EDP Sciences
Free Access
Volume 577, May 2015
Article Number A29
Number of page(s) 32
Section Extragalactic astronomy
DOI https://doi.org/10.1051/0004-6361/201425032
Published online 27 April 2015

Online material

Appendix A: Multiwavelength images

A selection of zoomed-in multiwavelength views towards AzTEC1630 is shown in Fig. A.1. In the first (top left) panel of each source we show the PdBI 1.3 mm image overlaid with the same contour levels as in Fig. 1. The PdBI images are also annotated with the source designations. The positive 1.3 mm contours are overlaid on the other wavelength images to guide the eye.

thumbnail Fig. A.1

Multiwavelength views towards AzTEC1630. The panels from top left to bottom right for each source are as follows: PdBI 1.3 mm, VLA 20 cm, VLA 10 cm, Spitzer 24 μm, Spitzer 8 μm, Spitzer 3.6 μm, UltraVISTA YJHKs colour composite, and HST/ACS I-band. The overlaid 1.3 mm contours are as in Fig. 1, and positive 1.3 mm contours are shown in all panels. The synthesised beam of the PdBI data is shown in the bottom left corner in the first panel for each source. A scale bar indicating the 1″ projected length is shown in the PdBI panel, and the corresponding proper length [kpc] at the indicated redshift is also denoted (except when only a lower limit to z could be derived). The catalogue positions of the Herschel/SPIRE 250 μm sources are marked with plus signs in the PdBI images towards AzTEC19, 20, and 24. The diamond symbol in the PdBI image towards AzTEC24 indicates the position of the ASTE/AzTEC 1.1 mm source AzTEC/C48 from Aretxaga et al. (2011).

Open with DEXTER

Appendix B: Updated redshifts of the 15 brightest JCMT/AzTEC-detected SMGs: AzTEC115

Here we provide the reader with an overview of the redshifts of the SMA-detected SMGs AzTEC115. Among these SMGs, there are eight spectroscopic redshifts reported in the literature: for AzTEC1, 2, 3, 5, 6, 8, 9, and 11 (see Smolčić et al. 2012b; their Tables 1 and 4 and references therein). Despite the partially extensive efforts and data coverage some of these redshift determinations are still uncertain. We discuss below the updated redshifts among AzTEC115 and the cases where there is some confusion about the source redshifts in the literature.

Smolčić et al. (2011) determined a spectroscopic redshift of 4.650 ± 0.005 for AzTEC1. The UV–NIR photometric redshift they derived, , was found to be very similar to the zspec value, although a secondary photo-z solution at zphot = 4.44 was also found. A somewhat lower photo-z of was derived by Smolčić et al. (2012b) using the same method as in the present paper (Sect. 4.2). The CO spectral-line observations using the Redshift Search Receiver (RSR) on the Large Millimetre Telescope (LMT) performed by Yun et al. (2012) yielded a spec-z value of 4.3421 for AzTEC1. Their SMA follow-up observations of C+ emission yielded a line detection at zspec = 4.3415, in very good agreement with the CO observations. Since it is based on interferometric observations, this last redshift is adopted in the present work. We note that the new spec-z of AzTEC1 explains the non-detection of the CO(5−4) line emission by Smolčić et al. (2011) because their PdBI and Combined Array for Research in Millimetre-wave Astronomy (CARMA) observations covered the redshift ranges 4.564.76 and 4.945.02.

The optical spectrum observed with the Deep Extragalactic Imaging Multi-Object Spectrograph (DEIMOS) on the 10 m Keck II telescope towards AzTEC2 exhibits an emission feature that can be assigned to the [ O ii ]λ3727 forbidden-line doublet at zspec = 1.124, and the J = 2−1 rotational line of CO detected with CARMA suggests a similar redshift (zspec = 1.126; Baloković et al., in prep.). Following Smolčić et al. (2012b), we adopt the value zspec = 1.125 as the redshift of AzTEC2. Koprowski et al. (2014) claimed that the target position of these spectral line observations was away from the SMA 890 μm position (Younger et al. 2007). They concluded that the SMG lies at a redshift of derived from the radio/submm flux-density ratio because the radio source is only away from the SMA position. The redshift we derived from the radio/submm flux-density ratio is z = 4.28 ± 0.82. The latter difference emerges because Koprowski et al. (2014) based their calculation on the average z ≃ 2−3 SMG spectral template derived by Michałowski et al. (2010), while we utilised the Carilli-Yun redshift indicator (Carilli & Yun 1999, 2000) as described in Sect. 4.2. However, as shown in Fig. E.1, the Keck/DEIMOS slit was centred from the SMA position, and the spectrum was extracted from the SMA peak of the SMG. Moreover, the EW-oriented DEIMOS slit did not cover the optically visible foreground galaxy on the southern side of AzTEC2 (UltraVISTA ID 232116, zphot = 0.34; cf. Fig. B1 in Koprowski et al. 2014). This implies that the redshift of AzTEC2 is close to unity instead of the higher value proposed by Koprowski et al. (2014).

