EDP Sciences
Issue A&A
Volume 518, July-August 2010
Herschel: the first science highlights
Article Number L14
Number of page(s) 5
Section Letters
DOI http://dx.doi.org/10.1051/0004-6361/201014681
Published online 16 July 2010

A&A 518, L14 (2010)

Herschel: the first science highlights

LETTER TO THE EDITOR

Deep Herschel view of obscured star formation in the Bullet cluster[*]

T. D. Rawle1 - S. M. Chung2 - D. Fadda3 - M. Rex1 - E. Egami1 - P. G. Pérez-González4,1 - B. Altieri5 - A. W. Blain6 - C. R. Bridge6 - A. K. Fiedler1 - A. H. Gonzalez2 - M. J. Pereira1 - J. Richard7 - I. Smail7 - I. Valtchanov5 - M. Zemcov6,8 - P. N. Appleton3 - J. J. Bock6,8 - F. Boone9,11 - B. Clement10 - F. Combes11 - C. D. Dowell6,8 - M. Dessauges-Zavadsky12 - O. Ilbert10 - R. J. Ivison13,14 - M. Jauzac10 - J.-P. Kneib10 - D. Lutz15 - R. Pelló9 - G. H. Rieke1 - G. Rodighiero16 - D. Schaerer12,9 - G. P. Smith17 - G. L. Walth1 - P. van der Werf18 - M. W. Werner8

1 - Steward Observatory, University of Arizona, 933 N. Cherry Ave, Tucson, AZ 85721, USA
2 - Department of Astronomy, University of Florida, Gainesville, 32611-2055, USA
3 - NASA Herschel Science Center, California Institute of Technology, MS 100-22, Pasadena, CA 91125, USA
4 - Departamento de Astrofísica, Facultad de CC. Físicas, Universidad Complutense de Madrid, 28040 Madrid, Spain
5 - Herschel Science Centre, ESAC, ESA, PO Box 78, Villanueva de la Cañada, 28691 Madrid, Spain
6 - California Institute of Technology, Pasadena, CA 91125, USA
7 - Institute for Computational Cosmology, Department of Physics, Durham University, South Road, Durham DH1 3LE, UK
8 - Jet Propulsion Laboratory, Pasadena, CA 91109, USA
9 - Laboratoire d'Astrophysique de Toulouse-Tarbes, Université de Toulouse, CNRS, 14 Av. Edouard Belin, 31400 Toulouse, France
10 - Laboratoire d'Astrophysique de Marseille, CNRS - Université Aix-Marseille, 38 rue Frédéric Joliot-Curie, 13388 Marseille Cedex 13, France
11 - Observatoire de Paris, LERMA, 61 Av. de l'Observatoire, 75014 Paris, France
12 - Geneva Observatory, University of Geneva, 51 Ch. des Maillettes, 1290 Versoix, Switzerland
13 - UK Astronomy Technology Centre, Science and Technology Facilities Council, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ, UK
14 - Institute for Astronomy, University of Edinburgh, Blackford Hill, Edinburgh EH9 3HJ, UK
15 - Max-Planck-Institut für extraterrestrische Physik, Postfach 1312, 85741 Garching, Germany
16 - Department of Astronomy, University of Padova, Vicolo dell'Osservatorio 3, 35122 Padova, Italy
17 - School of Physics and Astronomy, University of Birmingham, Edgbaston, Birmingham, B15 2TT, UK
18 - Sterrewacht Leiden, Leiden University, PO Box 9513, 2300 RA Leiden, The Netherlands

