensed by the group. We calculated a magnifi- 4.3 HS1700 cation factor by using glafic (Oguri 2010). We used a halo In HS1700, the overdensity peak of colour-selected bright mass of M/h = 1 × 1014 M⊙ (Israel et al. 2014) and a con- SPIRE sources coincides with that of the protocluster mem- centration parameter of c=6 (Bhattacharya et al. 2013). We ber galaxies (2.1′ offset, corresponding to ∼ 1.0 Mpc). found that the second nearest (1.5′ away from group center) We searched for the counterparts of SPIRE sources us- SPIRE source is affected only ∼5% magnification, so we con- ing UV-selected (BX/BM) star-forming galaxies (LBGs), clude that only the HS1700-SPIRE24 could be affected by HAEs, LAEs, and DRGs in the SPIRE overdense re- lensing. gion. There are 3010 LBGs within 256 arcmin2 , 123 LAEs are within 219 arcmin2 , 83 HAEs and 75 DRGs are 4.4 SSA22 within 72 arcmin2 . The LBG catalogues are from the KBSS (Keck Baryonic Structure Survey; Rudie et al. 2012; The maximum number of colour-selected bright SPIRE Steidel et al. 2014). HS1700 is one of the fifteen QSO sources found in on 6 comoving Mpc aperture was five, cor- fields intensively studied in KBSS. The Hα and Lyα NB responding to a 1.6σ overdensity. This suggests that SSA22 catalogues are from Milan BogosavljevićâĂŹs Ph.D thesis does not have any significant overdensities of colour-selected (Bogosavljević 2010). Individual properties of LBGs, LAEs bright 250 µm sources. But, this could be a problem for and observations are described in Shapley et al. (2005), SSA22 at z = 3.1 because DSFGs would start to drop out Erb, Bogosavljević, & Steidel (2011), Kulas et al. (2013), at 250 µm. Here we searched for an overdensity of colour- and Erb et al. (2014). In addition, for the Hα narrow band selected bright 500 µm detected sources with identical colour imaging, we used the Brγ filter (Palomar WIRC, center selection and luminosity cut in §3.2. We found six colour- wavelength 2.17 µm, FWHM=297 Å). The narrow band se- selected bright 500 µm sources in the SSA22 field, and three lection criteria are NB < 20.5 (Vega) and NB −Ks ≤ −0.75. sources are concentrated 3′ (∼1.4 Mpc) east to the LAEs The DRGs have selected with J − K > 2.3 to a limit of overdensity (Figure 3). For 350 µm sources, we did not find Ks = 21 (Vega). We define the quality criteria, and clas- any significant overdensities as same as 250 µm sources. MNRAS 000, 1–15 (2016) 8 Y. Kato et al. We investigated AzTEC 1.1 mm counterparts in Table 3. Estimated SFR and assumed volume for Figure 5. Umehata et al. (2014, 2015) for the five colour-selected bright 500 µm sources within ∼800 arcmin2 overlapped re- gion (Figure 3). We searched for the counterparts within Field z SFRa rb Vc SFRmin d (M⊙ yr−1 ) (Mpc) (Mpc3 ) (M⊙ yr−1 ) a radius of 14′′ (half of AzTEC 1.1 mm FWHM). We found that four colour-selected bright 500 µm sources have All sources been matched (SSA22-AzTEC1, 2, 5, and 34). Thus ∼ 80% 2QZCluster 2.230 ± 0.016 5000 1.0 4.2 150 colour-selected bright 500 µm sources are matched with HS1700 2.300 ± 0.015 4500 1.0 4.2 280 SSA22 3.09 ± 0.03 7500 1.0 4.2 310 AzTEC 1.1 mm sources. Field corrected values 2QZCluster 2.230 ± 