The spec-z of AzTEC5 was previously reported to be zspec = 3.971 (Smolčić et al. 2012b). The Keck/DEIMOS slit position and orientation are shown in Fig. E.1. We note that the slit does not include emission from galaxies other than AzTEC5, but the corresponding DEIMOS spectrum is of poor quality. Therefore, we adopt the photo-z of from Smolčić et al. (2012b). For comparison, Koprowski et al. (2014) derived a photo-z of . The redshift we derived from the radio/submm flux-density ratio using the Carilli & Yun (2000) formula (see Sect 4.2), 1.85 ± 0.23, is also lower than the value calculated by Koprowski et al. (2014).

Koprowski et al. (2014) argued that the spectroscopic redshift of AzTEC6, zspec = 0.802, is uncertain because it is measured towards an optically visible object about 1″ from the SMA position (Younger et al. 2007), and that the submm/radio flux ratio of AzTEC6 is inconsistent with a low redshift (they derived a value of from the radio/submm flux-density ratio, while we derive the value z> 3.52 because AzTEC6 is not detected at 20 cm). The photo-z value derived by Smolčić et al. (2012b), , is similar to the zspec value, while Koprowski et al. (2014) reported a value of zphot = 1.12 for this object. The above-mentioned optically visible galaxy lies only from the SMA position, and away from the Keck/DEIMOS slit centre (see Fig. E.1). The above spec-z value was derived from a high quality spectrum (flag 4; Kartaltepe et al., in prep.) extracted from a position that lies from the SMA position, and coincides with the optical galaxy. This implies that the spectral line emission originates in this foreground object as suggested by Koprowski et al. (2014). A redshift of zspec = 0.802 indeed conflicts with the non-detection of AzTEC6 at 20 cm, and we therefore adopt the redshift z> 3.52.

The DEIMOS spec-z of AzTEC9 was reported to be 1.357, and its corresponding photo-z was found to be (Smolčić et al. 2012b). However, the spec-z value is based on a relatively weak spectrum (Salvato et al., in prep.), and is therefore quite uncertain. Koprowski et al. (2014) reported that the above redshift values refer to an object as far as about from the SMA position (Younger et al. 2009). Again, this is not the case, but the DEIMOS spectrum was extracted from the SMA position (the slit centre was offset from the SMA peak by ; see Fig. E.1). Koprowski et al. (2014) stated that the submm/radio flux-density ratio of AzTEC9 is inconsistent with a redshift value close to unity (they derived , while our result based on the Carilli & Yun 2000 redshift formula is z = 2.82 ± 0.76). They also derived a high photo-z of for AzTEC9 (counterpart lying from the SMA position). There is a Spitzer/IRAC source from the SMA position, and the Wardlow et al. (2011) redshift formula gives a redshift of z ≃ 2.75, which is similar to the redshift we inferred from the radio/submm flux-density ratio, but considerably lower than the redshifts derived by Koprowski et al. (2014). In the present study, we adopt the photo-z of from Smolčić et al. (2012b) because the corresponding distribution exhibits a clear minimum at that value (see Fig. 6 in Smolčić et al. 2012b).

For AzTEC10, the photo-z derived by Smolčić et al. (2012b) is , while Koprowski et al. (2014) determined a photo-z of for the optical/NIR source about from the SMA position. The most up-to-date COSMOS spectroscopic-redshift catalogue gives a likely (quality flag 2) DEIMOS redshift value of zspec = 0.547 towards AzTEC10 (only offset from the SMA position; Salvato et al., in prep.). As illustrated in Fig. E.1, the DEIMOS slit however picked up emission from a foreground galaxy at zphot ≃ 0.51 (ID 302846 in the new UltraVISTA catalogue) that lies NW of AzTEC10. Since AzTEC10 is not detected at 1.4 GHz, it appears to lie at a high redshift. In the present study, we adopt the photo-z from Smolčić et al. (2012b), but note that because of the multiple nearby counterparts of this source (three within 2″) it is difficult to obtain accurate photometry for AzTEC10.

