Chinese Team:

NAOC: Xu Zhou (PI), Hu Zou (Deputy PI), Zhaoji Jiang, Jun Ma, Tianmeng Zhang, Zhimin Zhou, Jundan Nie, Jiali Wang

PKU: Xuebing Wu, Ran Wang, Linhua Jiang 

Tsinghua University: Xiaofeng Wang

US Team:

Xiaohui Fan (PI, Arizona), Ed Olszewski, Brenda Frye, Ian McGreer, Mike Lesser, Buell Jannuzi, Eduardo Rozo, Peter Nugent, David Schlegel, Dustin Lang, Arjun Dey

The Beijing-Arizona Sky Survey (BASS) is an ambitious wide-field multicolor survey of 5000 deg^2 of the Northern Galactic Cap using the 90prime imager on the 2.3m Bok Telescope at Kitt Peak, as a four-year collaboration between the Chinese team led by NAOC and the US team led by Steward Observatory, University of Arizona. The survey will use 400 nights, covering SDSS g, r, and z bands, reaching limiting AB magnitude of 24.4, 24.0 and 23.4, respectively. BASS will cover half of the SDSS extragalactic footprint at 1.2 to 2.9 magnitude deeper. It will produce a new legacy wide-field imaging dataset, with scientific impact comparable to the original SDSS imaging survey as the leading imaging survey of the Northern Galactic Cap for the foreseeable future. It will provide unique science opportunities in a wide range of topics in galactic and extragalactic astronomy, including Galactic structure, near-field cosmology, AGN evolution, high redshift quasars, large scale structure of the universe and time-domain astronomy, and produce high impact results in many of these areas. It will also compliment surveys such as GAIA and LAMOST. BASS will put China at the forefront of wide-field survey astronomy and help train a new generation of astronomers best prepared for the big data science enabled by LSST, Dome-A and the space station projects. The survey will follow the successful SCUSS model with strong participations and support from both China and US teams and produce high level science products for the general astronomical community. In addition, the survey data will allow effective target selection for the DESI dark energy spectroscopic survey project in the survey area, and be used as in-kind contribution for the Chinese community to joining DESI project.

SDSS, with its unique combination of wide-field spectroscopic and imaging surveys, has made revolutionary discovery in a wide range of areas in astronomy and cosmology, from dark energy, the most distant quasars, to near-field cosmology and cool dwarfs. However, the SDSS imaging survey was carried out more than a decade ago. SDSS only reaches r~22, sensitive to low-redshift galaxies, luminous quasars and only the tip of iceberg of the Galactic halo substructure. There is a glaring gap between what is available now in photometric data, and what LSST can offer; this is even more acute in the Northern hemisphere, as most of the new, deeper imaging survey being planned in the south (e.g. SkyMapper, KIDS, LSST); note that PanStarrs is no deeper than SDSS except in z band.

To fully exploit the existing spectroscopic data also needs deeper imaging. The SDSS/BOSS surveys contain ~2.8 million spectra, including 300,000 stars, 700,000 galaxies at z < 0.2, 500,000 galaxies at 0.2 < z < 0.5, 1 million galaxies at z > 0.5, 100,000 QSOs at z < 2, and 200,000 quasars at z > 2. SDSS-IV/eBOSS (2014-2020) will add another 600,000 galaxies at 0.6 < z < 1 and 750,000 quasars at z > 0.9. Although SDSS-I spectroscopy is relatively shallow, BOSS and eBOSS are pushing to the limit of the original SDSS photometry (Dawson et al. 2013). The existing imaging data is merely sufficient for target selection. We lack precise photometry, size measurements, morphology, and any detailed information of the galactic environment that are needed to study galaxy properties in addition to the large scale structure and cosmology. By imaging ~2 magnitudes fainter in three optical bands, we will increase the number of z > 0.5 by close to two order of magnitude. In addition, SDSS-IV/eBOSS has demonstrated that using g, r, and z band with the addition of WISE, we can efficiently isolate z>0.5 galaxies, measure stellar mass and select AGNs at high redshift.

