Near-UV Sources in the Hubble Ultra Deep Field: The Catalog
The catalog from the first high resolution U-band image of the Hubble Ultra Deep Field, taken with Hubble’s Wide Field Planetary Camera 2 through the F300W filter, is presented. We detect 96 U-band objects and compare and combine this catalog with a Great Observatories Origins Deep Survey (GOODS) B-selected catalog that provides B, V, i, and z photometry, spectral types, and photometric redshifts. We have also obtained Far-Ultraviolet (FUV, 1614 Å) data with Hubble’s Advanced Camera for Surveys Solar Blind Channel (ACS/SBC) and with Galaxy Evolution Explorer (GALEX). We detected 31 sources with ACS/SBC, 28 with GALEX/FUV, and 45 with GALEX/NUV. The methods of observations, image processing, object identification, catalog preparation, and catalog matching are presented.
The Hubble Ultra Deep Field (UDF) campaign (Beckwith et al. 2006) has produced the deepest optical images of our universe to date. The UDF was observed by Hubble in 412 orbits that were centered in a region of the Chandra Deep Field South (CDF-S) which was also the target of the Great Observatories Origins Deep Survey (GOODS, Giavalisco et al. 2004) known as the GOODS south or GOODS-S. The UDF used the same Advanced Camera for Surveys (ACS) filters as GOODS, F435W (B), F606W (V), F775W (i), and F850LP (z), but covered only one field of the 15 GOODS-S fields. The UDF reached approximately uniform limiting magnitudes m 29 for point sources, at least two magnitudes deeper than GOODS. Both campaigns, the UDF and GOODS, did not include deep imaging in the U bandpass. Taking U-band photometry is a time consuming task because longer integrations are required to achieve comparable depth to optical images. Only three HST U-band deep fields have been taken so far, the original Hubble Deep Field, the Hubble Deep Field South and the deepest U-band ever taken with Hubble which was part of the parallel campaign of the UDF and lies on the edge of the GOODS-S (Williams et al. 1996, Casertano et al. 2000, de Mello et al. 2006b). GOODS has only partial U-band coverage with HST obtained during parallel observations (de Mello et al. 2006a). Deep U-band ground-based images of the GOODS-S field, such as those taken with the CTIO 4m and ESO 2.2m available in the GOODS webpage555http://www.stsci.edu/science/goods/SupportObs/cdfsmosaic/, are included in the multiwavelength coverage of GOODS. Although ground-based observations can cover larger fields of view than Hubble’s cameras more efficiently, they do not possess the same angular resolution as space-based observations. Low-resolution ground-based images will blend together nearby detections leading to inaccurate photometric redshifts and morphological analysis. The U-band is a critical wavelength in studies at intermediate redshifts (z 2) since the rest-frame UV light is redshifted into the U bandpass. It is in the UV that short-lived, massive, O and B stars radiate most of their energy and therefore the U-band is necessary to probe the unobscured star-formation activity in galaxies at z 2.
After a redshift of z1.5-2, the star-formation rate of the universe began to steadily decline, decreasing by more than an order of magnitude (e.g. Hopkins & Beacom 2006, Wadadekar, Casertano & de Mello 2006). However, it is still an open question as to what population of objects contributes to the SFR density during the decline, and whether downsizing (the shift in star-formation being dominated from large to small mass galaxies as the universe aged) plays an important role in this era (Cowie et al. 1996, Savaglio et al. 2005, Mouri and Taniguchi 2006, Neistein, van den Bosch, & Dekel 2006). Therefore, studies that use U-band observations to better understand the nature of star-forming galaxies at intermediate-z can greatly contribute to connecting the early universe and the local universe.
In this paper we present the first HST targeted U-band image of the UDF. We present the U-band object catalog, and describe the methods used for observations, image processing, object identification, and catalog preparation.
The U-band observations were obtained with the HST Wide Field Planetary Camera 2 (WFPC2) in the Cycle 13 HST Treasury proposal (Teplitz, Program 10403). The UDF was imaged in the near UV (NUV) through the WFPC2/F300W filter ( = 2987Å, =740Å) in 12 HST orbits divided into 4 roll angles to compensate for the shape of the WFPC2 chevron and achieve uniform depth. On-chip binning (2x2) was applied during WFPC2 observations to reduce the effects of read-out noise, i.e. each Wide Field became 400 400 pixels. A total of 24 WFPC2 images were taken with individual exposure times of 1200s.