Both AzTEC13 and AzTEC14-E have neither optical nor IRAC counterparts, and we derived lower limits of z> 4.07 and z> 2.95 for their radio/submm flux-ratio based redshifts (these differ from the values z> 3.59 and z> 3.03 derived by Smolčić et al. 2012b because of the different assumptions that we used here). These lower limits are consistent with the corresponding values of Koprowski et al. (2014), i.e. and , respectively. The updated COSMOS spec-z catalogue gives a high quality (flag 4) DEIMOS redshift of zspec = 0.471 for a target that is away from the SMA position of AzTEC13 (Salvato et al., in prep.), where the spectroscopic slit centre was positioned away from the SMA peak). This redshift is much lower than the other estimates mentioned above. However, as shown in Fig. E.1, there are two foreground galaxies lying SE and SW from AzTEC13 (UltraVISTA IDs 268116 and 268129 with the photo-z values of 0.49 and 0.45, respectively); these could have contaminated the spectral line measurements, although they do not lie within the slit boundaries. A low redshift of AzTEC13 would indeed be inconsistent with the radio non-detection (cf. AzTEC6). For both AzTEC13 and AzTEC14-E, we adopt the redshifts derived from the radio/submm flux ratio (z> 4.07 and z> 2.95).

Appendix C: Multiwavelength counterparts and redshifts of the SMGs AzTEC1630

Below we describe the multiwavelength appearances of our PdBI SMGs and provide notes of their redshifts.

AzTEC16. The 5σ point-like 1.3 mm emission feature near the edge of the PB FWHM appears to have no counterparts at other wavelengths. Altogether four negative sources were found in this field with | S/N | = 4.0−6.2, three of which lie outside the PB. Hence, AzTEC16 could be spurious despite its relatively high significance. We note that west of the target field centre, there is a ~3σ (35.6 μJy beam-1) VLA 20 cm source, which also appears to be detected at Spitzer/IRAC and 24 μm wavelengths. This source can be identified as the galaxy J095950.03+024416.5 from the COSMOS optical/NIR catalogue (Capak et al. 2007); its UltraVISTA DR1 photometric-redshift catalogue ID is 319194 (z ≃ 1.62; Ilbert et al. 2013). The radio non-detection of AzTEC16 suggests a lower limit of z> 2.42 to its redshift. In Table 1 of Smolčić et al. (2012b), the source called AzTEC16 at a spectroscopic redshift of 1.505 (based on high-resolution CO observations with CARMA; Sheth et al., in prep.) corresponds to AzTEC42 in our nomenclature (cf. Scott et al. 2008; Table 1 therein).

AzTEC17. The 1.3 mm source AzTEC17a (lying SW of the pointing centre, and detected at a significance level of 6.2σ) is clearly associated with the 20 cm source COSMOSVLADPJ095939.19+023403.6 (S20 cm = 68 ± 13μJy), and a VLA 10 cm source (37 μJy beam-1 or 8.2σ). The source also shows Spitzer/MIPS and IRAC emission. There is a Herschel 250 μm source east of AzTEC17a (ID 1753 in the COSMOS SPIRE 250 μm Photometry Catalogue from HerMES; Oliver et al. 2012). The search for Herschel counterparts was performed by using a search radius of , i.e. half the SPIRE beam FWHM at λ = 250μm. AzTEC17a has an optical-NIR counterpart about SW of the 1.3 mm emission peak (ID 1475165 in the COSMOS photometry catalogue Capak et al. 2007). We note that the source visible in the UltraVISTA and ACS I-band images, lying NW in projection from AzTEC17a, is the galaxy COSMOS J095939.12+023405.1 (Capak et al. 2007), which has a photometric redshift of z = 0.793 (the source 271694 in the DR1 UltraVISTA photometric-redshift catalogue; Ilbert et al. 2013). Cross-correlation with the COSMOS photometry catalogue yielded a candidate optical counterpart for AzTEC17b (ID 1475223), about SW of the PdBI emission peak. However, AzTEC17b has no counterparts at UltraVISTA bands or at MIR or cm wavelengths. There are two negative sources (−4.5σ and −6.2σ) within ≲ 7″ of the phase centre. As AzTEC17a is a confirmed SMG, this could mean that the 4.5σ source AzTEC17b is spurious.