0.016 1000 1.0 4.2 - HS1700 2.300 ± 0.015 2000 1.0 4.2 - 5 DISCUSSION & SUMMARY SSA22 3.09 ± 0.03 4700 1.0 4.2 - We searched for DSFGs associated with three protoclus- Notes. (a): Integrated SFR derived from SPIRE sources which ters at z = 2 − 3 (2QZCluster, HS1700, SSA22) using Her- are included in column 4 radius. (b) − (c): Used radius and schel /SPIRE. In the 2QZCluster and HS1700 field, we found volume to calculate SFR density. (d): Minimum SFR of SPIRE 4σ and 5σ overdensities of the colour-selected bright SPIRE source which is included in integrated SFR. sources on a scale of 6 comoving Mpc. In the SSA22, we did not find any significant overdensities of 250 µm sources, but we found that three colour-selected bright 500 µm sources (2003)’s LFIR correction to follow Clements et al. (2014) are concentrated on a scale of 6 comoving Mpc. The results analysis. The derived SFR densities are 103 –104 times higher suggest possible activity associated with enhanced dusty than the global SFR density seen in Madau & Dickinson star-formation in protoclusters at z ∼ 2 − 3. (2014). This comparison indicates enhanced star-formation We derive the star-formation rate (SFR) density of 2QZ- activities in protoclusters at z ∼ 2 − 3. The enhance- Cluster, HS1700 and SSA22 to compare with the average ment of dusty star-formation activity within our proto- value of the Universe. The SFR of DSFGs is often converted clusters is consistent with that seen for Planck clumps from far-infrared luminosity, although the conversion from (Clements et al. 2014). Dannerbauer et al. (2014) presents LFIR to SFR is not straightforward and relies on the dust results based on APEX/LABOCA 870 µm observations composition and initial mass function (IMF). Most works on around MRC1138−262 at z = 2.16, showing that at least DSFGs assume the conversion given by SFR (M⊙ yr−1 ) = six DSFGs are likely part of the protocluster. The SFR den- 4.5 × 10−44 LFIR (erg s−1 ) or 1.7 × 10−10 LFIR (L⊙ ) sity ∼ 1500 M⊙ yr−1 Mpc−3 is similar to Clements et al. (Kennicutt 1998), where LFIR is the integrated luminosity (2014) and our protoclusters study. We also plot the error of 8 − 1000 µm. This conversion takes the radiative transfer bars in Figure 5. The high end shows the case when includ- models of Leitherer & Heckman (1995) and a Salpeter IMF ing all sources in 1 Mpc radius and the low end shows the (Salpeter 1955). case for subtracting field average values. We deduced field We used a simple estimation to calculate the integrated values by calculating average number of sources and average SFR densities by assuming that all colour-selected bright luminosity in 1 Mpc radius apertures in the COSMOS field. SPIRE sources in the overdensities are associated with pro- In Table 3, we summarized the estimated SFR and as- toclusters. In addition, we also included sources with S/N > sumed volume plotted in Figure 5. We also caluculated SFR 2 in all three SPIRE bands within a radius of 1 Mpc (phys- density in an aperture using a radius of 6 comoving Mpc in ical) following Clements et al. (2014) analysis of Planck the same manner. These radii are 1.9, 1.8 and 1.5 Mpc in clumps with Herschel /SPIRE. Because this radius of 1 Mpc physical units for 2QZCluster, HS1700 and SSA22, respec- circles are smaller than our search radius (6 comoving Mpc, tively, and the results do not change compared to the case ∼ 1.5 − 2 Mpc in physical), we adjusted the position of the of 1 Mpc. smaller aperture to contain as many colour-selected bright What causes the enhanced dusty star-forming activity SPIRE sources as possible. This enables us to select regions in our protoclusters? One possible answer to this is galaxies similar to Planck clumps in Clements et al. (2014). The 1σ mergers. For instance, Casey et al. (2015) investigated the instrumental noise of Clements’s work ranges from 2.5 − 2.8 morphology of the galaxies in a protocluster to search for in- mJy at 250 µm, 2.1−2.3 mJy at 350 µm and 3.0−3.3 mJy at teraction/merger state by using HST H-band imaging data. 500 µm (Clements et al. 2014), which is similar to our sur- Although the sample size is limited, they found that the frac- vey. For all SPIRE sources within a 1 Mpc radius, we fitted tion of irregular and interacting galaxies among the LBGs single grey body SEDs with fixed Td = 35 K, β = 1.5 and and DSFGs is 1.5 times higher in the protocluster than in the protocluster’s redshifts, and derived LFIR . We excluded the field. Webb et al. (2015) suggested that dusty star for- one colour-selected bright SPIRE source HS1700-SPIRE30 mation at the center of a z = 1.7 cluster is being driven from this discussion because it is very close to z = 0.08 SDSS by galaxy-galaxy interaction, involving a far-infrared lumi- galaxy, and its Spitzer /MIPS 24 µm flux density is very high nous Bright Cluster Galaxy (BCG). Because there are strong (∼ 200 µJy). We found that Mips 24 µm flux densities of correlation between the major-merger and the far-infrared other colour-selected bright SPIRE sources in HS1700 are luminosity (Kartaltepe et al. 2010, 2012; Engel et al. 2010; consistent with that of typical high-z star-forming galaxies. Hung et al. 2013), higher major mergers rate in protoclus- 2QZCluster does not have any Spitzer images. ters could induce dusty star-formation. N-body simulation Figure 5 shows our results and compare to previous also predicts that progenitors of cluster and group haloes at studies in the literature. We calculated SFR with Bell z > 2 have 3 − 5× higher major merger rates than isolated MNRAS 000, 1–15 (2016) Herschel protocluster survey 9 Figure 5. SFR density of clusters, protoclusters and global cosmic SFR densities. The SFR densities of our protoclusters are 103 –104 times higher than the global SFR density (Madau & Dickinson 2014). We show the compilation from Dannerbauer et al. (2014) and data from Clements et al. (2014), which original literature are following; IRAS measurements of Perseus from Meusinger, Brunzendorf, & Krieg (2000), BLAST measurements of A3112 from Braglia et al. (2011), ISO measurements of A1689 from Fadda et al. (2000), and Spitzer measurements of A1758 from Haines et al. (2009), Bullet cluster from Chung et al. (2010), Cl0024+16 and MS0451-03 from Geach et al. (2006). RXJ0057, RXJ0941, RXJ1218, RXJ1249 and RXJ1633 are based on JCMT/SCUBA from Stevens