FUV imaging was also obtained in this proposal with HST/ACS Solar Blind Channel (SBC) camera at which time WFPC2 parallel observations were also made. The UDF ACS/SBC images were taken in 50 HST orbits using the long-pass quartz filter (F150LP) ( = 1614Å, and FWHM=177Å). The FUV was imaged in 25 pointings. Each pointing had a four point dither pattern with two 650s exposures at each dither position. The total exposure time per pointing was 5200s (Siana et al. 2007).
3 Image Processing
The WFPC2 images were retrieved from the HST archive for further processing. We combined the 24 WFPC2 images with the MultiDrizzle code in the PyDrizzle package (Koekemoer et al. 2002). These images were taken using a dithering technique that reduces effects of pixel-to-pixel errors and allows one to better remove hot pixels, bad columns, and charge traps from the image. Dithering also allows the recovery of information lost to undersampling by pixels that are not small compared to the point spread function (PSF). MultiDrizzle simplifies and automates the detection of cosmic-rays of these dithered observations. We used calibrated flat-fields from the HST pipeline and ran the MultiDrizzle script through the following steps. First, a static mask was created to identify bad pixels, then each image was sky-subtracted, shifts were determined from header coordinates for each image and were applied in drizzling each image separately onto registered output images. Next, a median image was created from these separate drizzled images and was blotted back to each original input image. Finally, the blotted images were used to compute cosmic ray masks, and the final drizzle combination was performed using these masks.
Prior to running MultiDrizzle the Planetary Camera (PC) data was removed from all images because its inclusion greatly increased the noise level of the output drizzled image. This was achieved by replacing the PC in each of the 24 images with hot pixels, forcing MultiDrizzle to automatically include the PC in the pixel mask for each individual image. We set the user-inputs to MultiDrizzle so the script would output separate science and weight images for these data that we have made available online at: http://goods.gsfc.nasa.gov/release/UDFF300W.
The drizzled WFPC2 image shown in Figure 1 has a total exposure time of 28,800s and a pixel scale of 0.1 arcsec. Due to the nature in which the 24 WFPC2 pointings were positioned, the final drizzled image does not have of a uniform depth. The majority of pointings overlap in the mid-upper region of the combined image. Consequently, the majority of UV sources are detected in this area of the UDF. Also, the upper-most section of the U-band image lies just outside the UDF footprint, and sources from this area are marked as such in the catalog. The reduction procedure used for the UDF ACS/SBC images is outlined in Teplitz et al. (2006).
4 Object Identification And Catalog Preparation
The catalog of U-band sources was produced using SourceExtractor version 2.5 (Bertin & Arnouts 1996, hereafter SE). Initially, we created both a low- (1.5 above background noise) and a high- (3 above background noise) catalog. The high- catalog contains all visually confirmed sources in the image, while the low- catalog may contain spurious detection. The difference between the SE parameters specified for these two catalogs was the detection threshold relative to the background RMS. For both catalogs, the minimum area of adjoining pixels for a detection was 15 pixels, and the minimum deblending parameter was set at 10%, except for a few cases where we set the deblending value in order to avoid multiple detections in one single object. This is particularly critical for source detection in U-band images since star-forming regions can appear as multiple clumps in one object. SE might detect those clumps as individual objects making false identifications.
The GAIN, MAGZEROPOINT, and SATURLEVEL parameters which are specific to the WFPC2 camera were set to 7 e/ADU, 20.77 mag, and 2 ADU/s, respectively. The weight map produced by the drizzling process was used by setting the weight map type to MAPWEIGHT. Photometric measurements of each source were calculated using SE’s automatic aperture magnitudes (MAGAUTO). MAGAUTO uses a Kron (1980) flexible elliptical aperture to measure the total magnitude of each source. Instead of using the classical aperture photomery with a fixed aperture radius, MAGAUTO has the advantage of limiting the background noise while detecting light from faint sources more effectively. The size of the background mesh which is subtracted from the photometry of each source (BACKSIZE) was set to 64 pixels, and its RMS value is used to calculate photometric errors. We cleaned the high- and low- catalogs removing sources with photometric errors 1.0.
The resulting U-band catalog includes all detections from the high- catalog, and the remaining objects visually confirmed from the low- catalog. The U-band 1.5 limiting magnitude measured within a 1 diameter aperture is 23.5mag (AB) (Figure 2).
4.1 Visual Identification Of U-Band Sources
We have visually checked each SE U-band detection in order to decide (i) if a single source detection is actually multiple sources, (ii) if multiple source detections are single sources, (iii) if a detection is too noisy, or (iv) if there are any faint UV sources which are not detected. When such cases occur, SE parameters can be adjusted to maximize U-band source detections in the UDF image, and non-detections can be omitted from the catalog.