For AzTEC17a, the primary photo-z solution is . Because the source of this photo-z lies about from the PdBI position, it is questionable whether it is related to AzTEC17a (although it is within the synthesised beam). The primary photo-z value is however comparable to the very secure (quality flag 3) spectroscopic redshift zspec = 0.834 measured towards AzTEC17a ( offset from the PdBI position) with Keck/DEIMOS (Salvato et al., in prep.). We note that for the measurements the slit was centred on a position away from AzTEC17a (see Appendix E), and that there are no HST/ACS I-band sources within the slit boundaries. For comparison, the angular offset between the 1.4 GHz radio source and the PdBI detection peak is only , and the redshift derived from the radio/submm flux density ratio is z = 2.29 ± 0.42. A comparable redshift of z ≃ 2.77 was derived from the Wardlow et al. (2011; their Eq. (1)) redshift estimator based on the Spitzer/IRAC 3.6 μm and 8 μm flux densities. The last two values agree with a shallow “bowl” in the distribution at z ≃ 2.7, but the zspec value is adopted in the present work. For AzTEC17b we derive a photo-z of . There is a dip in the distribution also at z ≃ 0.4. However, the 1.4 GHz non-detection towards AzTEC17b results in a lower limit to its redshift of z> 2.49, which is consistent with the above photo-z value of .

AzTEC18. The 1.3 mm 4.5σ source seen towards AzTEC18 has counterparts at optical and NIR wavelengths ( from the PdBI position), and is therefore considered a potential SMG. This is further supported by the fact that only one negative source, being of −4.2σ significance, was detected in this field.

AzTEC18 has a photo-z solution of , which is consistent with the redshift derived from the radio/submm flux-density ratio of z> 2.20.

AzTEC19. AzTEC19a lies NE of the SCUBA-2 source SMMJ100028.6+023201 (or 450.00 or 850.07) identified by Casey et al. (2013). This angular offset is within the JCMT/SCUBA-2 beam size (FWHM) of ~7″ at 450 μm. With the deboosted flux densities of S450 μm = 37.54 ± 6.58 mJy and S850 μm = 9.21 ± 1.45 mJy, this was the strongest 450 μm source found by Casey et al. (2013) in the COSMOS field. These authors identified two possible optical counterparts to SMMJ100028.6+023201 (see their Table 6), but our higher-resolution observations show that only one of them – lying west of the PdBI peak position – can be taken as a candidate counterpart (the other source lies SW of the PdBI peak). AzTEC19a was also detected by Herschel. In the COSMOS SPIRE 250 μm Photometry Catalogue, the source ID is 2277 ( offset). AzTEC19a is associated with both a 20 and 10 cm radio-continuum source. In the VLA Deep Catalogue, the corresponding source has the name COSMOSVLADPJ100028.70+023203.7 (S20 cm = 78 ± 12μJy). The 10 cm peak flux density is S10 cm = 28.8μJy beam-1, making it a 6.4σ detection. The source is also associated with Spitzer IR emission. The 9.7σ 1.3 mm source AzTEC19b lying at the border of the PdBI PB has some Spitzer/IRAC 3.6 μm and lower wavelength emission just north of it. This emission can be associated with the galaxy COSMOS J100029.24+023211.5 (Capak et al. 2007), which has a photo-z of about 1.27 (source 262768; Ilbert et al. 2013). As it lies NW of the 1.3 mm peak, it is probably unrelated to AzTEC19b. However, there is also a NIR source within the 3σ contour of 1.3 mm emission, about from the mm peak. In the COSMOS ACS I-band photometry catalogue (Leauthaud et al. 2007), the ID of this source is 1486, while in the DR1 UltraVISTA catalogue its ID is 262766, and its reported photo-z value is about 1.30 (Ilbert et al. 2013). Three negative sources (−4.8σ, −6.6σ, −9.5σ) were detected in the AzTEC19 field. However, only one of them (−6.6σ) lies within the PB FWHM ( from the phase centre), while the remaining two lie outside the PB ( away from the phase centre). Because AzTEC19a is confirmed, and AzTEC19b has a high 1.3 mm detection S/N of 9.7 and is associated with multiwavelength emission, we are not expecting to have any spurious sources in this field.