et al. (2010). halos (Gottlöber, Klypin, & Kravtsov 2001). The facts that port of the W.M. Keck Foundation. The authors wish to the colour-selected bright SPIRE sources are significantly recognize and acknowledge the very significant cultural role concentrated in HS1700 and 2QZCluster would support such and reverence that the summit of Mauna Kea has always had simultaneously major merging phenomenon. However, fur- within the indigenous Hawaiian community. We are most ther observations are needed to investigate the processes of fortunate to have the opportunity to conduct observations dusty star-forming activity in these protoclusters. from this mountain. This research was supported in part by a grant from the Hayakawa Satio Fund awarded by the Astronomical So- ciety of Japan. YM acknowledges support from JSPS KAK- ACKNOWLEDGMENTS ENHI Grant Number 20647268. IRS acknowledges support We thank the anonymous referee for helpful comments which from STFC (ST/L00075X/1), the ERC Advanced Grant significantly improved the clarity of this paper. We thank DUSTYGAL (321334) and a Royal Society/Wolfson Merit Scott Chapman, James Colbert, Emanuele Daddi, Koichiro Award. BH acknowledges support from JSPS KAKENHI Nakanishi and Kazuhiro Shimasaku for useful discussions Grant Number 15K17616. HU acknowledges support from and supports. Rhythm Shimakawa and Mariko Kubo made Grant-in-Aid for JSPS Fellows, 26.11481. enormous contribution to analyses. We acknowledge Masaru Kajisawa for our use of his MOIRCS fringe-removal software (Kajisawa et al. 2015) during the data reduction. REFERENCES Herschel is an ESA space observatory with science in- Bai L., et al., 2007, ApJ, 664, 181 struments provided by European-led Principal Investigator Bell E. F., 2003, ApJ, 586, 794 consortia and with important participation from NASA. Béthermin M., et al., 2012, A&A, 542, AA58 SPIRE has been developed by a consortium of institutes led Bhattacharya S., Habib S., Heitmann K., Vikhlinin A., 2013, ApJ, by Cardiff University (UK). Subaru Telescope, which is op- 766, 32 erated by the National Astronomical Observatory of Japan. Braglia F. G., et al., 2011, MNRAS, 412, 1187 UKIRT is funded by the STFC (UK). The W.M. Keck Ob- Brodwin M., et al., 2013, ApJ, 779, 138 servatory was made possible by the generous financial sup- Bogosavljević M., 2010, PhDT, MNRAS 000, 1–15 (2016) 10 Y. Kato et al. Casey C. M., et al., 2012, ApJ, 761, 140 Saito T., et al., 2015, MNRAS, 447, 3069 Casey C. M., et al., 2015, ApJ, 808, L33 Salpeter E. E., 1955, ApJ, 121, 161 Casey C. M., 2016, arXiv, arXiv:1603.04437 Shapley A. E., Steidel C. C., Erb D. K., Reddy N. A., Adelberger Chapman S. C., et al., 2015, MNRAS, 449, L68 K. L., Pettini M., Barmby P., Huang J., 2005, ApJ, 626, 698 Chung S. M., Gonzalez A. H., Clowe D., Markevitch M., Zaritsky Simpson J. M., et al., 2012, MNRAS, 426, 3201 D., 2010, ApJ, 725, 1536 Smail I., Ivison R. J., Blain A. W., Kneib J.