THE FIRST ULTRA-COOL BROWN DWARF DISCOVERED BY THE WIDE-FIELD INFRARED SURVEY EXPLORER

Brown dwarfs (BDs) are star-like objects with masses too low to sustain core fusion of hydrogen to helium. Although first predicted to exist in the early 1960's (Kumar 1963; Hayashi & Nakano 1963), BDs were not discovered in bulk until decades later by deep, large-area surveys such as the Two Micron All Sky Survey (2MASS; Skrutskie et al. 2006), the Sloan Digital Sky Survey (SDSS; York et al. 2000), and the DEep Near-Infrared Southern Sky Survey (DENIS; Epchtein et al. 1997). Strong methane absorption characterizes the coolest observed class of BDs, providing the hallmark of the "T" spectral type. To date, more than 200 T dwarfs have been discovered (DwarfArchives.org). T dwarfs have effective temperatures ranging from ∼1300 K down to ∼450 K (Lucas et al. 2010; Burgasser et al. 2006; Burningham et al. 2008; Delorme et al. 2008; Golimowski et al. 2004; Warren et al. 2007), and the coldest T dwarfs are characterized by deep absorption bands of CH₄, H₂O, H₂, and NH₃ appearing throughout the near- and mid-infrared portions of their spectrum (Borysow et al. 1997; Irwin et al. 1999; Leggett et al. 2007a). No census of our solar neighborhood can be complete without discovering the lowest temperature BDs, as it is anticipated that they exist in numbers rivaling those of normal stars (Burningham et al. 2010b; Lodieu et al. 2007; Moraux et al. 2003; Reid et al. 1999). Additionally, the spatial density of ultra-cool and low-mass BDs can be used as a sensitive probe of the lowest masses that can be produced via different star formation processes (Bate 2009; Burgasser 2004).

One of the two primary science objectives for the Wide-field Infrared Survey Explorer (WISE) is to find these coldest BDs, which represent the final link between the lowest mass stars and the giant planets in our own solar system. WISE is a NASA Medium-class Explorer mission designed to survey the entire sky in four infrared wavelengths, 3.4, 4.6, 12, and 22 μm (Wright et al. 2010; Liu et al. 2008; Mainzer et al. 2005). WISE consists of a 40 cm telescope that images all four bands simultaneously every 11 s. It covers nearly every part of the sky a minimum of eight times, ensuring high source reliability, with more coverage at the ecliptic poles. Astrometric errors are less than 0.5 arcsec with respect to 2MASS (Wright et al. 2010). The preliminary estimated S/N = 5 point source sensitivity on the ecliptic is 0.08, 0.1, 0.8, and 5 mJy in the four bands (assuming eight exposures per band; Wright et al. 2010). Sensitivity improves away from the ecliptic due to denser coverage and lower zodiacal background.

WISE's two shortest wavelength bands, centered at 3.4 and 4.6 μm (W1 and W2, respectively), were specifically designed to optimize sensitivity to the coolest types of BDs (J. D. Kirkpatrick et al. 2010, in preparation). In particular, BDs cooler than ∼1500 K exhibit strong absorption due to the ν₃ band of CH₄ centered at 3.3 μm, with the onset of methane absorption at this wavelength beginning at ∼1700 K (Noll et al. 2000). The as-measured sensitivities of 0.08 and 0.1 mJy in W1 and W2 allow WISE to detect a 300 K BD out to a distance of ∼8 pc, according to models by Marley et al. (2002) and Saumon & Marley (2008).