The optical/IR counterpart of AzTEC19a (CFHT iK = 1.49), located only from the PdBI position, has a photo-z solution of . There is a spec-z value of 1.048 measured for a source only from the PdBI position of AzTEC19a with the VLT Visible Multi-Object Spectrograph (VIMOS) in the zCOSMOS project. However, the corresponding quality flag is 1.1, meaning that the zspec value is insecure (< 25% reliability). For the 1.4 GHz source – situated from the PdBI position – we derived a radio/submm-based redshift of 4.22 ± 0.91, which is comparable within the errors with our photo-z solution. For comparison, Casey et al. (2013) derived a photometric redshift of for their SMG source SMMJ100028.6+023201, which is associated with AzTEC19a. A comparable value of z ~ 2.3 can be derived from the Spitzer IRAC/MIPS flux densities (Pope et al. 2006; their Eq. (2)), and the value z ≃ 2.90 is obtained when using the IRAC 3.6 μm and 8 μm flux densities (Wardlow et al. 2011). In the present paper, we adopt our photo-z solution of for AzTEC19a. For AzTEC19b, the photo-z solution of zphot = 1.11 ± 0.10 is adopted as the redshift of the aforementioned SMG, while the 1.4 GHz non-detection suggests a very high radio/submm-based redshift of z> 6.57, where no SMGs have been discovered to date. There is, however, a degeneracy between the dust temperature of the source and its redshift (both affecting the source SED), and the radio dimness of AzTEC19b could in principle be the result of a low dust temperature (e.g. Blain et al. 2002; Kovács et al. 2006).

AzTEC20. Interestingly, the 6σ source about north of the phase centre shows no emission at other wavelengths. The ~3σ features of 1.3 mm emission seen near the phase centre appear instead to be associated with a source seen at several different wavelengths from 20 cm to NIR. The UltraVISTA catalogue ID of the latter source is 306331 (z ≃ 1.98; Ilbert et al. 2013). The 20 cm source has a peak flux density of 42.3 μJy beam-1, hence has a S/N ratio of about 3.5. We note that the VLA Deep catalogue contains sources down to 4σ or about 48 μJy beam-1 (Schinnerer et al. 2010). About west of this source, a slightly stronger 20 cm source candidate (45 μJy beam-1 or ~3.8σ) can be seen. The 10 cm source near the PdBI phase-tracking centre is a ~6.2σ detection (28.1 μJy beam-1). The Spitzer/MIPS 24 μm emission is quite extended, but is clearly resolved into two sources in the IRAC 3.6 μm image and yet more sources in the UltraVISTA NIR images. The Herschel/HerMES/SPIRE 250 μm catalogue (Oliver et al. 2012) contains a source (ID 3076) near the phase centre, from our PdBI source. There is one negative source of −4.8σ significance in this field, located on the SE side of our 1.3 mm source. Although the positive source is more significant (6σ) compared to the negative feature, and it fulfils our detection criterion of S/N> 4.5, it has no optical-to-IR counterparts that could confidently confirm that it is real. The lack of radio emission from AzTEC20 yields a lower redshift limit of z> 2.35.

AzTEC21. The south-western clump of the detected filamentary structure, AzTEC21a, is associated with a 20 cm source of peak flux density of 63 μJy beam-1 (~3.9σ). This also coincides with the position of a Spitzer IR source. AzTEC21b is probably part of the same structure (see below). The bright galaxy lying SE of AzTEC21c is COSMOS J100002.93+024639.9 (V = 20.681; Capak et al. 2007). The ID of this galaxy in the DR1 UltraVISTA catalogue is 327783, and its photo-z is about 0.34 (Ilbert et al. 2013). We note that the source ID in the zCOSMOS catalogue is 846495, but its spectroscopic redshift measurement could not be attempted (confidence class 0; Lilly et al. 2007, 2009).

For AzTEC21a, the photo-z solution is (for a source with the CFHT colour iK = 2.39). The optical/NIR source lies only away (NE) from the PdBI source according to the previous COSMOS/UltraVISTA catalogue (ID 1688587), but in the most recent UltraVISTA-TERAPIX DR the nearest source (ID 328878) lies NE, i.e. 8.5 times further away from our source, making the reliability of the proposed counterpart questionable. The 1.4 GHz non-detection (only 3.9σ) suggests a high redshift of z> 3.45. On the other hand, the Spitzer photometric redshift is estimated to be about 1.5 (Pope et al. 2006), and the value z ≃ 2 is derived using the Wardlow et al. 2011 IRAC flux-density method. The last value is comparable to our photo-z of . For AzTEC21b, we derived a redshift of (source from the PdBI peak; CFHT/iK = 1.74), while the radio/submm based value is z> 2.47, consistent with our photo-z solution. The most recent UltraVISTA DR does not contain a nearby () counterpart to AzTEC21b. We note that the overall 1.3 mm emission morphology (Fig. A.1) could indicate a relation between AzTEC21a and 21b, and this is further supported by their comparable photo-z values of and (i.e. their redshifts could be identical). For AzTEC21c, we derived a lower limit of z> 1.93 from the upper limit to the 1.4 GHz flux density.