-P., 1998, ApJ, 507, Clements D. L., et al., 2014, MNRAS, 439, 1193 L21 Croom S. M., Smith R. J., Boyle B. J., Shanks T., Loaring N. S., Smail I., et al., 2014, ApJ, 782, 19 Miller L., Lewis I. J., 2001, MNRAS, 322, L29 Smith A. J., et al., 2012, MNRAS, 419, 377 Croom S. M., Smith R. J., Boyle B. J., Shanks T., Miller L., Steidel C. C., Adelberger K. L., Dickinson M., Giavalisco M., Outram P. J., Loaring N. S., 2004, MNRAS, 349, 1397 Pettini M., Kellogg M., 1998, ApJ, 492, 428 Dannerbauer H., et al., 2014, A&A, 570, AA55 Steidel C. C., Adelberger K. L., Shapley A. E., Pettini M., Dick- Digby-North J. A., et al., 2010, MNRAS, 407, 846 inson M., Giavalisco M., 2000, ApJ, 532, 170 Downes A. J. B., Peacock J. A., Savage A., Carrie D. R., 1986, Steidel C. C., Adelberger K. L., Shapley A. E., Erb D. K., Reddy MNRAS, 218, 31 N. A., Pettini M., 2005, ApJ, 626, 44 Elbaz D., et al., 2007, A&A, 468, 33 Steidel C. C., et al., 2014, ApJ, 795, 165 Ellis R. S., Smail I., Dressler A., Couch W. J., Oemler A., Jr., Stevens J. A., et al., 2003, Natur, 425, 264 Butcher H., Sharples R. M., 1997, ApJ, 483, 582 Stevens J. A., Jarvis M. J., Coppin K. E. K., Page M. J., Greve Engel H., et al., 2010, ApJ, 724, 233 T. R., Carrera F. J., Ivison R. J., 2010, MNRAS, 405, 2623 Erb D. K., Bogosavljević M., Steidel C. C., 2011, ApJ, 740, L31 Swinyard B. M., et al., 2010, A&A, 518, L4 Erb D. K., et al., 2014, ApJ, 795, 33 Tamura Y., et al., 2009, Natur, 459, 61 Fadda D., Elbaz D., Duc P.-A., Flores H., Franceschini A., Ce- Tran K.-V. H., et al., 2010, ApJ, 719, L126 sarsky C. J., Moorwood A. F. M., 2000, A&A, 361, 827 Umehata H., et al., 2014, MNRAS, 440, 3462 Geach J. E., et al., 2006, ApJ, 649, 661 Umehata H., et al., 2015, ApJ, 815, L8 Gottlöber S., Klypin A., Kravtsov A. V., 2001, ApJ, 546, 223 Valtchanov I., et al., 2013, MNRAS, 436, 2505 Granato G. L., Ragone-Figueroa C., Domı́nguez-Tenreiro R., Webb T., et al., 2015, ApJ, 809, 173 Obreja A., Borgani S., De Lucia G., Murante G., 2015, MN- Wylezalek D., et al., 2013, MNRAS, 428, 3206 RAS, 450, 1320 Yamada T., Nakamura Y., Matsuda Y., Hayashino T., Yamauchi Griffin M. J., et al., 2010, A&A, 518, L3 R., Morimoto N., Kousai K., Umemura M., 2012, AJ, 143, 79 Haines C. P., Smith G. P., Egami E., Okabe N., Takada M., Ellis R. S., Moran S. M., Umetsu K., 2009, MNRAS, 396, 1297 Hayashino T., et al., 2004, AJ, 128, 2073 Hildebrand R. H., 1983, QJRAS, 24, 267 Hung C.-L., et al., 2013, ApJ, 778, 129 Israel H., Reiprich T. H., Erben T., Massey R. J., Sarazin C. L., Schneider P., Vikhlinin A., 2014, A&A, 564, A129 Ivison R. J., Dunlop J. S., Smail I., Dey A., Liu M. C., Graham J. R., 2000, ApJ, 542, 27 Ivison R. J., et al., 2013, ApJ, 772, 137 Kajisawa M., Morishita T., Taniguchi Y., Kobayashi M. A. R., Ichikawa T., Fukui Y., 2015, ApJ, 801, 134 Kartaltepe J. S., et al., 2010, ApJ, 721, 98 Kartaltepe J. S., et al., 2012, ApJ, 757, 23 Kennicutt R. C., Jr., 1998, ARA&A, 36, 189 Koyama Y., Kodama T., Tadaki K.-i., Hayashi M., Tanaka M., Smail I., Tanaka I., Kurk J., 2013, MNRAS, 428, 1551 Kulas K. R., et al., 2013, ApJ, 774, 130 Lehmer B. D., et al., 2009, ApJ, 691, 687 Lehmer B. D., et al., 2009, MNRAS, 400, 299 Lehmer B. D., et al., 2013, ApJ, 765, 87 Leitherer C., Heckman T. M., 1995, ApJS, 96, 9 Lilly S., et al., 1998, ApJ, 500, 75 Lutz D., et al., 2001, A&A, 378, 70 Ma C.