In order to facilitate the search for new BDs in the WISE data set, we first examined the WISE fluxes and colors of known BDs. The W1 − W2 and IRAC [3.6] − [4.5] colors of a sample of L and T dwarfs are shown in Figure 1(a) along with their Spitzer/IRAC colors (Patten et al. 2006). The WISE W1 band spans the deep H₂O and CH₄ absorption features found in the coolest BDs, while the W2 band is centered on the window near 5 μm where much of the flux of these objects is emitted; it also includes the 4.6 μm CO absorption feature and the 4.2 μm CO₂ feature (Yamamura et al. 2010). While the W1 and W2 bands are superficially similar to the Spitzer/IRAC bands 1 and 2 (which were also designed to isolate cool BDs; Fazio et al. 1998), the WISE W1 band is wider to improve discrimination between normal stars and ultra-cool BDs (Wright et al. 2010). This results in a systematically larger W1 − W2 color compared to [3.6]−[4.5]. Figures 1(b) and (c) show that J − W2 and H − W2 increase with spectral type, with H − W2 showing the smoothest correlation between color and spectral type for BDs later than T5, in good agreement with Leggett et al. (2010); these indices are likely to be a valuable means of identifying new L and T dwarfs. Table 1 contains the WISE photometry of the small sample of L and T dwarfs shown in the figures, and Figure 2 shows a model of a cool BD along with the WISE W1 and W2 transmission curves.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Several methods were employed for searching for new BD candidates in the WISE data set. The WISE preliminary data processing pipeline produces calibrated frames for each of the individual exposures. These images have been dark-subtracted, flat-fielded, distortion-corrected, have had artifacts flagged and sources extracted. Once the individual images are produced, they are co-added to produce a single frame for each inertial position.

Searches for new BD candidates were performed on both the single-frame and multi-frame co-added mosaic source lists. The co-add search provided the maximum depth and sensitivity since the co-added frames generally feature the combination of at least eight frames per position. The co-add search was performed by searching the source lists for objects with a W1 − W2 ⩾ 2.0, a W2 signal-to-noise ratio (S/N) ⩾10, although sources were not required to be detected in W1 (limiting magnitudes are given to 2.5σ). Sources were required to have been observed a minimum of four times in order to ensure reliability. Sources affected by artifacts such as diffraction spikes and latent images were excluded, and the reduced chi-squared point-spread function goodness-of-fit parameter was required to be consistent with a point source (w2rchi2 ⩽ 3, where the quantity w2rchi2 represents the goodness of fit of a point source model to the set of observed pixel values in band W2, whereby a value close to unity indicates consistency). No requirement on the W1 S/N was levied in order to increase sensitivity to the coolest BDs, which can fall below the W1 sensitivity limits. Subsequent searches were performed with varied criteria, including requiring the W4 S/N to be ⩽5, and only sources with high reliability were accepted (as determined by comparing the number of detections to the number of coverages). Once sources meeting these criteria were extracted from the WISE data, they were compared with images from the Digitized Sky Survey (DSS), 2MASS, and SDSS. BD candidates were required to have blue and red optical fluxes from DSS and SDSS (where SDSS data were available) consistent with expectations for very late T dwarfs observed with WISE. Cold WISE BD candidates were required to be unseen in DSS or SDSS, or in some cases visible at the longest wavelengths only in these surveys.

A source, given the provisional designation WISEPC J045853.90+643451.9 (hereafter WISEPC J0458+64), was located in the co-added source list that met all of the above criteria; Table 2 summarizes the astrometry and Table 3 summarizes the WISE magnitudes and colors for the source. Although WISEPC J0458+64 is faintly visible in the 2MASS J-band original scans, slightly offset from the WISE-derived position, the source was not bright enough to have been included in either the 2MASS All-Sky Point Source Catalog or Reject Table. Aperture photometry of the J-band flux using standard IDL astronomy library routines reveals it to be detected at 3σ. Figure 3 shows the WISE, 2MASS, and DSS finder images. At the position of the BD candidate, there are 13 separate images from WISE. The S/N in the resulting co-added data in W1 are 11.9 and 36.3 in W2. The source was not detected in W3 due to contaminating cirrus, which can confound photometry at these wavelengths. The object is not detected in W4 to a limiting magnitude of 9.3 (Vega system) or 1.70 mJy. The images of WISEPC J0458+64 were searched for widely separated companions, but none were found. No known nearby stars with common proper motion were found.