AzTEC22. The candidate point-like 5.1σ PdBI source at the southern edge of the field has no counterparts at other wavelengths. The detection of two negative features with | S/N | = 5.7 and 5.9 in this field provides a hint that the 1.3 mm detection is spurious. A submm source was detected with Herschel towards our phase centre (250 μm ID 5470), about 8″ north from the above mentioned PdBI feature. Moreover, about from our pointing centre, there is the 20 cm source COSMOSVLADPPJ095950.57+022827.5 (S20 cm = 124 ± 12μJy), which is also seen in the VLA 10 cm image (58.6 μJy beam-1 or ~13σ). The Spitzer IR images show a “double source”, and a stronger 24 μm emitter is associated with the above-mentioned radio-continuum source J095950.57+022827.5. A trace of 20 cm emission (43.8 μJy beam-1 peak surface brightness) can also be seen towards the position of the weaker 24 μm source. The two Spitzer sources can be seen in the UltraVISTA NIR images: the NW one has the ID 244762 (zphot ≃ 1.8−1.9), while the SE source is 244405 at zphot ≃ 2 (Ilbert et al. 2013). For our PdBI detection the lack of a radio counterpart suggests a redshift of z> 3.0.

AzTEC23. The 1.3 mm feature seen towards this source ( NW of the AzTEC centroid) has counterparts at optical-NIR wavelengths ( from the PdBI position). Cross-correlation with the Herschel/SPIRE 250 μm catalogue shows the presence of a source (ID 2659) about 3″ NW from the 1.3 mm feature. A visual inspection of the VLA 20 cm image reveals an EW-oriented, elongated emission feature, whose western emission peak ( from the 1.3 mm source) has a peak surface brightness of 43.5 μJy beam-1 and the eastern peak has the surface brightness of 39.8 μJy beam-1.

We derived a primary photo-z of with a secondary solution at z ≃ 4.3, while the radio non-detection implies the lower limit z> 2.06. We adopt the redshift but note that the distribution of AzTEC23 is quite complex.

AzTEC24. This source is called AzTEC/C48 in the ASTE/AzTEC 1.1 mm catalogue of Aretxaga et al. (2011). The ASTE/AzTEC peak position lies NE of the JCMT/AzTEC centroid. We have found three candidate PdBI sources of 4.9−5.1σ significance (outside/at the border of the PB), but none of them have counterparts at other wavelengths. Altogether five negative sources with | S/N | = 4.3−5.9 were detected in the field, and two of them lie outside the PB FWHM. Hence, the identified positive 1.3 mm sources might be spurious and should be treated with caution. About SW from AzTEC24b there is a Herschel submm source (250 μm ID 4991). Moreover, SW from AzTEC24b there is the VLA 20 cm source COSMOSVLADPJ100039.28+023845.3 (S20 cm = 63 ± 13μJy). Aretxaga et al. (2011) associated this source with the radio counterpart of their source AzTEC/C48 ( separation). In the VLA 10 cm image, a source with a peak flux density of 28.5 μJy beam-1 (6.3σ) can be seen at the 20 cm source position. The cm radio-continuum source is also associated with Spitzer IR emission. When referring to the COSMOS catalogue of Capak et al. (2007), the source can be identified as the galaxy COSMOS J100039.29+023845.4, and in the UltraVISTA catalogue the source ID is 293896 (zphot ≃ 2.1; Ilbert et al. 2013). For the three radio non-detected components AzTEC24a, 24b, and 24c we derived lower redshift limits of z> 2.35, z> 2.28, z> 3.17, respectively.

AzTEC25. In this field, none of the 1.3 mm point-like features fulfilled our detection criteria. We note that of the 4.2σ feature (the most south-western feature shown in Fig. A.1) there is the source 1925434 from the COSMOS+UltraVISTA catalogue, but its CFHT/i-band magnitude (24.27 ± 0.49 mag) with respect to the CFHT/K-band and UltraVISTA Ks-band magnitudes (23.56 ± 0.20 mag and 24.30 ± 0.23 mag, respectively) suggests a very blue colour index, hence it is unlikely related to the 1.3 mm feature9. There are also two negative sources (−4.3σ and 5.0σ) in the field, inside the PB FWHM. In the VLA 20 cm map, a source candidate with a peak surface brightness of 58.7 μJy beam-1 (only ~3.3σ) is visible close to our pointing centre (1″ away). However, no shorter-wavelength emission can be seen towards this source.

AzTEC26. Two 1.3 mm point sources were detected around the phase centre of this source, and the source we called AzTEC26a has an optical-to-NIR counterpart ( from the 1.3 mm peak). The weaker, 4.8σ source AzTEC26b has no multiwavelength counterparts. The VLA 20 cm image detects some emission near both sources, but the S/N ratio of the radio detection is less than 2.

