Fumiaki Nakata

Subaru Prime Focus Camera — Suprime-Cam

We have built an 80-mega pixels (10240 ×8192) mosaic CCD camera, called Suprime-Cam, for the wide-field prime focus of the 8.2m Subaru telescope. Suprime-Cam covers a field of view 34 � ×27 � , a unique facility among the 8-10m class telescopes, with a resolution of 0. �� 202 per pixel. The focal plane consists of ten high-resistivity 2k ×4k CCDs developed by MIT Lincoln Laboratory, which are cooled by a large Stirling-cycle cooler. The CCD readout electronics was designed to be scalable, which allows the multiple read-out of tens of CCDs. It takes 50 seconds to readout entire arrays. We designed a filter-exchange mechanism of the jukebox type that can hold up to ten large filters (205 ×170 ×15mm 3 ). The wide-field corrector is basically a three-lens Wynne-type, but has a new type of atmospheric dispersion corrector. The corrector provides a flat focal plane and an un-vignetted field of view of 30 � in diameter. The achieved co-planarity of the focal array mosaic is smaller than 30 µm peak-to-peak, which realizes mostly the seeing limited image over the entire field. The median seeing in the Ic-band, measured over one year and ah alf, is 0. 61. The PSF anisotropy in Suprime-Cam images, estimated by stellar ellipticities, is about 2% under this median seeing condition. At the time of commissioning, Suprime-Cam had the largest survey speed, which is defined as the field of view multiplied by the primary mirror area of the telescope, among those cameras built for

The Subaru/XMM-Newton Deep Survey (SXDS) - II. Optical Imaging and Photometric Catalogs

We present multiYwave band optical imaging data obtained from observations of the Subaru/XMM-Newton Deep Survey (SXDS). The survey field, centered at R:A: ¼ 02 h 18 m 00 s , decl: ¼� 05 � 00 0 00 00 , has been the focus of a wide range of multiwavelength observing programs spanning from X-ray to radio wavelengths. A large part of the optical imaging observations are carried out with Suprime-Cam on Subaru Telescope at Mauna Kea in the course of Subaru Telescope ‘‘Observatory Projects.’’ This paper describes our optical observations, data reduction and analysis procedures employed, and the characteristics of the data products. A total area of 1.22 deg 2 is covered in five contiguous subfields,eachof whichcorrespondstoasingleSuprime-Camfieldof view(� 34 0 ; 27 0 ),infivebroadbandfilters,B, V,Rc,i 0 ,andz 0 ,tothedepthsof B ¼ 28:4,V ¼ 27:8,Rc ¼ 27:7,i 0 ¼ 27:7,andz 0 ¼ 26:6,respectively(AB,3 � , � ¼ 2 00 ). The data are reduced and compiled into five multiYwave band photometric catalogs, separately for each SuprimeCampointing.Thei 0 -bandcatalogscontainabout900,000objects,makingtheSXDScatalogsoneof thelargestmultiY wavebandcatalogsincorrespondingdepthandareacoverage.TheSXDScatalogscanbeusedforanextensiverangeof astronomicalapplicationssuchasthenumberdensityof theGalactichalostarstothelarge-scalestructuresatthedistant universe. The number counts of galaxies are derived and compared with those of existing deep extragalactic surveys. The optical data, the source catalogs, and configuration files used to create the catalogs are publicly available via the SXDS Web page (http://www.naoj.org/Science/SubaruProject/SXDS/index.html). Subject headingg cosmology: observations — galaxies: evolution — galaxies: formation — galaxies: photometry — large-scale structure of universe

Down‐sizing in galaxy formation at z∼ 1 in the Subaru/XMM–Newton Deep Survey (SXDS)

We use the deep wide-field optical imaging data of the Subaru/XMM‐Newton Deep Survey to discuss the luminosity- (mass-)dependent galaxy colours down to z � = 25.0 (5 × 10 9 h −2 70 M� ) for z ∼ 1 galaxies in colour-selected high-density regions. We find an apparent absence of galaxies on the red colour‐magnitude sequence below z � ∼ 24.2, corresponding to ∼M ∗ + 2( ∼ 10 10 M� ) with respect to passively evolving galaxies at z ∼ 1. Galaxies brighter than M ∗ − 0.5 (8 × 10 10 M� ), however, are predominantly red passively evolving systems, with few blue star-forming galaxies at these magnitudes. This apparent age gradient, where massive galaxies are dominated by old stellar populations while less massive galaxies have more extended star formation histories, supports the ‘downsizing’ idea where the mass of galaxies hosting star formation decreases as the Universe ages. Combined with the lack of evolution in the shape of the stellar mass function for massive galaxies since at least z ∼ 1, it appears that galaxy formation processes (both star formation and mass assembly) should have occurred in an accelerated way in massive systems in highdensity regions, while these processes should have been slower in smaller systems. This result provides an interesting challenge for modern cold dark matter based galaxy formation theories which predict later formation epochs of massive systems, commonly referred to as ‘bottom-up’.