Zoom In Zoom Out Reset image size

4.1. Fan Mountain–FanCam

4.2. Large Binocular Telescope—LUCIFER

4.3. NASA Infrared Telescope Facility–SpeX

We obtained a 0.8–2.5 μm spectrom of WISEPC J0458+64 on 2010 September 12 (UT) using SpeX (Rayner et al. 2003) on the 3.0 m NASA Infrared Telescope Facility on Mauna Kea. We used the low-resolution prism mode with the 05 wide slit which yielded an average resolving power of R = 150. A series of 120 s exposures at two positions along the 15'' slit were obtained to facilitate sky subtraction. Observations of the A0 star HD 14632 were obtained after the science exposures for telluric correction and flux calibration purposes as well as flat and arclamp exposures.

The data were reduced using Spextool (Cushing et al. 2004), the IDL-based data reduction package for SpeX. Given the faintness of WISEPC J0458+64, the images were first pair-subtracted and then averaged together robustly in order to facilitate extraction. The spectra were then optimally extracted (e.g., Horne 1986) and wavelength calibrated using argon lamp exposures. The raw spectrum was then corrected for telluric absorption and flux-calibrated using the observations of the A0 V star HD14632 and the technique described in Vacca et al. (2003). The S/N of the final spectrum ranges from a maximum of ∼40 in the J band to ∼5 in the K band. The spectrum was absolutely flux calibrated as described in Rayner et al. (2009, see Equation (1)) using the FanCam J- and H-band magnitudes. The agreement between the LUCIFER and SpeX data is excellent (Figure 4).

Zoom In Zoom Out Reset image size

5.1. Spectral Analysis

The LUCIFER and SpeX spectra are shown in Figure 4. The spectra exhibit the deep absorption features of CH₄ and H₂O indicative of a very late T dwarf. The spectrum of WISEPC J0458+64 can be compared to the NIR spectra of the very late T dwarfs 2MASS J0415−09 (Burgasser et al. 2002), ULAS J003402.77−005206.7 (hereafter ULAS J0034−00; Warren et al. 2007), CFBDS J005910.90−011401.3 (CFBDS J0059−00; Delorme et al. 2008), ULAS J133553.45+113005.2 (ULAS J1335+11; Burningham et al. 2008), and SDSS J1416+13B (Burningham et al. 2010a; Scholz 2010). As shown in Figure 5, WISEPC J0458+64 appears to best match the spectrum of the T9 dwarfs, although formal assignment of a spectral type will have to wait for the identification of spectral standards for objects this late.

Zoom In Zoom Out Reset image size

The atmospheric parameters of WISEPC J0458+64 were estimated by comparing the SpeX spectrum and the WISE photometry to the model spectra of Marley et al. (2002) and Saumon & Marley (2008). We used a grid of 80 cloudless models with T_eff ranging from 500 to 1000 K in steps of 100 K, log g = 4.0, 4.5, 5.0, and 5.5 (cm s⁻²), [M/H] = 0, −0.3, and K_zz = 0, 10⁴ cm² s⁻¹. K_zz is the eddy diffusion coefficient and ranges from ∼10² cm² s⁻¹ to 10⁵ cm² s⁻¹ in planetary atmospheres (e.g., Saumon et al. 2006). K_zz = 0 corresponds to an atmosphere without vertical mixing in the radiative zone where chemical equilibrium prevails. The models were first smoothed to the resolution of the spectra and interpolated onto the wavelength scale of the data. In order to include the WISE photometry in the comparison, we converted the observed W1 and W2 magnitudes to average fluxes using the following zero points on the Vega system: 306.681, 170.663, 29.0448, and 8.2839 Jy in bands W1 through W4 and using the relative spectral response curves from Wright et al. (2010) and T. Jarrett et al. (2010, in preparation).

We identified the best-fitting model spectra in the grid using the goodness-of-fit statistic G_k (w_i = 1; all data points receiving equal weight) described in Cushing et al. (2008). For each model k, we compute the scale factor C_k = (R/d)², where R and d are the radius and distance of the dwarf, respectively, that minimizes G_k. The best-fitting model is identified as having the global minimum G_k value. To access the uncertainty in the selection process, we run a Monte Carlo simulation using both the uncertainties in the individual spectral and photometric points and the overall absolute flux calibration of the SpeX spectrum (see Cushing et al. 2008; Bowler et al. 2009). We report the quality of the fit between the data and model k as f_MC, the fraction of Monte Carlo simulations in which model k is identified as the best-fitting model.