-J., et al., 2015, ApJ, 806, 257 Madau P., Dickinson M., 2014, ARA&A, 52, 415 Matsuda Y., et al., 2004, AJ, 128, 569 Matsuda Y., et al., 2011, MNRAS, 416, 2041 Meusinger H., Brunzendorf J., Krieg R., 2000, A&A, 363, 933 Nguyen H. T., et al., 2010, A&A, 518, LL5 Oguri M., 2010, PASJ, 62, 1017 Oliver S. J., et al., 2012, MNRAS, 424, 1614 Ott S., 2010, ASPC, 434, 139 Peter A. H. G., Shapley A. E., Law D. R., Steidel C. C., Erb D. K., Reddy N. A., Pettini M., 2007, ApJ, 668, 23 Pilbratt G. L., et al., 2010, A&A, 518, L1 Polletta M., et al., 2007, ApJ, 663, 81 Rigby E. E., et al., 2014, MNRAS, 437, 1882 Rudie G. C., et al., 2012, ApJ, 750, 67 MNRAS 000, 1–15 (2016) Herschel protocluster survey 11 Table 4. Number of sources which have fluxes above 12 mJy in each bands. COSMOS SSA22 HS1700 2QZCluster 250 µm 2552 699 356 365 350 µm 1944 529 259 260 500 µm 961 252 133 116 Table 5. Number counts at 250 µm. The errors take into account the statistical uncertainties, including the Poisson noise. Central flux Flux bin Number counts (dN/dS) [mJy−1 deg−2 ] (mJy) (mJy) SSA22 HS1700 2QZCluster COSMOS 16.8 12.0 – 21.6 148.1 ± 7.2 137.2 ± 10.2 157.1 ± 10.7 158.1 ± 4.2 23.8 21.6 – 26.0 70.0 ± 7.3 92.1 ± 12.3 73.0 ± 10.8 87.5 ± 4.6 33.6 26.0 – 41.2 31.5 ± 2.6 37.6 ± 4.2 34.0 ± 4.0 36.4 ± 1.6 47.4 41.2 – 53.6 6.5 ± 1.3 11.7 ± 2.6 7.9 ± 2.1 10.4 ± 0.9 67.0 53.6 – 80.4 1.6 ± 0.5 4.0 ± 1.0 2.3 ± 0.8 2.7 ± 0.3 94.6 80.4 – 108.8 0.12 ± 0.12 0.51 ± 0.36 0.74 ± 0.43 0.74 ± 0.17 133.7 108.8 – 158.6 0.07 ± 0.07 0.15 ± 0.15 0.42 ± 0.24 0.06 ± 0.04 188.8 158.6 – 219.0 - - - - Table 6. Number counts at 350 µm. The errors take into account the statistical uncertainties, including the Poisson noise. Central flux Flux bin Number counts (dN/dS) [mJy−1 deg−2 ] (mJy) (mJy) SSA22 HS1700 2QZCluster COSMOS 16.8 12.0 – 21.6 113.6 ± 6.3 107.0 ± 9.0 116.4 ± 9.2 127.4 ± 3.7 23.8 21.6 – 26.0 60.1 ± 6.8 69.1 ± 10.7 68.3 ± 10.4 62.2 ± 3.9 33.6 26.0 – 41.2 23.1 ± 2.3 23.8 ± 3.4 18.8 ± 2.9 28.5 ± 1.4 47.4 41.2 – 53.6 3.2 ± 0.9 8.8 ± 2.3 3.4 ± 1.4 5.9 ± 0.7 67.0 53.6 – 80.4 0.9 ± 0.3 2.4 ± 0.8 2.3 ± 0.8 1.5 ± 0.2 94.6 80.4 – 108.8 - - - 0.04 ± 0.04 133.7 108.8 – 158.6 - - 0.14 ± 0.14 - 188.8 158.6 – 219.0 - - - - Table 7. Number counts at 500 µm. The errors take into account the statistical uncertainties, including the Poisson noise. Central flux Flux bin Number counts (dN/dS) [mJy−1 deg−2 ] (mJy) (mJy) SSA22 HS1700 2QZCluster COSMOS 16.8 12.0 – 21.6 68.3 ± 4.9 70.8 ± 7.3 53.1 ± 6.2 75.2 ± 2.9 23.8 21.6 – 26.0 23.6 ± 4.2 21.4 ± 5.9 39.7 ± 7.9 27.3 ± 2.6 33.6 26.0 – 41.2 3.7 ± 0.9 10.0 ± 2.2 6.4 ± 1.7 9.7 ± 0.8 47.4 41.2 – 53.6 1.3 ± 0.6 2.3 ± 1.2 1.7 ± 1.0 1.6 ± 0.4 67.0 53.6 – 80.4 0.1 ± 0.1 0.27 ± 0.27 0.26 ± 0.26 0.08 ± 0.06 94.6 80.4 – 108.8 - - - - 133.7 108.8 – 158.6 - - - - 188.8 158.6 – 219.0 - - - - MNRAS 000, 1–15 (2016) 12 Y. Kato et al. Table 8. SPIRE sources catalogue of 2QZCluster. The full table is available online. ID R.A. Dec. S250 S350 S500 (J2000) (J2000) (mJy) (mJy) (mJy) 2QZCluster-SPIRE1 150.81846968368112 0.029424092103377656 127.0 ± 2.1 54.2 ± 1.9 24.6 ± 2.2 2QZCluster-SPIRE2 150.76480950482065 0.22501254698200338 139.1 ± 2.4 117.9 ± 1.8 71.4 ± 2.3 2QZCluster-SPIRE3 150.91517959664077 0.3909833462929028 111.4 ± 2.4 55.2 ± 1.8 21.2 ± 2.4 