Received 31 March 2010 / Accepted 11 May 2010

Abstract
We use deep, five band (100-500 $\mu $m) data from the Herschel Lensing Survey (HLS) to fully constrain the obscured star formation rate, ${\it SFR}_{\rm FIR}$, of galaxies in the Bullet cluster (z = 0.296), and a smaller background system (z = 0.35) in the same field. Herschel detects 23 Bullet cluster members with a total ${\it SFR}_{\rm FIR} = 144 \pm 14~M_{\odot}$ yr-1. On average, the background system contains brighter far-infrared (FIR) galaxies, with $\sim$50% higher ${\it SFR}_{\rm FIR}$ (21 galaxies; $207 \pm 9~ M_{\odot}~\rm yr^{-1}$). SFRs extrapolated from 24 $\mu $m flux via recent templates ( ${\it SFR}_{24~{\mu}{\rm m}}$) agree well with ${\it SFR}_{\rm FIR}$ for $\sim$60% of the cluster galaxies. In the remaining $\sim$40%, ${\it SFR}_{24~{\mu}{\rm m}}$ underestimates ${\it SFR}_{\rm FIR}$ due to a significant excess in observed $S\!_{100}/S\!_{24}$ (rest frame $S\!_{75}/S\!_{18}$) compared to templates of the same FIR luminosity.

Key words: galaxies: clusters: individual: Bullet cluster - galaxies: star formation - infrared: galaxies - submillimeter: galaxies

1 Introduction

In the last decade many studies have attempted to quantify the star formation rate (SFR) within cluster galaxies. Ultraviolet and optical observations have successfully identified trends between unobscured star formation and local environment, suggesting that star formation in cluster core galaxies is generally more quenched (e.g. Porter & Raychaudhury 2007; Kodama et al. 2004). However, star formation can be obscured by dust, which re-emits stellar light in the far-infrared (FIR), peaking at a rest frame $\lambda_0 \sim 100~\mu$m. Mid-infrared surveys (e.g. Fadda et al. 2008; Metcalfe et al. 2005; Geach et al. 2006) have explored obscured star formation by estimating total FIR luminosity from template spectra. These templates are often based on small numbers of well constrained local galaxies, e.g. Rieke et al. (2009).

The PACS (Poglitsch et al. 2010) and SPIRE (Griffin et al. 2010) instruments, onboard the ESA Herschel Space Observatory (Pilbratt et al. 2010), enable unprecedented multi-band coverage of the FIR. The Herschel Lensing Survey (HLS; PI: E Egami) consists of 5-band observations (100-500 $\mu $m) of 40 nearby clusters ( $z \sim 0.2{-}0.4$). Nominally devised to exploit the gravitational lensing effect of massive clusters to observe high redshift galaxies (see Egami et al. 2010, for details on survey design), a useful by-product is deep FIR observations of the clusters themselves. At these redshifts, Herschel photometry spans the peak of the dust component, allowing an accurate constraint of far infrared luminosity, $L_{\rm FIR}$, and hence obscured SFR.

During the Herschel science demonstration phase, HLS observed the Bullet cluster (1E0657-56; z = 0.296). The reason for this choice was two-fold. First, previous studies report bright submillimeter galaxies in the background (e.g. Rex et al. 2009), with HLS analysis presented in Rex et al. (2010). Second, the Bullet cluster is a recent collision of two clusters (Markevitch et al. 2002), offering a unique laboratory for the study of star formation within a dynamic environment. The sub-cluster has conveniently fallen through the main cluster perpendicular to the line of sight ( ${<} 8^\circ$ from the sky plane; Markevitch et al. 2004). Analysis of X-ray emission shows that a supersonic bow shock precedes the hot gas, while the weak lensing mass profile indicates that this X-ray bright component lags behind the sub-cluster galaxies due to ram pressure (Barrena et al. 2002; Markevitch et al. 2002). A recent mid-infrared study by Chung et al. (2009) concluded that ram pressure from the merger event had no significant impact on the star formation rates of nearby galaxies. We can re-evaluate these previous studies by using Herschel data to constrain $L_{\rm FIR}$ directly. In this letter, we present an exploration of obscured star formation in this cluster environment.

\begin{figure}
\par\includegraphics[angle=270,width=7.5cm,clip]{14681fg1.eps}\vspace*{-2mm}\vspace*{-3mm}
\end{figure} Figure 1:

Distribution of spectroscopic redshifts ( 0.27 < z < 0.37) for galaxies within the Bullet cluster field (outline). Herschel detected galaxies are also shown (filled). In addition to the Bullet cluster (z = 0.296), there is a background system at z = 0.350. Dotted lines show our membership limits of 3000 km s-1 and 2000 km s-1 respectively.