We have also visually identified the U-band sources within a Hubble ACS/B-band image of the GOODS-S field that overlaps the UDF. The B-band sources had been cataloged by the GOODS team using SE and have matched aperture photometry in multiple ACS bands (V,i,z) (Dahlen et al. 2007). The B-band catalog also lists spectral types (see §5 for definition) and photometric redshifts with a typical GOODS accuracy of =0.8 (where z-z/(1+z)) (Dahlen et al. 2007) for each source. If a U-band detection could not be visually identified as one of the objects in the B-band catalog it was removed from the U-band catalog. We did this because the B-band data is much deeper (limiting 10 sensativity is 27.8; Giavalisco et al. 2004) than the U-band, and we would not expect to detect a source in the U-band without also seeing it in the B-band. During this cleaning the majority of spurious U-band detections located in the borders of the WFPC2 image were removed. We also discovered five B-band objects that corresponded to multiple detections in the U-band. This was a result of setting SE’s deblending parameters to a low value in an effort to detect as many U-band sources as possible. This parameter was adjusted in an additional SE run to obtain single detections of these sources. The final U-band catalog contains 96 objects.
4.2 Catalog Matching
We matched the final U-band catalog to the far-UV catalog of the UDF created from the ACS/SBC observations described in Section 2 (Siana et al. 2007). Each ACS/SBC source was matched to the nearest U-band object within a 2.5 radius. The typical difference between WFPC2/F300W coordinates (RA and Dec) and ACS/SBC coordinates of the same object is 0.1. Thirteen U-band sources do not have FUV detections because they are outside the ACS/SBC footprint. Any ACS/SBC sources with signal to noise 3 were not considered. In total, 31 of the 96 U-band objects have resolved matching ACS/SBC detections.
The UDF has also been observed with the Galaxy Evolution Explorer (GALEX) mission in the far and near ultraviolet (FUV: =1528Å =269Å NUV: =2271Å =616Å; GALEX field of view is 1.28 and 1.24 in FUV and NUV, and pixel scale is 1.5/pixel) and is publicly available in the GALEX Release 4 (GR4) at the Multimission Archive at STScI (MAST). These data are from two different surveys, the All Sky Survey (AIS, 4.3 FWHM) and the Deep Sky Survey (DIS, 5.3 FWHM). The 5 limiting magnitudes of the AIS data are 20.8 in the NUV and 19.9 in the FUV, and for the DIS data 24.4 in the NUV and 24.8 in the FUV (Morrissey et al. 2007).
We searched for all U-band objects in the GR4 and found 28 detected by GALEX/FUV and 45 in the NUV. Since GALEX resolution is significantly lower than Hubble’s (GOODS ACS image has 0.03/pixel) we have searched for objects where confusion might be problematic and one should use the GALEX data with caution. We found 22 objects where confusion was an important factor in the NUV and FUV and have flagged them in the catalog. In Figure 3 we show an example of a single GALEX NUV detection of at least four identified objects in the U and B-band.
5 The Catalog
Table 1 presents the U-band catalog of the UDF NUV sources.
Columns (1) and (2) are the GOODS World Coordinate System (WCS) Right Ascension and
Declination in degrees. Three objects with Chandra X-ray detections
(Koekemoer et al. 2004) are flagged as “a” 111Seven other Chandra X-ray sources found in the UDF were not detected in our U-band image..
Columns (3)-(8) are the U, FUV, B, V, i, and z magnitudes and photometric errors (MAGAUTO), respectively.
The B, V, i, and z magnitudes were obtained from the GOODS-S
B-selected catalog. Note that not all U-band detections have FUV photometry because they are either outside the ACS/SBC
footprint, are non-detections (S/N 3), or are not resolved
in the FUV imaging.
Columns (9) and (10) are GALEX NUV and FUV magnitudes from the GR4. Sources with confusion are flagged as “c” in column 1.
Columns (11) and (12) list photometric redshifts (z) for all objects, and available spectroscopic redshifts (z) for
21 objects. The z are from the GOODS collaboration (taken from the ESO/GOODS-S spectroscopy masters catalog666See http://www.eso.org/science/goods/spectroscopy/CDFSMastercat.),
and Figure 4 plots z as a function of z for these objects. Column (14)
are spectral types from the GOODS B-selected catalog based on spectral energy distributions from Coleman et al. (1980) and Kinney et al. (1996).
Type 1 galaxies are early-types (E, S0, Sa), type 2 are Sbc, type 3 are Scd, type 4 are irregulars, and types 5 and 6 are starbursts SB1 and SB2.