2QZCluster-SPIRE4 151.02463506680982 0.1206541823404277 97.4 ± 2.2 54.2 ± 1.8 26.5 ± 2.4 2QZCluster-SPIRE5 151.04455762206453 0.4047339937359634 93.6 ± 2.3 61.3 ± 1.8 24.2 ± 2.2 2QZCluster-SPIRE6 151.07930718745527 0.12006911913976963 84.1 ± 2.3 58.9 ± 1.8 31.3 ± 2.3 2QZCluster-SPIRE7 151.11225122198036 0.17674219593981788 66.1 ± 2.2 36.2 ± 2.1 13.2 ± 2.1 2QZCluster-SPIRE8 150.7066091154151 0.2830585502030584 67.4 ± 2.3 54.5 ± 1.8 18.4 ± 2.2 2QZCluster-SPIRE9 150.94968775985714 0.33116978814976 64.4 ± 2.2 22.2 ± 1.9 0.0 ± 2.1 2QZCluster-SPIRE10 150.91631977926835 0.3524558479517708 60.1 ± 2.1 64.7 ± 1.8 45.6 ± 2.0 Table 9. SPIRE sources catalogue of HS1700. The full table is available online. ID R.A. Dec. S250 S350 S500 (J2000) (J2000) (mJy) (mJy) (mJy) HS1700-SPIRE1 255.53646683056928 64.2060083253161 422.6 ± 3.1 181.4 ± 1.9 67.3 ± 2.4 HS1700-SPIRE2 255.07094202680116 64.36337246187729 112.4 ± 2.2 49.0 ± 1.9 19.5 ± 2.2 HS1700-SPIRE3 255.10511462429938 64.14686204692669 86.2 ± 2.4 51.5 ± 2.0 21.4 ± 2.1 HS1700-SPIRE4 255.15806474143926 64.22916135482777 85.8 ± 2.5 60.5 ± 1.8 27.9 ± 2.2 HS1700-SPIRE5 254.99437853070503 64.26410380634614 70.8 ± 2.1 76.4 ± 1.7 51.2 ± 2.2 HS1700-SPIRE6 254.91863041158092 64.21174308418698 76.7 ± 2.4 44.4 ± 1.7 17.6 ± 2.3 HS1700-SPIRE7 255.22359843297363 64.38646521955137 65.7 ± 2.2 21.3 ± 1.9 13.0 ± 2.3 HS1700-SPIRE8 254.82938021291324 64.17873107989598 68.7 ± 2.3 59.4 ± 1.7 27.6 ± 2.2 HS1700-SPIRE9 255.20168755944053 64.25859055126669 66.3 ± 2.2 35.3 ± 2.1 16.2 ± 2.2 HS1700-SPIRE10 255.1823543841818 64.03153372522755 64.7 ± 2.2 64.8 ± 1.9 31.2 ± 2.1 Table 10. SPIRE sources catalogue of SSA22. The full table is available online. ID R.A. Dec S250 S350 S500 (J2000) (J2000) (mJy) (mJy) (mJy) SSA22-SPIRE1 334.264444128561 0.6764415586203842 135.3Âś2.2 71.3Âś1.9 30.4Âś2.3 SSA22-SPIRE2 334.57160684240256 0.49219161422544583 93.3Âś2.2 76.7Âś1.7 48.7Âś2.3 SSA22-SPIRE3 334.344178441303 0.3528652232847903 79.2Âś2.2 34.4Âś1.9 10.4Âś2.2 SSA22-SPIRE4 334.34457572927244 0.607307593915126 79.7Âś2.4 56.9Âś1.7 21.9Âś2.0 SSA22-SPIRE5 334.69745115684617 0.3983836901706686 72.9Âś2.2 29.8Âś1.9 12.8Âś2.1 SSA22-SPIRE6 334.392819716858 0.3042332652494624 68.9Âś2.1 33.2Âś1.8 0.0Âś2.1 SSA22-SPIRE7 334.23836274598824 0.40194158606064945 61.8Âś2.1 45.8Âś1.8 8.8Âś2.1 SSA22-SPIRE8 334.55462830631444 0.47701863826781765 62.9Âś2.1 33.3Âś1.8 18.3Âś2.0 SSA22-SPIRE9 334.3848999033597 0.2912395866246295 57.5Âś2.1 49.6Âś1.7 45.7Âś2.2 SSA22-SPIRE10 334.58002216598203 0.35006798763473185 64.3Âś2.3 51.3Âś2.0 30.6Âś2.3 MNRAS 000, 1–15 (2016) MNRAS 000, 1–15 (2016) Table 11. Matching results for the SPIRE sources which have HAEs counterparts in 2QZCluster. ID R.A. Dec S250 S350 S500 Counterpart R.A. Dec Sepa.a p-valueb Memberc (SPIRE) (J2000) (J2000) (mJy) (mJy) (mJy) (J2000) (J2000) (arcsec) Colour-selected bright SPIRE sources 2QZCluster-SPIRE10 150.916320 0.352456 60.1 ± 2.1 64.7 ± 1.8 45.6 ± 2.0 QZC-C1-HAE02(QSO) 150.915790 0.353000 2.73 <0.01 Secure at least one SPIRE band above 12 mJy sources 2QZCluster-SPIRE124 150.965998 0.249463 21.9 ± 2.2 14.2 ± 1.8 9.7 ± 2.2 QZC-C1-HAE03(QSO) 150.964920 0.250583 5.60 <0.01 Secure 2QZCluster-SPIRE239 151.069470 