Open with DEXTER

2 Observations

2.1 Photometric data

Five band Herschel imaging was obtained using two instruments: PACS (100 160 $\mu $m) covering approximately 8$^\prime$ $\times$ 8$^\prime$ and SPIRE (250 350 500 $\mu $m) with a wider $\sim$17$^\prime$ $\times$ 17$^\prime$ field. We also use Magellan IMACS optical, Spitzer IRAC and MIPS 24 $\mu $m maps with similar coverage to SPIRE, and high resolution HST ACS images of the central 4$^\prime$ $\times$ 4$^\prime$. Egami et al. (2010) provides details of all the data, and presents Herschel FIR maps.

The deep SPIRE maps have detection limits well below the instrument confusion limits. To avoid compiling sourcelists from confused maps, Herschel fluxes are measured at all Spitzer MIPS 24 $\mu $m source positions. For a typical galaxy SED at $z \sim 0.3$, the 24 $\mu $m map is much deeper than SPIRE, so even with a relatively high ${\it S/N} > 10$ cut (flux limit $\sim$100 $\mu $Jy), we can assume the inclusion of all sources contributing significant FIR flux. The use of mid-infrared source positions has the added advantage of decreasing the significance of flux boosting, which has not been addressed in this study.

Photometric analysis followed the same procedure in all 5 Herschel bands. An average PSF, measured from the brightest unblended sources in the image, was simultaneously fit to all positions in the 24 $\mu $m catalogue (without re-centering) using D AOPHOT A LLSTAR. At the longer SPIRE wavelengths, there is a higher probability of more than one 24 $\mu $m source falling within the FIR beam. In these instances, the objects are grouped together at the 24 $\mu $m ${\it S/N}$-weighted mean position, treated as a single source, and flagged (see following sub-section). For more details on the photometry technique see Rex et al. (2010).

2.2 Spectroscopy and sample selection

The spectroscopic redshift catalogue combines observations from three campaigns: Magellan IMACS multi-slit (856 targets, Chung et al. 2010, Chung et al. in prep.), CTIO Hydra multi-fiber (202, Fadda et al. in prep.) and VLT FORS multi-slit (14, J. Richard, private communication). Egami et al. (2010) provides further details. The merged catalogue comprises 929 sources within the SPIRE field.

Figure 1 presents the distribution of spectroscopic redshifts for the range 0.27 < z < 0.37. An important aspect of this study is confidence in the cluster membership of galaxies. The Bullet cluster distribution peaks at z = 0.296, and we limit membership to $\pm$3000 km s-1 ( 0.286 < z < 0.306). In addition, this study also analyzes galaxies from a system at z = 0.350 in the same field, limiting membership to $\pm$2000 km s-1 ( 0.343 < z < 0.357). The systems have 362 and 95 known members respectively.

The sample for this analysis consists of MIPS 24 $\mu $m sources with spectroscopically confirmed cluster redshifts. These two catalogues were merged by identifying the closest 24 $\mu $m source, within the rms pointing error of MIPS (1.4 $^{\prime \prime }$), to the spectroscopic position. For sample members grouped during the FIR photometry (previous sub-section), we examined the optical and IRAC colours of each group member, identifying the likely source of the mid- and far-IR flux. In cases where the sample member was not considered to be the source, or when the situation was unclear, the object was rejected from the sample.

In the final sample, there are 47 confirmed Bullet cluster members, and an additional 28 sources in the z = 0.35 system. Of these, 23 and 21 galaxies respectively are detected in the Herschel bands, highlighted by the filled distribution in Fig. 1. The background system has a much higher fraction of Herschel detections than the Bullet cluster (75% of 24 $\mu $m sources, compared to 50%).

\begin{figure}
\par\includegraphics[angle=270,width=15cm,clip]{14681fg2.eps}
\end{figure} Figure 2:

Photometric data for five of the most FIR luminous galaxies in the sample. Blue = PACS; red = SPIRE; grey = IRAC/MIPS. Flux densities are as observed (i.e. not de-magnified). Redshift and, for background system galaxies, magnification factor, $\mu $, are displayed at the top of each panel. $L_{\rm FIR}$ and ${\it SFR}_{\rm FIR}$ derived from the best fit blackbody (black line) are also shown and have been de-magnified where necessary.

