High-sensitivity wide-band X-ray spectroscopy is the key feature of the Suzaku X-ray observatory, launched on 2005 July 10. This paper summarizes the spacecraft, in-orbit performance, operations, and data processing that are related to observations. The scientific instruments, the high-throughput X-ray telescopes, X-ray CCD cameras, non-imaging hard X-ray detector are also described.

We present a 0.72 deg2 contiguous 1.1‐mm survey in the central area of the Cosmological Evolution Survey field carried out to a 1σ≈ 1.26 mJy beam−1 depth with the AzTEC camera mounted on the 10‐m Atacama Submillimeter Telescope Experiment. We have uncovered 189 candidate sources at a signal‐to‐noise ratio (S/N) ≥ 3.5, out of which 129, with S/N ≥ 4, can be considered to have little chance of being spurious (≲2 per cent). We present the number counts derived with this survey, which show a significant excess of sources when compared to the number counts derived from the ∼0.5 deg2 area sampled at similar depths in the Submillimetre Common‐User Bolometer Array (SCUBA) HAlf Degree Extragalactic Survey (SHADES). They are, however, consistent with those derived from fields that were considered too small to characterize the overall blank‐field population. We identify differences to be more significant in the S1.1mm≳ 5 mJy regime, and demonstrate that these excesses in number counts are related to the areas where galaxies at redshifts z≲ 1.1 are more densely clustered. The positions of optical–infrared galaxies in the redshift interval 0.6 ≲z≲ 0.75 are the ones that show the strongest correlation with the positions of the 1.1‐mm bright population (S1.1mm≳ 5 mJy), a result which does not depend exclusively on the presence of rich clusters within the survey sampled area. The most likely explanation for the observed excess in number counts at 1.1‐mm is galaxy–galaxy and galaxy–group lensing at moderate amplification levels, which increases in amplitude as one samples larger and larger flux densities. This effect should also be detectable in other high‐redshift populations.

We report the results of the counterpart identification and a detailed analysis of the physical properties of the 48 sources discovered in our deep 1.1‐mm wavelength imaging survey of the Great Observatories Origins Deep Survey‐South (GOODS‐S) field using the AzTEC instrument on the Atacama Submillimeter Telescope Experiment. One or more robust or tentative counterpart candidate is found for 27 and 14 AzTEC sources, respectively, by employing deep radio continuum, Spitzer/Multiband Imaging Photometer for Spitzer and Infrared Array Camera, and Large APEX Bolometer Camera 870 μm data. Five of the sources (10 per cent) have two robust counterparts each, supporting the idea that these galaxies are strongly clustered and/or heavily confused. Photometric redshifts and star formation rates (SFRs) are derived by analysing ultraviolet(UV)‐to‐optical and infrared(IR)‐to‐radio spectral energy distributions (SEDs). The median redshift of zmed∼ 2.6 is similar to other earlier estimates, but we show that 80 per cent of the AzTEC–GOODS sources are at z≥ 2, with a significant high‐redshift tail (20 per cent at z≥ 3.3). Rest‐frame UV and optical properties of AzTEC sources are extremely diverse, spanning 10 mag in the i‐ and K‐band photometry (a factor of 104 in flux density) with median values of i= 25.3 and K= 22.6 and a broad range of red colour (i−K= 0–6) with an average value of i−K≈ 3. These AzTEC sources are some of the most luminous galaxies in the rest‐frame optical bands at z≥ 2, with inferred stellar masses M*= (1–30) × 1010 M⊙ and UV‐derived SFRs of SFRUV≳ 101‐3 M⊙ yr−1. The IR‐derived SFR, 200–2000 M⊙ yr−1, is independent of z or M*. The resulting specific star formation rates, SSFR ≈ 1–100 Gyr−1, are 10–100 times higher than similar mass galaxies at z= 0, and they extend the previously observed rapid rise in the SSFR with redshift to z= 2–5. These galaxies have a SFR high enough to have built up their entire stellar mass within their Hubble time. We find only marginal evidence for an active galactic nucleus (AGN) contribution to the near‐IR and mid‐IR SEDs, even among the X‐ray detected sources, and the derived M* and SFR show little dependence on the presence of an X‐ray bright AGN.

Lyman-alpha emitters are thought to be young, low-mass galaxies with ages of approximately 10(8) yr (refs 1, 2). An overdensity of them in one region of the sky (the SSA 22 field) traces out a filamentary structure in the early Universe at a redshift of z approximately 3.1 (equivalent to 15 per cent of the age of the Universe) and is believed to mark a forming protocluster. Galaxies that are bright at (sub)millimetre wavelengths are undergoing violent episodes of star formation, and there is evidence that they are preferentially associated with high-redshift radio galaxies, so the question of whether they are also associated with the most significant large-scale structure growing at high redshift (as outlined by Lyman-alpha emitters) naturally arises. Here we report an imaging survey of 1,100-microm emission in the SSA 22 region. We find an enhancement of submillimetre galaxies near the core of the protocluster, and a large-scale correlation between the submillimetre galaxies and the low-mass Lyman-alpha emitters, suggesting synchronous formation of the two very different types of star-forming galaxy within the same structure at high redshift. These results are in general agreement with our understanding of the formation of cosmic structure.

The source counts of galaxies discovered at submillimetre and millimetre wavelengths provide important information on the evolution of infrared-bright galaxies. We combine the data from six blank-field surveys carried out at 1.1 mm with AzTEC, totalling 1.6 deg 2 in area with root-mean-square depths ranging from 0.4 to 1.7 mJy, and derive the strongest constraints to date on the 1.1 mm source counts at flux densities S 1100 = 1-12 mJy. Using additional data from the AzTEC Cluster Environment Survey to extend the counts to S 1100 ∼ 20 mJy, we see tentative evidence for an enhancement relative to the exponential drop in the counts at S 1100 ∼ 13 mJy and a smooth connection to the bright source counts at >20 mJy measured by the South Pole Telescope; this excess may be due to strong-lensing effects. We compare these counts to predictions from several semi-analytical and phenomenological models and find that for most the agreement is quite good at flux densities 4 mJy; however, we find significant discrepancies ( 3σ ) between the models and the observed 1.1-mm counts at lower flux densities, and none of them is consistent with the observed turnover in the Euclideannormalized counts at S 1100 2 mJy. Our new results therefore may require modifications to existing evolutionary models for low-luminosity galaxies. Alternatively, the discrepancy between the measured counts at the faint end and predictions from phenomenological models could arise from limited knowledge of the spectral energy distributions of faint galaxies in the local Universe.