On the evolution and environmental dependence of the star formation rate versus stellar mass relation since Z ∼ 2

This paper discusses the evolution of the correlation between galaxy star formation rates (SFRs) and stellar mass (M) over the last ∼10 Gyr, particularly focusing on its environmental dependence. We first present the mid-infrared (MIR) properties of the Hα-selected galaxies in a rich cluster Cl 0939+4713 at z = 0.4. We use wide-field Spitzer/MIPS 24 μm data to show that the optically red Hα emitters, which are most prevalent in group-scale environments, tend to have higher SFRs and higher dust extinction than the majority population of blue Ha sources. With an MIR stacking analysis, we find that the median SFR of Hα emitters is higher in higher density environment at z = 0.4. We also find that star-forming galaxies in highdensity environment tend to have higher specific SFR (SSFR), although the trend is much less significant compared to that of SFR. This increase of SSFR in high-density environment is not visible when we consider the SFR derived from Hα alone, suggesting that the dust attenuation in galaxies depends on environment; galaxies in high-density environment tend to be dustier (by up to ∼0.5 mag), probably reflecting a higher fraction of nucleated, dusty starbursts in higher density environments at z = 0.4. We then discuss the environmental dependence of the SFR-M relation for star-forming galaxies since z ∼ 2, by compiling our comparable, narrow-band-selected, large Haemitter samples in both distant cluster environments and field environments. We find that the SSFR of Hα-selected galaxies (at the fixed mass of log(M*/M⊙) = 10) rapidly evolves as (1 + z)3, but the SFR-M* relation is independent of the environment since z ∼ 2, as far as we rely on the Hα-based SFRs (with M*-dependent extinction correction). Even if we consider the possible environmental variation in the dust attenuation, we conclude that the difference in the SFR-M* relation between cluster and field star-forming galaxies is always small (≲0.2 dex level) at any time in the history of the Universe since z ∼ 2.

Discovery of a Very Extended Emission-Line Region around the Seyfert 2 Galaxy NGC 4388

We found a very large, D35 kpc, emission-line region around the Seyfert type 2 galaxy NGC 4388, using deep narrowband imaging with the prime focus camera (Suprime-Cam) of the Subaru Telescope. This region consists of many faint gas clouds, or —laments, and extends northeastward from the galaxy. The typical Ha luminosity L (Ha) of the clouds is D1037 ergs s~1, and the total L (Ha) of the region within 10 kpc of the nucleus is D2 ] 1038 ergs s~1, which corresponds to an ionized gas mass of D105 The map of the emission-line intensity ratio I([O III])/I(Ha) indicates that the inner (r \ 12 kpc) M _ . region of the very extended emission-line region (VEELR) may be excited by nuclear ionizing radiation. The excitation mechanism of the outer (r [ 12 kpc) region is unclear, but it is likely that the nuclear radiation is also a dominant source of its ionization. We discuss the origin of the ionized gas. Two plausible origins of the gas in the VEELR are (1) the tidal debris resulting from a past interaction with a gas-rich dwarf galaxy, i.e., a minor merger, and (2) the interstellar medium of NGC 4388, stripped by the ram pressure of the hot intracluster medium of the Virgo Cluster.