Open with DEXTER

3 Results and discussion

3.1 Far-infrared (FIR) spectral energy distributions

For each source, the FIR spectral energy distribution (SED) is fit to all available Herschel data points, taking into account the upper limits for non-detections. The dust component is modeled by a modified, single-temperature, blackbody

\begin{displaymath}S\!_{\nu} = N({\nu}/{\nu_0})^{\beta}B_{\nu}(T)
\end{displaymath} (1)

where $S\!_{\nu}$ is flux density, $\beta$ is dust emissivity index (fixed at 1.5; using $\beta = 2.0$ would vary $L_{\rm FIR}$ by <15% on average) and $B_{\nu}(T)$ is the Planck blackbody radiation function for a source at temperature T. The shape of this optically thin (rather than thick) blackbody imitates the inclusion of a secondary (warm) dust component. As we are concerned only with $L_{\rm FIR}$ and SFR, the parameterization of the data is the most important aspect, and T is used purely as a fit parameter. Galaxies within the PACS field have well constrained fits, and T is allowed to float freely. For those without PACS data ($\sim$40%), $T_{\rm0}$ has been forced to a narrow range centered on the mean value from the constrained SEDs ($30 \pm 1$ K). Forcing $T_{\rm0}$ to a similarly narrow range about values $\sim$1$\sigma$ from the constrained mean, varies $L_{\rm FIR}$ by <25%. Bias in $L_{\rm FIR}$ due to model priors is comparable in scale to systematics from instrument calibration.

$L_{\rm FIR}$ is integrated over (rest frame) $\lambda_0 = 8{-}1000~\mu$m from which ${\it SFR}_{\rm FIR}$ is derived using the Kennicutt (1998) relation. As an illustration, Fig. 2 displays the FIR SED fits for five of the most luminous galaxies in the sample. These simple fits may underestimate $L_{\rm FIR}$ by up to a factor of 1.8 (Rex et al. 2010), as they lack a mid-infrared component. Future analysis will fully account for additional components. For the purposes of this study, a blackbody fit is sufficient. The luminosities of galaxies in the background system have been de-magnified using the Bullet cluster lensing model of Paraficz et al. (in prep.). The remaining figures in this paper present de-magnified values.

\begin{figure}
\par\includegraphics[angle=270,width=8.5cm,clip]{14681fg3.eps}
\vspace*{-3mm}
\end{figure} Figure 3:

Smoothed density maps for IMACS B-band flux ( left panels), 24 $\mu $m flux ( central) and SFR calculated from Herschel data ( right). The sources are binned by confirmed system membership, with the number of contributing galaxies displayed in the upper-right of each panel: upper row for Bullet cluster (over-plot in orange by the weak lens mass map); lower row for z = 0.35 system. All maps are Gaussian smoothed to the SPIRE 250 $\mu $m beam size (18 $^{\prime \prime }$  FWHM).

Open with DEXTER

3.2 Star formation rates within the systems

The system at z = 0.35 contains IR galaxies brighter than those in the Bullet cluster, with three galaxies meeting the LIRG criterion [ $\log(L_{\rm FIR}/L_{\odot}) \ga 11.0$] and an additional two within 1$\sigma$. A further 10 members have $\log(L_{\rm FIR}/L_{\odot}) > 10.5$. In contrast, the Bullet cluster contains two LIRGs, and only six other galaxies brighter than $\log(L_{\rm FIR}/L_{\odot}) = 10.5$.