BASS imaging will offer unique opportunities in a wide-range of areas ranging from Galactic structure and near-field cosmology, to galaxy clustering and high-redshift quasars, to time-domain astronomy. As a natural successor of the SDSS imaging survey, it will have board impact in galactic and extragalactic astronomy prior to the LSST era. By providing photometry 2 magnitude deeper in the three most critical optical bands over almost half of the original SDSS footprint, BASS will enable qualitative new imaging science, and at the same time greatly enhance the scientific reach of existing spectroscopy.

The SDSS revolutionized our view of the Milky Way, especially its halo and environment. SDSS has found numerous stellar halo streams (e.g. Newberg et al. 2002; Yanny et al. 2003, Grillmair 2009) and has doubled the number of dwarf galaxies (Willman 2010). BASS data will reach more than a factor two further into the halo, and when combined with the equatorial data from the approved NOAO/DECam survey, will increase the volume probed by close to an order of magnitude. We will use these data to study:

• Galactic structure through proper motion

Understanding the origin and evolution of the Galaxy is one of the major goals of modern astronomy. The Galaxy stellar system is broadly composed of four distinct parts although in reality there is likely to be considered overlapping among them: the bulge, thin disc, thick disc, halo. The stellar kinematic and chemical abundance distributions are very important to characterize these components. Combining BASS multi-epoch observations with SDSS, we will be able to measure proper motion over >15 years of baseline for objects at g<22 down to the level of 6mas/year. This unique dataset will be combined with brighter objects from GAIA, allowing detailed studies of galactic dynamics deep in the Galactic halo, a key probe to the evolution of our Galaxy and near-field cosmology.

• Searching satellite dwarf galaxies of the Milky Way

Despite the ΛCDM model being highly successful in describing the observations of the large-scale structure and evolution of the Universe, it has some problems in describing structures at small scales. One of the main problems is the “missing satellite problem”, namely the discrepancy between the number of predicted subhalos in N-body simulations and the one observed. Klypin et al. (1999) and Moore et al. (1999) first noticed that numerical simulations predicted much more subhaloes in galactic and cluster haloes. The number of satellites having circular velocities larger than Ursa-Minor and Draco were about 500, while as it is well known the MW dSphs are much less (9 bright dSphs, Sagittarius, the LMC and SMC). We expect to find 4−10 new dwarf galaxies in these data.

• Searching substructures or tidal streams in the Galactic halo

One of the key results from the SDSS is the discovery of ultra-faint dwarfs and tidal stream features in the Galactic halo which revolutionized our understanding of galaxy assembly (Figure 1). However, at the depths of SDSS, we could only effectively probe the inner halo. BHB stars are excellent tracers to probe the structure of the Galactic halo since they are luminous and have a well-determined color-magnitude relation (or absolute magnitude). With BASS data, our photometric parallax-based 3-D halo map will reach ~40kpc using main sequence stars (Juric et al. 2008), ~80kpc for turnoff stars, selected using g and r colors (Bell et al. 2008) and to ~150kpc in the area where u-band data is available (e.g. from CFHT) for BHB selection (Ruhland et al. 2006). Theoretical models predict a dramatic increase in halo substructure at larger distance (Helmi et al. 2011). Our data will strongly test this CDM prediction. Deeper imaging data on known streams will be used on test the presence of additional satellite galaxies through their impact on the stream structure (e.g. Yoon et al. 2011).

The evolution of galaxies is tightly linked to the growth of their central super-massive black holes. Quasars might represent a key transitional phase in the evolution of massive galaxies (e.g. Hopkins et al. 2006). SDSS provided the most comprehensive view of AGN evolution so far (Ross et al. 2013), and opened the window of studies of high-redshift quasars close to the end of reionization (Fan et al. 2006). BASS will create an unprecedented sample of photometrically selected AGNs (a few million) over a wide range of redshift, AGN type, and AGN/galaxy luminosity. It also leverages the upcoming eROSITA all-sky survey, which will make a near-complete catalog of X-ray bright AGN over the last 8 Gyr. At the proposed depths, we will identify ~97% of AGN (~ 95% type 2 AGN, Brusa et al. 2010), dramatically improving our ability to compile cross-identification of AGN across the electromagnetic spectrum. The AGN sample will allow us to study AGN clustering as a function of redshift which will provide strong tests to the model of AGN evolution and black hole/galaxy co-evolution and to study how the overall SED of AGN, correlate with redshift, luminosity, host galaxy properties and environment (Figure 2, Aird et al. 2012).