Narrowband Imaging in [O III] and H-alpha to Search for Intracluster Planetary Nebulae in the Virgo Cluster

We have identified intracluster planetary nebulae (PNs) in a Virgo Cluster core field by imaging with the Subaru Suprime-Cam through two narrowband filters centered at the redshifted wavelengths of the [O III] λ5007 and the Hα 6563 A lines; broadband images in V and R were acquired to check for emission in the adjacent continuum. Emission-line objects in Virgo are then selected from the two-color diagram [O III]-Hα versus [O III]-(V+R), which is calibrated using PNs in M84 (Jacoby et al.). Using both [O III] and Hα allows us to distinguish bona fide planetary nebulae from high-redshift emission-line galaxies at the bright end of the [O III] luminosity function. Spectroscopic observations of a subsample of these objects were made at the Telescopio Nazionale Galileo and at the Very Large Telescope in a region around M84 and in an intracluster field, respectively. The observations confirm the efficiency of the combined [O III]+Hα imaging to identify true PNs. We also obtained the first spectrum of an intracluster PN that shows the [O III] doublet with a signal-to-noise ratio greater than 10 and its Hα emission. From the results based on the spectroscopic follow-up, we derive a lower limit to the fraction of the Virgo Cluster light contributed by the intracluster stars at the surveyed position in the cluster core: it amounts to 10%.

Optical Spectropolarimetry of SN 2002ap: A High-Velocity Asymmetric Explosion*

We present spectropolarimetry of the Type Ic supernova SN 2002ap and give a preliminary analysis: the data were taken at two epochs, close to and 1 month later than the visual maximum (2002 February 8). In addition, we present June 9 spectropolarimetry without analysis. The data show the development of linear polarization. Distinct polarization profiles were seen only in the O I λ7773 multiplet/Ca II IR triplet absorption trough at maximum light and in the O I λ7773 multiplet and Ca II IR triplet absorption troughs a month later, with the latter showing a peak polarization as high as ~2%. The intrinsic polarization shows three clear position angles: 80° for the February continuum, 120° for the February line feature, and 150° for the March data. We conclude that there are multiple asymmetric components in the ejecta. We suggest that the supernova has a bulk asymmetry with an axial ratio projected on the sky that is different from 1 by an order of 10%. Furthermore, we suggest very speculatively that a high-velocity ejecta component moving faster than ~0.115c (e.g., a jet) contributes to polarization in the February epoch.

SUBARU DEEP SURVEY. III. EVOLUTION OF REST-FRAME LUMINOSITY FUNCTIONS BASED ON THE PHOTOMETRIC REDSHIFTS FOR A K 0 -BAND-SELECTED GALAXY SAMPLE 1

We have constructed a very deep K 0 -selected multicolor BVRIz 0 JK 0 sample of 439 field galaxies. Based on this multicolor data, a photometric redshift for each sample galaxy was estimated. The overall redshift distribution N(z) for the K 0 � 21.0 sample is consistent with previous observations, and for the first time we derive N(z) down to K 0 = 24.0. After taking account of the dust extinction and selection effects of the sample, the observed N(z) distribution is well described with the pure luminosity evolution (PLE) model, while the hierarchical galaxy formation model shows an apparent deficiency of galaxies especially at z e 2. The photometric redshift and the best-fit spectral energy distribution model evaluations allow the derivation of the restframe K 0 , B, and UV (2000 A ˚ ) band luminosity functions and their evolutions. The rest K 0 -band luminosity function shows almost no evolution up to z = 3, while the rest B luminosity function shows mild luminosity evolution, and the rest UV luminosity function shows strong luminosity evolution. These findings seem to be qualitatively in favor of the PLE model. No evolution in the rest K 0 -band luminosity function can also be consistent with the hierarchical galaxy formation model if M/LK decreases with redshift, that is, if the number density of K 0 -band luminosity-selected galaxies does not significantly change with redshift while the number density of stellar mass–selected galaxies decreases with look-back time. This trend corresponds to the evolution of the rest UV (2000 A ˚ )� K 0 color, which gets bluer with increasing redshift. We also found that more massive galaxies are redder in this rest-frame color in every epoch. The rest UV luminosity function of our K 0 -selected galaxies shows a much shallower faint-end slope at z = 3 than that of previous estimations for rest UV–selected Lyman break galaxies. As a consequence, the contribution to the global star formation rate of our K 0 -selected galaxies is about 42% of that derived from the integration of the luminosity function of Lyman break galaxies at z = 3. This result suggests that a large fraction of the star formation rate density at z > 1.5 comes from the contribution from a faint (M2000 > � 20) blue galaxy population at high redshift that has not yet obviously been identified.

A Search for Lyα Emitters at Redshift 3.7

We present the results of a survey for emission-line objects based on optical intermediate-band (

Early-type Galaxies at z = 1.3. I. The Lynx Supercluster: Cluster and Groups at z = 1.3. Morphology and Color-Magnitude Relation

\lambda_{\rm c}