The total star formation rate of the 23 Bullet cluster galaxies is $144 \pm 14~ M_{\odot}$ yr-1. The 21 galaxies in the background system are, on average, $\sim$50% more active, with a total ${\it SFR} = 207 \pm 9~ M_{\odot}$ yr-1. Only five of these galaxies have been magnified by more than 20%, and the minimum detected SFRs are similar in each system. Therefore, it is unlikely that the higher total SFR is due to a decreased lower limit caused by magnification. The difference is likely to reflect the mass of the systems, although the lower SFR in the Bullet cluster may indicate that cluster-cluster mergers are not important for triggering FIR starbursts.

Figure 3 displays the spatial distribution of the Herschel-derived SFR for the two systems. Flux densities in optical B-band and 24 $\mu $m are shown for comparison. An initial examination suggests that the Bullet cluster exhibits a radial trend in ${\it SFR}_{\rm FIR}$ (lacking significant FIR detection towards the centre), reminiscent of that found in other contemporary studies (Pereira et al. 2010; Braglia et al. 2010). The gradient in the Bullet cluster SFR is examined in detail in Chung et al. (in prep.).

\begin{figure}
\par\includegraphics[angle=270,width=68mm,clip]{14681fg4.eps}\vspace*{-3mm}
\end{figure} Figure 4:

Ratio of ${\it SFR}_{\rm FIR}$ (from blackbody fit) to ${\it SFR}_{24~{\mu}{\rm m}}$ (via Rieke et al. 2009) versus the flux ratio $S\!_{100}/S\!_{24}$. For galaxies outside the PACS field, 100 $\mu $m is predicted from the blackbody fit. Dashed line is equality and shaded region indicates 50% difference in the SFRs. All galaxies with under-predicted ${\it SFR}_{24~{\mu}{\rm m}}$ have redder $S\!_{100}/S\!_{24}$.

Open with DEXTER

The IR and optical flux of the background system trace similar distributions, whereas in the Bullet cluster, the B-band flux is more centrally concentrated, away from the IR sources. This may indicate a different trend in dust retention for the two systems. While the 24 $\mu $m and Herschel SFR density maps generally trace the same distribution, there are significant outliers: bright 24 $\mu $m sources with relatively lower SFRs, and vice versa. In the following section, we compare the SFR estimated from 24 $\mu $m (through the Rieke et al. 2009, templates) to the ${\it SFR}_{\rm FIR}$.

3.3 24 $\mu $m as a L $_\mathsf{FIR}$ predictor in nearby clusters

The mid-infrared bands, e.g. MIPS 24 $\mu $m, are often used to estimate far infrared luminosity, $L_{\rm FIR}$, and hence obscured SFR, via template FIR SEDs such as Rieke et al. (2009). Those authors provide a simple formula (their Eq. (14)) to convert 24 $\mu $m flux directly to SFR. The templates are based on local (U)LIRGs ($z \la 0.1$), and at high redshift, may not be valid. Here, we test the template accuracy for cluster galaxies at z = 0.3.

In Sect. 3.2 (Fig. 3), we suggested that while 24 $\mu $m flux and ${\it SFR}_{\rm FIR}$ follow the same general distribution, they are not perfectly correlated. A direct comparison of ${\it SFR}_{\rm FIR}$ to ${\it SFR}_{24~{\mu}{\rm m}}$ (Fig. 4; plotted against the dust-peak-mid-IR flux ratio) leads to the same conclusion. For $\sim$60% of galaxies, the two SFRs agree well. However, there are several galaxies ($\sim$30%) that have severely underestimated ${\it SFR}_{24~{\mu}{\rm m}}$, and these also display systematically redder $S\!_{100}/S\!_{24}$. If ${\it SFR}_{\rm FIR}$ is underestimated by the simple blackbody fits (Sect. 3.1), the ${\it SFR}_{24~{\mu}{\rm m}}$ predictions are correspondingly worse.