The best-fitting model has T_eff = 600 K, log g = 5.0, [Fe/H] = 0, K_zz = 10⁴ cm² s⁻¹, and f_MC = 0.66 while the second best-fitting model has T_eff = 600 K, log g = 5.5, [Fe/H] = 0, K_zz = 10⁴ cm² s⁻¹, and f_MC = 0.34. Given their f_MC values, these are the only two models in the grid that fit the data given the uncertainties. Figure 6 shows the SpeX spectrum and WISE photometry of WISEPC J0458+64 along with the best-fitting model. The spectrum is well matched by the model, but the W2 point is discrepant by over 10σ. Although Figure 5 shows that WISEPC J0458+64 appears to show suppressed K-band flux relative to other late T dwarfs, the interpretation of the feature is uncertain as the models for objects in this temperature regime are as yet not well validated. Suppressed K-band flux has previously been ascribed to low metallicity and/or high surface gravity (Leggett et al. 2009; Stephens et al. 2009); however, our model fits do not indicate unusual metallicity. The atmospheric parameters derived herein for WISEPC J0458+64 are consistent with those derived for other late-type T dwarfs (e.g., Leggett et al. 2009; Burningham et al. 2009). In particular, there is evidence for the presence of vertical mixing in the atmosphere of WISEPC J0458+64 as K_zz = 10⁴ cm² s⁻¹. However, the poor match between the data and the model at 4.6 μm, a spectral region particularly sensitive to vertical mixing (Leggett et al. 2007b; Geballe et al. 2009), makes drawing any conclusions difficult.

Zoom In Zoom Out Reset image size

Following Bowler et al. (2009), we can also estimate a spectrophotometric distance to WISEPC J0458+64 using the scale factor C_k. We first estimate the radius of WISEPC J0458+64 using the derived (T_eff, log g) values and the evolutionary models of Saumon & Marley (2008). Using errors given by half the grid spacing of 50 K and 0.25 dex, we estimate a distance of 6–8 pc for WISEPC J0458+64. This distance is in rough agreement with the photometric distance of 9.0 ± 1.9 pc, using the apparent J magnitude of 17.47 ± 0.05 and absolute magnitude M_J = 17.7 ± 0.45. This was computed from a weighted average of M_J for the late T dwarfs Wolf 940B (Burningham et al. 2009), ULAS0034−00 (Smart et al. 2010), and SDSS J1416+13B (Scholz 2010), using the worst-case error.

5.2. Astrometry

One of the key questions in BD science is whether or not the spectral sequence ends with the T dwarfs, or whether another spectral type (with a proposed designation of Y; Kirkpatrick et al. 1999) is justified. Currently, only about two dozen objects with spectral types later than T7 are known, and four objects are known to have type T9 or later: ULAS J0034−00 (Warren et al. 2007), CFBDS J0059−00 (Delorme et al. 2008), ULAS J1335+11 (Burningham et al. 2008), and UGPS J0722−05 (Lucas et al. 2010). New spectral indices for typing these late-type T dwarfs have been developed as the CH₄ absorption band depths may not continue to increase in the NIR with temperatures lower than those of T8/T9 dwarfs (Burningham et al. 2010b). We compute the H₂O−J, CH₄ − J, W_J, H₂O−H, CH₄ − H, and CH₄ − K indices given by Burgasser et al. (2006), Burningham et al. (2008), and Warren et al. (2007) for WISEPC J0458+64 (see Table 3). However, as noted in Burningham et al. (2008), these indices must be viewed with caution for extremely late T dwarfs; the values we compute for WISEPC J0458+64 tell us only that it is later than T7/T8. Given the paucity of very late T dwarfs, assignment of a definitive spectral type will await the identification and definition of spectral standards. As this remarkably cool object was found easily in some of the first WISE data, WISE is likely to find many more similar objects, as well as cooler ones, inevitably producing abundant candidates for the elusive Y class BDs. Scaling from the Spitzer sample of 4.5 μm selected ultra-cool BD candidates (Eisenhardt et al. 2010), we expect to find hundreds of new ultra-cool BDs with WISE. We compute that for most likely initial mass functions, WISE has a better than 50% chance of detecting a cool BD which may actually be closer to our Sun than Proxima Centauri (Wright et al. 2010).