For AzTEC26a, we derived a photo-z value of . Although the Subaru/i+-CFHT/K colour of this source, 1.06, is relatively blue, the radio/submm-based redshift of z> 1.87 supports the aforementioned zphot solution. The radio non-detection of AzTEC26b gives a lower redshift limit of z> 1.79.

AzTEC27. There is a hint of 20 cm emission associated with the elongated 1.3 mm source, although its 20 cm peak surface brightness is only 32.1 μJy beam-1 or about 2.5σ. No other wavelength counterparts are detected. At the projected distance of NE of the 1.3 mm peak position, there is the bright (V = 20.359) galaxy COSMOS J100039.47+024055.5 (Capak et al. 2007). The UltraVISTA catalogue gives a photo-z of about 0.25 for this galaxy (ID 303584; Ilbert et al. 2013). About SW of the 1.3 mm peak position, there is another galaxy, namely COSMOS J100039.07+024050.2 (Capak et al. 2007), whose photometric redshift is about unity (source 303782 in the UltraVISTA catalogue; Ilbert et al. 2013). As described in Appendix D, AzTEC27 appears to be subject to gravitational lensing by the foreground galaxies J100039.47 and J100039.07, and our lens model suggests that their combined lens effect amplifies the λobs = 1.3 mm flux density by a factor of ~2. At the distance of north from AzTEC27, there is a Herschel/HerMES 250 μm source (ID 1500; Oliver et al. 2012). Based on its radio dimness, we derived a redshift of z> 4.17 for AzTEC27, which makes this source potentially the highest-redshift SMG among AzTEC1630.

AzTEC28. This clearly detected 1.3 mm source, SE of the phase centre, has no counterparts at other wavelengths shown in Fig. A.1. Casey et al. (2013) detected AzTEC28 with SCUBA-2 (their source SMMJ100004.5+023042 or 450.20), with the 450 μm peak lying about NW of the PdBI peak position. The deboosted flux density at 450 μm was reported to be 19.11 ± 5.91 mJy. The optical counterpart (at ) reported by Casey et al. (2013) lies north of the PdBI 1.3 mm peak, hence is unrelated to the SMG. We also note that the ASTE-detected source AzTEC/C150 from Aretxaga et al. (2011) lies NE of our PdBI 1.3 mm source – still within the 34″ beam of ASTE/AzTEC at 1.1 mm. About NE of AzTEC28, there is the Herschel 250 μm source 4388 from the HerMES survey (Oliver et al. 2012).

AzTEC29. There are no clear signatures of PdBI 1.3 mm emission inside the PB. The source candidate AzTEC29a (4.7σ) lies at the border of the PB and the strong source AzTEC29b (7.3σ) at the northern edge of the map, is outside the PB. The latter could be associated ( offset) with the source 1685295 from the COSMOS+UltraVISTA catalogue. Two negative sources of −4.3σ and −5.4σ were detected in the field outside the PB FWHM. Hence, AzTEC29a, which does not show up at other wavelengths, could be spurious. The VLA 20 cm image of the source region detects emission in its south-western corner (peak surface brightness of 50.5 μJy beam-1 or ~3.4σ). The Spitzer/MIPS 24 μm emission near the field centre can be associated with the galaxy COSMOS J100026.79+023749.4 (Capak et al. 2007) or 289240 in the UltraVISTA catalogue (zphot ≃ 0.58; Ilbert et al. 2013).

For AzTEC29a, we derived a radio/submm-based redshift of z> 2.96. The photo-z of derived for AzTEC29b (Subaru/i+-CFHT/K = 2.67) is much lower than the unrealistically high value z> 7.25 derived from the radio/submm flux-density ratio, hence the photo-z value is adopted.

AzTEC30. A 4.6σ candidate point source with no counterparts is detected in this field. There is a Spitzer/MIPS 24 μm source in the NW part of the field, SE of our tentative source, that is also visible at the IRAC wavelengths but has no catalogue identification. The radio non-detection of AzTEC30 gives a redshift of z> 2.51.

Finally, we note that none of our sources are detected in X-rays, implying that none host a prominent AGN. Cross-correlation with the Chandra-COSMOS Bright Source Catalogue v2.1 and COSMOS XMM Point-like Source Catalogue v2.0 revealed that the nearest X-source to any of our SMGs is XMMU J100002.8+024635, lying south of the PdBI phase centre towards AzTEC21. However, it is possible that some of the studied SMGs host an extremely Compton-thick AGN (i.e. with obscuration due to high column densities of dust) that remains undetected in the existing X-ray images.