Are the under-predicted ${\it SFR}_{24~{\mu}{\rm m}}$ caused by the redder $S\!_{100}/S\!_{24}$ colours? Figure 5 examines the Rieke et al. templates more closely, comparing them to the Herschel fluxes. For templates spanning the $L_{\rm FIR}$ range of the observations, the agreement is good for $\lambda_{\rm0} \ga 200~\mu$m. However, at 100 $\mu $m there are 8 significant outliers; we define 100 $\mu $m excess galaxies as those with $S\!_{100}/S\!_{24}$ (rest frame $S\!_{75}/S\!_{18})> 30$ as the templates predict $S\!_{100}/S\!_{24} \la 20$. Only templates with very high luminosities, i.e. $\log(L_{\rm FIR}/L_{\odot}) \ge 12$, match the observed $S\!_{100}/S\!_{24}$, but even the brightest sample galaxy has only $\log(L_{\rm FIR}/L_{\odot}) \sim 11.5$, while most are $\log(L_{\rm FIR}/L_{\odot}) < 11$. Although high $L_{\rm FIR}$ templates have $S\!_{100}/S\!_{24} \ga 30$, their lower peak wavelength leads to an under-prediction at $\lambda \ga 200~\mu$m in at least three observed SEDs. We also compare to the least active Dale & Helou (2002) FIR templates ( $\alpha = 1.8{-}2.5$). The locus of these are substantially similar to the low luminosity Rieke et al. templates and thus also only under-predict $S\!_{100}/S\!_{24}$. We stress that, unlike $L_{\rm FIR}$ and ${\it SFR}_{\rm FIR}$, the presence of a 100 $\mu $m excess is independent of the blackbody fits and the systematic uncertainties therein.

\begin{figure}
\par\includegraphics[angle=270,width=70mm,clip]{14681fg5.eps}\vspace*{-2mm}
\vspace*{1.7mm}
\end{figure} Figure 5:

Observed Herschel fluxes normalized by 24 $\mu $m for galaxies in both systems. Symbols as described at top-left. Rieke et al. (2009) average templates for $9.75 \le \log(L_{\rm FIR}) \le 11.0$ (red-pink) plus one example high- $L_{\rm FIR}$ template (cyan). Locus of low-activity templates from Dale & Helou (2002) ( $\alpha = 1.8{-}2.5$) is shaded green. All templates are normalized at $\lambda_{\rm obs} = 24~\mu$m ( $\lambda_{\rm 0} = 18~\mu$m). Templates in $L_{\rm FIR}$ range of the observations under-predict $S\!_{100}/S\!_{24}$ for 40% of sources. High $L_{\rm FIR}$ templates do not match the shape at $\lambda \ge 200~\mu$m.

Open with DEXTER

Galaxies with a 100 $\mu $m excess account for $\sim$40% of cluster members detected with PACS, and cover the entire range of $L_{\rm FIR}$ sampled. Above a nominal luminosity limit of 10 $^{10}~L_{\odot}$, 55% of Bullet cluster galaxies have the 100 $\mu $m excess. The fraction in the background system is lower at 36%. This may indicate a trend with environment, or could be due to the off-centre view of the latter system (i.e. a potential radial trend). High resolution HST imaging covers five of the eight 100 $\mu $m excess galaxies (Fig. 6). Despite the small number, the galaxies span a broad range of types and morphologies. Further examples are required for a firm conclusion, but these suggest that the 100 $\mu $m excess is not due to a single population of galaxies.

The $S\!_{100}/S\!_{24}$ colours alone may have led to the conclusion that the 100 $\mu $m excess was due to galaxies with generally colder dust. However, fits to the combined HLS PACS+SPIRE photometry suggest that this is not the case. Rather, the excess may be due to an additional warm dust component or active galactic nuclei (AGN) which are not considered in the templates. Using a simple power law to parameterize flux in the range 24-100 $\mu $m, we estimate the AGN contribution to total bolometric luminosity via the $S\!_{60}/S\!_{25}$ indicator for ULIRGs (Veilleux et al. 2009, Fig. 36). None of the 100 $\mu $m excess galaxies have predicted AGN fractions >30%. However, we may be under-predicting the contribution if the mid-IR SED steepens beyond 60 $\mu $m, or if the indicator breaks down for galaxies in this luminosity range.