Because of the increased survey volume and depth, BASS will be especially suitable for the studies of the rarest types of AGNs and quasars. A key limiting factor of the SDSS quasar survey is the shallow z-band data. BASS will reach more than 2 magnitude deeper in z, comparable to the stripe 82 z-band data. BASS data, combined with the full depths WISE data, will allow the discoveries of large number of quasars at the redshift up to 6.5. We estimate that the survey will produce 5 times more z>6 quasars compared to SDSS, providing key reionization probes as well as mapping the detailed luminosity function of faint quasars which provides key constraints on early growth of supermassive black holes.

The relationship between the physical properties of galaxies and the dark matter halos in which they reside is a key problem in galaxy evolution. The contents (and shapes) of galaxy dark matter halos can be revealed from the cross-correlation of spectroscopic and imaging maps. These methodologies benefit substantially from deeper imaging. Our proposed depth of z = 23 will increase these samples by factors of 15 or more from SDSS. Cross-correlation studies use angular correlations to tie deep photometric catalogs to overlapping spectroscopic maps, measuring the mean environments and clustering of galaxies and AGN with great accuracy (e.g. Jiang et al. 2012). The proposed survey will extend this to far larger spectroscopic and photometric samples at high redshift, measuring the satellite distributions around the central galaxies as a function of redshift, luminosity, stellar mass, color and other physical properties.

SDSS has obtained redshifts of 1.5 million massive galaxies. These are often the central and brightest galaxies in groups and clusters; however, current imaging often cannot detect their satellites. The proposed imaging will significantly improve stellar mass models for these galaxies and enable a sensitive search for cluster members around them. We will construct volume-limited cluster catalogs out to z ≈ 0.7, with richer systems detected to yet higher redshifts (z>=1), providing key cosmological probes (e.g. Becker et al. 2007).

With a flexible schedule on the telescope, we will be able to embed a time-domain program into our survey, either by spreading the required exposure time over the entire duration of the survey or scanning a small continuous area with multiple observations. One interesting subject of this time-domain program is that we can discover young SNe immediately after the explosion to probe a new parameter space for understanding the explosion physics of SNe and for making links to their progenitor stars. The very early discovery of a SN Ia by our survey and immediate follow-up spectroscopy may allow us to put rather tight constraints on the emission lines from the H-rich CSM. With SNe light curves, we can also constrain both the rise time to maximum light and the peak luminosity of core collapse SNe to constrain their explosion models. Besides SNe, we can also discover a large set of variable stars, asteroids, and transients such as CVs, quasar/AGN, and nova.

BASS proposed here will not only create the deepest survey of the northern SDSS footprint, but also be sufficient for identifying the primary LRG, ELG and QSO targets for the Dark Energy Spectroscopic Instrument (DESI) Key Project and for any possible public access time on the instrument. DESI, a 5000-fiber spectrograph destined for the Kitt Peak Mayall telescope, will undertake a spectroscopic survey an order of magnitude larger than SDSS, measuring spectra and redshifts for 20 million emission line galaxies (ELGs), 4 million luminous red galaxies (LRGs) and 2 million QSOs. The survey is designed to probe the expansion history at 0 < z < 3.5 using the baryon acoustic oscillation scale, and to map the dark matter and gravitational growth through redshift space distortions (Schlegel et al. 2011). DESI will have revolutionary capabilities for broad investigations on the origin and evolution of galaxies, Galactic structure, rare-object discovery, and result in one of the most complete high-z AGN catalogs to date (Pilachowski et al. 2012).

BASS will find a large number of rare classes objects, including halo and cool white dwarfs, to cool L/T brown dwarfs, to carbon stars, gravitationally-lensed quasars, high-z quasars etc. It will produce large and valuable dataset for spectroscopic follow-up observations. DESI spectroscopic survey will provide a key follow-up facility: DESI will have a wide range of ancillary fibers as part of the science operation, and DESI bright time allocation is expected to be focused on relatively bright galaxies and stars, with millions of fibers, all in the BASS survey area. If China joins DESI through BASS contribution, it will be in the best position to propose and to take advantage of the DESI spectroscopic capabilities on non-cosmological science.  