Appendix D: Gravitational lens modelling of AzTEC27

There are two foreground galaxies seen in projection close to the SMG AzTEC27 (see Fig. D.1). The north-eastern galaxy, at from the 1.3 mm peak position of AzTEC27, is COSMOS J100039.47+024055.5. The photometric redshift and stellar mass of J100039.47 are zphot ≃ 0.25 and log (M/ M) = 10.084 (Ilbert et al. 2013). On the south-western side, at from AzTEC27, the foreground galaxy is COSMOS J100039.07+024050.2 at zphot = 0.998 with a stellar mass of log (M/ M) = 10.713. AzTEC27 is potentially subject to gravitational lensing by these two intervening galaxies, and therefore to better understand its intrinsic physical properties requires a lens model.

thumbnail Fig. D.1

HST/ACS I-band (λcentral = 8333 Å) image towards AzTEC27 (linear colour scale), overlaid with white contours showing the PdBI 1.3 mm emission. The positive contour levels are as in Fig. 1. The green dashed circles of radius (20.31 physical kpc at z = 0.25) and (22.26 physical kpc at z = 0.998) are centred on the galaxies COSMOS J100039.47+024055.5 and J100039.07+024050.2, respectively. The synthesised PdBI beam (, PA ) is shown in the bottom left corner.

Open with DEXTER

Table E.1

Slit parameters of the Keck II/DEIMOS spectral line observations.

To estimate the strength of the lensing effect for AzTEC27, we carried out a gravitational lens modelling using the publicly available Python software called uvmcmcfit10 that will be described in detail by Bussmann et al. (in prep.). The code is a modified version of the one used in the papers by Bussmann et al. (2012, 2013) and uses the visibilities to determine the goodness of fit. To sample the posterior probability density function of our model parameters, we used the MCMC sampling code emcee (Foreman-Mackey et al. 2013).

The lensed background source in the model is assumed to be an elliptical Gaussian source, described by the following parameters: the total intrinsic (unlensed) flux density (Sin), the effective radius (, where a and b are the semi-major and semi-minor axes), the projected angular offset from the model image centroid, the axial ratio (b/a), and the position angle (PA measured E of N). The lens is assumed to be a singular isothermal ellipsoid (SIE), parameterised by the angular Einstein radius (θE), the angular offset from the model image centroid, the axial ratio, and the PA The magnification factor is then computed as μ = Sout/Sin, where Sout is the source’s total lensed flux density in the best-fit model.

The modelling was performed assuming three different scenarios: i) J100039.07 at zphot = 0.998 is acting as a lens; ii) J100039.47 at zphot ≃ 0.25 is responsible for lensing; and iii) both the above intervening galaxies act as lenses. The magnification factor in these three cases was found to be μ = 1.36 ± 0.11, 1.17 ± 0.04, and 2.04 ± 0.16, respectively. Both the galaxies J100039.07 and J100039.47 are therefore causing only a weak lensing effect, the former one, having a ~4.3 times higher stellar mass and lying at a higher redshift (i.e. closer to AzTEC27) than the latter, being the slightly more stronger lens. The weak lensing is consistent with the fact that we see only a single image of AzTEC27 in the PdBI 1.3 mm map. In the present paper we assume the two-lens system and adopt the value μ = 2.04 ± 0.16 for AzTEC27.

Appendix E: The DEIMOS spectrograph slit parameters

In Table E.1 we list the central coordinates, sizes (length and width), and position angles of the slits used for the Keck/DEIMOS spectral line observations towards AzTEC2, 5, 6, 9, 10, 13, and 17a. The DEIMOS slits are also illustrated in Fig. E.1 where we show the UltraVISTA Y-band NIR images towards the above sources.

thumbnail Fig. E.1

UltraVISTA Y-band (λeff = 1.02μm) images towards AzTEC2, 5, 6, 9, 10, 13, and 17a shown in linear scale. All the images are 15″ on a side. The white rectangles indicate the DEIMOS slit positions, sizes, and orientations (see Table E.1). We note that all the slits were aligned horizontally along the east-west direction. The central position of the slit is marked with a white plus sign, while the green circle of radius shows the SMA 890 μm peak position (Younger et al. 2007, 2009) except in the case of AzTEC17a where it marks the PdBI 1.3 mm peak position. The red circle in the AzTEC6 panel represents the optical galaxy discussed by Koprowski et al. (2014); see Appendix B for details.

Open with DEXTER

© ESO, 2015

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.