The SED templates of SB1s and SB2s are differentiated by the values of intrinsic color excess, E(B-V).
SB1 has E(B-V)0.10, and SB2 has 0.11E(B-V)0.21 (Kinney et al. 1996).
The U-band catalog, including U, FUV, B, and BVi postage stamp images of each source,
is available online at: http://goods.gsfc.nasa.gov/release/UDFF300W
- Bertin et al. (1996) Bertin, E., & Arnouts, S. 1996, A&AS, 117, 393
- Casertano et al. (2000) Casertano, S., et. al. 2000, AJ, 120, 2747
- Coleman et al. (1980) Coleman, G.D., Wu, C-C., & Wedman, D.W. 1980, ApJS, 43, 393
- Cowie et al. (1996) Cowie, L.L., Songalia, A., Hu, E.M., & Cohen, J.G. 1996, AJ, 112, 839
- Dahlen et al. (2007) Dahlen, T., Mobasher, B., Dickinson, M., Ferguson, H.C., Giavalisoc,M., Kretchmer, C., & Ravindranath, S. 2007, ApJ, 654, 172D
- de Mello et al. (2006a) de Mello, D.F., Dahlen, T., Gardner, J.P., & Grogin, N.A. 2006a, AJ, 132, 2014
- de Mello et al. (2006b) de Mello, D.F., Wadadekar, Y., Dahlen, T., Casertano, S., & Gardner, J.P. 2006b, AJ, 131, 216
- Giavalisco et al. (2004) Giavalisco, M., et al. 2004, ApJ, 600, L93
- Hopkins & Beacom (2006) Hopkins, A.M., & Beacom, J.F. 2006, ApJ, 651, 142
- Kinney et al. (1996) Kinney, A., Calzetti, D., Bohlin, R.C., McQuade, K., Storchi-Bergmann, T., & Schmitt, H. 1996, ApJ, 467, 38
- Koekemoer et al. (2002) Koekemoer, A. M., Fruchter, A. S., Hook, R. N., Hack, W. 2002, HSTC Conf., 337K
- Koekemoer et al. (2004) Koekemoer, A. M., et al. 2004, ApJ, 600, L123
- Kron (1980) Kron, R.G. 1980, ApJS, 43, 305
- Morrissey et al. (2007) Morrissey, P., et al. 2007, ApJS, 173, 682
- Mouri and Taniguchi (2006) Mouri, H., & Taniguchi, Y. 2006, A&A, 459, 371
- Neistein et al. (2006) Neistein, E., van den Bosch, F.C., & Dekel, A. 2006, MNRAS, 372, 933
- Savaglio et al. (2005) Savaglio, S., et al. 2005, ApJ, 635, 260S
- Siana et al. (2007) Siana, B., et al. 2007, ApJ, 668, 62S
- Teplitz et al. (2006) Teplitz, H.I., et al. 2006, AJ, 132, 853
- Wadadekar et al. (2006) Wadadekar, Y., Casertano, S., & de Mello, D.F. 2006, AJ, 132, 1023
- Williams et al. (1996) Williams, R. et al. 1996, AJ, 112, 1335
|53.1470947ccConfusion in GALEX image||-27.7784290||22.460.08||…||24.430.06||24.280.04||23.930.06||23.730.06||24.910.25||…||1.12||…||6.00|
|53.1478577ccConfusion in GALEX image||-27.7740345||20.580.03||…||22.590.02||22.330.01||21.760.02||21.300.01||23.030.07||…||1.02||…||4.00|
|53.1518784ccConfusion in GALEX image||-27.7754364||21.970.05||…||23.760.05||23.360.03||22.580.03||22.170.02||22.930.06||…||1.00||…||3.33|
|53.1518898ccConfusion in GALEX image||-27.7828751||22.730.04||…||24.440.04||24.140.03||23.450.03||23.230.03||24.110.13||…||0.71||…||4.00|
|53.1518974ccConfusion in GALEX image||-27.7819862||23.170.08||…||24.600.06||24.320.04||23.720.05||23.320.04||24.190.18||…||0.84||…||4.00|
|53.1520653ccConfusion in GALEX image||-27.7747822||21.650.03||…||22.640.02||22.150.01||21.370.01||21.160.01||22.930.06||…||0.70||…||3.67|
|53.1523628ccConfusion in GALEX image||-27.7779751||22.990.06||…||24.710.08||24.840.07||24.960.16||24.290.10||22.930.06||…||1.55||…||6.00|
|53.1528244ccConfusion in GALEX image||-27.7826958||23.570.06||28.010.26||25.520.10||25.210.07||24.610.09||24.250.08||24.110.13||…||0.70||…||6.00|