Herschel PACS observations of $z \sim 0.2$ LoCuSS clusters (without the advantage of complementary SPIRE data), display a similar fraction of 100 $\mu $m excess galaxies (Pereira et al. 2010; Smith et al. 2010). However, the high redshift field sample from the HLS Bullet cluster observations (Rex et al. 2010) lacks a comparable excess at $\lambda_{\rm0} \approx 75~\mu$m. These results suggest that the effect could be either redshift dependent or cluster-specific. HLSis well placed for further analysis of the $S\!_{100}/S\!_{24}$ phenomenon, as the combined PACS+SPIRE data ensures that both the excess and entire FIR component can be constrained simultaneously.

\begin{figure}
\par\includegraphics[width=90mm,clip]{14681fg6.eps}\vspace*{-2mm}
\vspace*{-3mm}
\end{figure} Figure 6:

HST ACS thumbnails of five 100 $\mu $m excess galaxies ( $S\!_{100}/S\!_{24} > 30$; increasing from left) do not suggest a single source population.

Open with DEXTER

4 Conclusions

Using deep Herschel observations (100-500 $\mu $m) to fully constrain the FIR component, we derive obscured SFRs for galaxies in the Bullet cluster (z = 0.296), and a background system (z = 0.35) in the same field. Herschel detects 23 Bullet cluster members, with a total ${\it SFR}_{\rm FIR} = 144 \pm 14~M_{\odot}$ yr-1, while the background system contains 21 detections but $\sim$50% higher SFR ( $207 \pm 9~ M_{\odot}$ yr-1). The relative distributions of ${\it SFR}_{\rm FIR}$ and optical flux suggest a difference in dust retention between the two systems. For $\sim$60% of galaxies, ${\it SFR}_{\rm FIR}$ agrees well with estimated SFRs from 24 $\mu $m flux via recent templates. However, the remaining galaxies display a significant excess at 100 $\mu $m ( $\lambda_{\rm0} \approx 75$ $\mu $m) compared to templates, which causes an under-prediction in ${\it SFR}_{24~{\mu}{\rm m}}$. We note that such an excess is not found in the high redshift, field sample (Rex et al. 2010). Future studies will exploit the full range of 5-band Herschel cluster observations available in HLS, to form a more complete understanding of the environmental effect on obscured star formation rates, and explore the origin and dependencies of the 100 $\mu $m excess.

Acknowledgements
This work is based in part on observations made with Herschel, a European Space Agency Cornerstone Mission with significant participation by NASA. Support for this work was provided by NASA through an award issued by JPL/Caltech.

References

Footnotes

... cluster[*]
Herschel is an ESA space observatory with science instruments provided by European-led Principal Investigator consortia and with important participation from NASA. Data presented in this paper were analyzed using ``The Herschel Interactive Processing Environment (HIPE)'', a joint development by the Herschel Science Ground Segment Consortium, consisting of ESA, the NASA Herschel Science Center, and the HIFI, PACS and SPIRE consortia.

All Figures

  \begin{figure}
\par\includegraphics[angle=270,width=7.5cm,clip]{14681fg1.eps}\vspace*{-2mm}\vspace*{-3mm}
\end{figure} Figure 1:

Distribution of spectroscopic redshifts ( 0.27 < z < 0.37) for galaxies within the Bullet cluster field (outline). Herschel detected galaxies are also shown (filled). In addition to the Bullet cluster (z = 0.296), there is a background system at z = 0.350. Dotted lines show our membership limits of 3000 km s-1 and 2000 km s-1 respectively.

Open with DEXTER
In the text

  \begin{figure}
\par\includegraphics[angle=270,width=15cm,clip]{14681fg2.eps}
\end{figure} Figure 2:

Photometric data for five of the most FIR luminous galaxies in the sample. Blue = PACS; red = SPIRE; grey = IRAC/MIPS. Flux densities are as observed (i.e. not de-magnified). Redshift and, for background system galaxies, magnification factor, $\mu $, are displayed at the top of each panel. $L_{\rm FIR}$ and ${\it SFR}_{\rm FIR}$ derived from the best fit blackbody (black line) are also shown and have been de-magnified where necessary.