We report the discovery of the first ultra-cool BD with WISE, an object which is consistent with an extremely late T dwarf. The object is easily detected by WISE, with S/N of ∼40 at 4.6 μm. Further, WISEPC J0458+64 lies in projection only 13° off the Galactic midplane. WISE can pick out ultra-cool BDs like this one easily even in regions that are traditionally confused, as the number of sources per square degree at 4.6 μm is significantly lower than in optical and near-IR wavelengths. The WISE survey offers an All-Sky Survey with extremely uniform and stable instrumental characteristics and survey cadence. By surveying the entire sky, including the galactic plane, WISE can provide an accurate census of cold BDs in the solar neighborhood, which is necessary for determining the initial mass function at low masses. WISEPC J0458+64, along with UGPS J0722−05 and the few other BDs like this, represents the first in what is anticipated to be a significant population of objects revealed by the WISE survey. We expect that BDs discovered by WISE will at last close the remaining gap in temperature between the coolest objects currently known (∼500 K) and potentially those of the giant planets in our own solar system (Wright et al. 2010).

This publication makes use of data products from the Wide-field Infrared Survey Explorer, which is a joint project of the University of California, Los Angeles, and the Jet Propulsion Laboratory/California Institute of Technology, funded by the National Aeronautics and Space Administration. We are deeply grateful for the outstanding contributions of all the members of the WISE team. We thank Roger Griffith for assistance with Figure 2 and Beth Fabinsky for assistance with early searches. Support for the modeling work of D.S. was provided by NASA through the Spitzer Science Center. M.C. was supported by an appointment to the NASA Postdoctoral Program at the Jet Propulsion Laboratory, administered by Oak Ridge Associated Universities through a contract with NASA. This publication makes use of data products from the Two Micron All Sky Survey (2MASS). 2MASS is a joint project of the University of Massachusetts and the Infrared Processing and Analysis Center/California Institute of Technology, funded by the National Aeronautics and Space Administration and the National Science Foundation. This research has made use of the NASA/IPAC Infrared Science Archive (IRSA), which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration. Our research has been benefited from the M, L, and T dwarf compendium housed at DwarfArchives.org whose server was funded by a NASA Small Research Grant, administered by the American Astronomical Society. We are also indebted to the SIMBAD database, operated at CDS, Strasbourg, France. The Digitized Sky Surveys were produced at the Space Telescope Science Institute under U.S. Government grant NAG W-2166. The images of these surveys are based on photographic data obtained using the Oschin Schmidt Telescope on Palomar Mountain and the UK Schmidt Telescope. The Second Palomar Observatory Sky Survey (POSS-II) was made by the California Institute of Technology with funds from the National Science Foundation, the National Geographic Society, the Sloan Foundation, the Samuel Oschin Foundation, and the Eastman Kodak Corporation. The Oschin Schmidt Telescope is operated by the California Institute of Technology and Palomar Observatory. We thank Richard Green and the LBT staff for making the LUCIFER observations possible. The LBT is an international collaboration among institutions in the United States, Italy, and Germany. LBT Corporation partners are the University of Arizona on behalf of the Arizona University System; Istituto Nazionale di Astrofisica, Italy; LBT Beteiligungsgesellschaft, Germany, representing the Max-Planck Society, the Astrophysical Institute Potsdam, and Heidelberg University; The Ohio State University, and The Research Corporation, on behalf of The University of Notre Dame, University of Minnesota and University of Virginia. The LUCIFER Project is funded by the Bundesministerium für Bildung und Forschung (BMBF). It is a collaboration of five German institutes: Landessternwarte Heidelberg, Max Planck Institut für Astronomie (Heidelberg), Max Planck Institut für Extraterrestrische Physik (Garching), Fachhochschule für Technik und Gestaltung (Mannheim), and Astronomisches Institut der Universität Bochum.