|53.1559105ccConfusion in GALEX image||-27.7948895||21.820.08||…||24.020.04||23.790.03||23.430.05||23.110.04||24.630.12||…||0.91||…||5.67|
|53.1567726ccConfusion in GALEX image||-27.7955532||22.170.07||…||23.790.04||23.310.02||22.660.03||22.080.02||23.340.15||…||1.09||1.0970||3.33|
|53.1615295ccConfusion in GALEX image||-27.7676716||22.950.10||26.580.15||24.960.07||24.450.04||23.780.05||23.700.05||21.860.00||…||0.69||…||4.00|
|53.1615944aaX-ray source (Koekemoer et al. 2004)||-27.7922535||20.890.02||22.650.02||21.810.01||21.370.01||21.020.01||20.770.01||21.980.04||23.080.09||0.42||…||5.33|
|53.1619873ccConfusion in GALEX image||-27.7925415||21.630.04||24.740.05||24.910.04||24.160.02||23.750.03||23.710.03||21.970.03||22.950.04||0.25||…||5.00|
|53.1619949ccConfusion in GALEX image||-27.7739410||21.870.06||23.950.04||23.580.03||22.830.02||22.620.02||22.530.03||22.390.04||23.120.05||0.26||…||5.67|
|53.1623802ccConfusion in GALEX image||-27.7750893||21.160.03||23.660.03||22.450.02||21.480.01||20.970.01||20.690.01||22.390.04||23.120.05||0.37||…||3.00|
|53.1675873ccConfusion in GALEX image||-27.7925072||23.270.54||…||23.760.04||23.450.02||23.010.03||22.720.03||23.480.24||…||0.71||…||5.67|
|53.1680603ccConfusion in GALEX image||-27.8074017||22.910.11||25.320.08||24.770.09||23.970.04||23.720.06||23.700.07||24.350.17||25.890.27||0.41||…||6.00|
|53.1704826bbSource is outside the UDF footprint||-27.7613792||21.580.03||…||22.530.01||21.840.01||21.490.01||21.390.01||23.160.09||23.910.09||0.25||…||4.67|
|53.1721077ccConfusion in GALEX image||-27.7969379||23.110.13||25.210.06||24.580.08||23.710.03||23.420.05||23.320.05||22.940.05||23.500.06||0.19||…||3.33|
|53.1725159ccConfusion in GALEX image||-27.7963371||21.050.06||23.660.03||23.010.03||22.150.01||21.850.02||21.670.02||22.890.08||23.270.10||0.29||0.3469||3.67|
|53.1778793bbSource is outside the UDF footprint||-27.7577496||23.500.12||…||25.020.05||25.050.05||24.440.06||24.270.06||…||…||0.91||…||6.00|
|53.1799660bbSource is outside the UDF footprint||-27.7573910||23.550.13||…||25.280.08||24.760.05||24.310.06||24.330.08||…||…||0.68||…||6.00|
|53.1841164bbSource is outside the UDF footprint||-27.7559299||22.820.13||…||24.890.06||24.730.05||24.410.07||24.330.08||…||…||1.89||…||6.00|
|53.1855202bbSource is outside the UDF footprint||-27.7554283||22.010.07||…||23.960.03||23.070.01||22.520.02||22.360.02||…||…||0.43||0.5328||3.33|
|53.1869583ccConfusion in GALEX image||-27.7910004||19.030.01||21.670.01||20.210.01||19.190.00||18.650.00||18.440.00||21.320.02||21.890.02||0.23||…||2.33|
|53.1877899ccConfusion in GALEX image||-27.7940979||19.830.02||22.230.02||21.490.01||20.680.01||20.450.01||20.230.01||21.950.02||22.490.03||0.27||0.3446||3.67|
|53.1879463bbSource is outside the UDF footprint||-27.7615261||21.550.06||…||21.500.01||20.610.01||20.220.01||20.050.01||23.590.11||24.470.21||0.22||…||3.33|
|53.1898384bbSource is outside the UDF footprint||-27.7588539||20.640.03||…||21.840.01||21.410.01||21.170.01||21.170.01||22.430.04||22.650.03||0.24||…||6.00|
|53.1901550bbSource is outside the UDF footprint||-27.7651958||20.490.06||…||22.010.02||20.950.01||20.520.01||20.290.01||22.820.07||23.690.13||0.36||…||3.00|