Open with DEXTER
In the text

  \begin{figure}
\par\includegraphics[angle=270,width=8.5cm,clip]{14681fg3.eps}
\vspace*{-3mm}
\end{figure} Figure 3:

Smoothed density maps for IMACS B-band flux ( left panels), 24 $\mu $m flux ( central) and SFR calculated from Herschel data ( right). The sources are binned by confirmed system membership, with the number of contributing galaxies displayed in the upper-right of each panel: upper row for Bullet cluster (over-plot in orange by the weak lens mass map); lower row for z = 0.35 system. All maps are Gaussian smoothed to the SPIRE 250 $\mu $m beam size (18 $^{\prime \prime }$  FWHM).

Open with DEXTER
In the text

  \begin{figure}
\par\includegraphics[angle=270,width=68mm,clip]{14681fg4.eps}\vspace*{-3mm}
\end{figure} Figure 4:

Ratio of ${\it SFR}_{\rm FIR}$ (from blackbody fit) to ${\it SFR}_{24~{\mu}{\rm m}}$ (via Rieke et al. 2009) versus the flux ratio $S\!_{100}/S\!_{24}$. For galaxies outside the PACS field, 100 $\mu $m is predicted from the blackbody fit. Dashed line is equality and shaded region indicates 50% difference in the SFRs. All galaxies with under-predicted ${\it SFR}_{24~{\mu}{\rm m}}$ have redder $S\!_{100}/S\!_{24}$.

Open with DEXTER
In the text

  \begin{figure}
\par\includegraphics[angle=270,width=70mm,clip]{14681fg5.eps}\vspace*{-2mm}
\vspace*{1.7mm}
\end{figure} Figure 5:

Observed Herschel fluxes normalized by 24 $\mu $m for galaxies in both systems. Symbols as described at top-left. Rieke et al. (2009) average templates for $9.75 \le \log(L_{\rm FIR}) \le 11.0$ (red-pink) plus one example high- $L_{\rm FIR}$ template (cyan). Locus of low-activity templates from Dale & Helou (2002) ( $\alpha = 1.8{-}2.5$) is shaded green. All templates are normalized at $\lambda_{\rm obs} = 24~\mu$m ( $\lambda_{\rm 0} = 18~\mu$m). Templates in $L_{\rm FIR}$ range of the observations under-predict $S\!_{100}/S\!_{24}$ for 40% of sources. High $L_{\rm FIR}$ templates do not match the shape at $\lambda \ge 200~\mu$m.

Open with DEXTER
In the text

  \begin{figure}
\par\includegraphics[width=90mm,clip]{14681fg6.eps}\vspace*{-2mm}
\vspace*{-3mm}
\end{figure} Figure 6:

HST ACS thumbnails of five 100 $\mu $m excess galaxies ( $S\!_{100}/S\!_{24} > 30$; increasing from left) do not suggest a single source population.

Open with DEXTER
In the text


Copyright ESO 2010

What is OpenURL?

The OpenURL standard is a protocol for transmission of metadata describing the resource that you wish to access. An OpenURL link contains article metadata and directs it to the OpenURL server of your choice. The OpenURL server can provide access to the resource and also offer complementary services (specific search engine, export of references...). The OpenURL link can be generated by different means.
  • If your librarian has set up your subscription with an OpenURL resolver, OpenURL links appear automatically on the abstract pages.
  • You can define your own OpenURL resolver with your EDPS Account. In this case your choice will be given priority over that of your library.
  • You can use an add-on for your browser (Firefox or I.E.) to display OpenURL links on a page (see http://www.openly.com/openurlref/). You should disable this module if you wish to use the OpenURL server that you or your library have defined.

Editor-in-Chief: T. Forveille
Letters Editor-in-Chief: J. Alves
Managing Editor: C. Bertout

ISSN: 0004-6361 ; e-ISSN: 1432-0746
Frequency: 12 volumes per year
Published by: EDP Sciences

Mirror sites: CDS | EDP Sciences
  RSS feeds
© The European Southern Observatory (ESO)