0.310488 15.4 ± 2.2 4.1 ± 1.8 6.9 ± 2.3 QZC-C1-HAE20 151.069750 0.309167 4.86 <0.01 Secure 2QZCluster-SPIRE251 150.885551 0.168325 15.5 ± 2.2 6.8 ± 2.0 3.7 ± 2.3 QZC-C1-HAE09 150.887370 0.167917 6.71 <0.01 Secure 2QZCluster-SPIRE261 150.852526 0.155903 15.8 ± 2.3 13.8 ± 1.8 12.1 ± 2.2 QZC-C1-HAE05 150.854380 0.155694 6.72 <0.01 Secure 2QZCluster-SPIRE353 150.916183 0.187292 14.4 ± 2.7 12.1 ± 1.8 3.0 ± 2.1 QZC-C1-HAE16 150.915750 0.187722 2.20 <0.01 Secure 2QZCluster-SPIRE372 150.935267 0.238716 10.2 ± 2.2 16.2 ± 1.8 14.0 ± 2.1 QZC-C1-HAE15 150.932670 0.238944 9.39 0.01 Secure Notes. (a): Separation between SPIRE source and counterparts. (b): p-value based on Downes et al. (1986). (c): We classify sources with p ≤ 0.05 as secure, and those with 0.05 < p ≤ 0.10 as tentative counterparts. Herschel protocluster survey 13 14 Table 12. Matching results for the SPIRE sources which have emitter counterparts within overdensity in HS1700. ID R.A. Dec S250 S350 S500 ID R.A. Dec za Sepa.b p-valuec Memberd Y. Kato et al. (SPIRE) (J2000) (J2000) (mJy) (mJy) (mJy) (Count.) (J2000) (J2000) (arcsec) Colour-selected bright SPIRE sources HS1700-SPIRE21 255.3951 64.2484 52.5 ± 2.2 41.2 ± 1.7 21.1 ± 2.2 BNB1 255.3953 64.2479 ∼2.00e 1.63 < 0.01 Secure BX980 255.3987 64.2490 - 6.13 0.32 Not HS1700-SPIRE42 255.2428 64.2196 40.0 ± 2.3 46.4 ± 1.8 34.3 ± 2.1 DRG46 255.2419 64.2195 - 1.37 < 0.01 Secure HaNB2 255.2417 64.2196 - 1.75 < 0.01 Secure HS1700-SPIRE24 255.3349 64.2403 51.5 ± 2.3 63.0 ± 2.0 49.3 ± 2.2 DRG53 255.3378 64.2395 ∼ 2.3f 5.36 0.03 Secure HaNB10* 255.3399 64.2384 2.289(abs) 10.43 0.10 Tentative BX913* 255.3399 64.2385 2.289(abs), 2.291(neb) 10.29 0.05 Tentative BX928 255.3318 64.2405 2.755(Lyα) 4.92 0.22 Not Colour-selected faint SPIRE sources HS1700-SPIRE321 255.2732 64.2046 13.7 ± 2.4 20.2 ± 1.8 15.9 ± 2.1 BNB41* 255.2688 64.2027 2.287(abs) 9.64 0.05 Secure MD109* 255.2687 64.2026 2.293(abs,neb) 10.09 0.02 Secure DRG38+ 255.2669 64.2032 2.286(emi) 10.98 0.10 Tentative BNB16+ 255.2670 64.2033 2.290(emi) 10.53 0.05 Tentative at least one SPIRE band above 12 mJy sources HS1700-SPIRE70 255.2827 64.2586 31.1 ± 2.2 21.4 ± 1.9 11.0 ± 2.0 BNB155 255.2811 64.2574 2.290(Lyα) 5.13 0.01 Secure HaNB27 255.2873 64.2578 - 7.81 0.06 Tentative HS1700-SPIRE78 255.4358 64.2607 30.3 ± 2.2 16.9 ± 1.7 6.0 ± 2.1 BNB139 255.4331 64.2622 - 6.85 0.02 Secure HS1700-SPIRE142 255.2668 64.2469 22.7 ± 2.2 12.0 ± 1.8 1.8 ± 2.2 HaNB83 255.2653 64.2487 - 6.67 0.04 Secure HaNB76 255.2661 64.2461 - 3.06 0.01 Secure HS1700-SPIRE179 255.3768 64.1988 19.8 ± 2.2 9.4 ± 1.8 0.5 ± 2.2 HaNB45 255.3796 64.1996 - 5.41 0.03 Secure Notes. (a): Redshift information of counterparts. (b): Separation between SPIRE source and counterparts. (c): p-value based on Downes et al. (1986). (d): We classify sources with p ≤ 0.05 as secure, and those with 0.05 < p ≤ 0.10 as tentative counterparts. (e): BNB1 is identified as a âĂIJLo-BALâĂİ QSO and therefore it is difficult to measure redshift. (f ): Photometric redshift from Chapman et al. (2015). MNRAS 000, 1–15 (2016) Herschel protocluster survey 15 This paper has been typeset from a TEX/LATEX file prepared by the author. MNRAS 000, 1–15 (2016)