Figure 1. A composite, filtered surface density map of stars in the SDSS. Stars have been filtered to select stellar populations at different distances with color-magnitude sequences similar to that of the globular cluster M 13 (Grillmair 2009). The distances of the streams range from 4 kpc for Acheron, to 9 kpc for GD-1 and the Anticenter Stream, to 50 kpc for Sagittarius and Styx. BASS will cover half of the area to 2 magnitude deeper, allow discovery of 4-10 new dwarfs and reveal many new halo substructures.

Figure 2. The Clustering of Active Galactic Nuclei. The bias (top panels) and comoving clustering length (lower panels) of quasars in three models of quasar growth, for samples of various limiting magnitude. BASS will be able to probe to limiting magnitudes of m~23 for quasar selection reliably, which is 2 mag deeper than the SDSS quasar sample. One needs to go appreciably fainter than the SDSS magnitude limit to break the degeneracy in models. The three models are (left to right): an efficient feedback model (in which infall to the SMBH halts immediately after a quasar episode); a model in which SMBHs grow smoothly to z = 2; and a model in which black hole growth is tied to that of the dark matter halo to z = 2. Figure from Hopkins et al. (2007).  

We will use the Steward 2.3m Bok Telescope on Kitt Peak, with the 90Prime 8kx8k CCD camera that covers 1 deg2. For BASS g and r band survey, we will use the existing blue-sensitive CCD, which we plan to have a number of minor improvements on CCD controller and computer. For BASS z band survey, the Steward Observatory, in collaboration with LBNL, is constructing a new Dewar equipped with four 4kx4k complete depletion red sensitive CCDs, which will be ready from late 2015 and on the sky from the second year of our survey. The new red camera will improve z-band observing efficiency by a factor of three. During the operation, we will switch between the two Dewars in the dark and grey/bright half of the month.

For this large survey, the Steward Observatory will continue to carry out a number of improvements on the telescope control system and focusing mechanisms of the camera. These upgrades will be carried out as the survey progresses. The goal is for semi-automated observing with remote observing/monitoring and only minimum on-site support after one or two seasons of observing for this project. This will reduce the operation cost on the observing team.

Figure 3 shows the entire SDSS photometric and spectroscopic footprint, which covers the high-galactic latitude sky accessible in the Northern hemisphere. Our goal is to create a legacy photometric dataset that is about 2 magnitudes deeper than the SDSS. For BASS, we focus on the northern 5000 deg2 that can only be reached by a Northern telescope. Meanwhile, NOAO has approved a survey project using DECam on the 4m Blanco telescope at CTIO which will survey the equatorial region to a depth similar to BASS; the DECam data will be publically available without proprietary period.

Figure 3. SDSS footprint. BASS will cover the northern 5000 deg2 to 2 magnitude deeper.

BASS will obtain data in 3 SDSS bands (g, r, and z); this is the minimum number of bands needed for (a) color selection of galaxies by types and redshift; (b) obtaining photometric parallax of main sequence stars, and © color selection of AGNs. We choose the depth by requiring: (i) g-r-z color selection of sub-L* galaxies at z~1; (ii) high signal-to-noise data for the measurement of structural parameters of galaxies with known redshifts on the faintest z = 20 mag eBOSS galaxy targets; (iii) color-separating main sequence turnoff stars in distant halo (~80kpc), (iv) detections of nearly 100% of the WISE sources for AGN selection. These requirements translate to a 5-σ point source depths of g, r, z ≈ 24.4, 24.0, 23.4 AB mag.

We then carried out a comprehensive simulation of our survey using existing Bok/90Prime data, seeing and weather statistics, and assuming realistic galaxy morphology for galaxy photometry. We find that BASS will need:

• g and r band survey: 180 nights for 5000 deg2, using 90prime+blue dewar, m_lim = 24.4 and 24.0 (5-σ). When including overhead and weather, we find that we need 75 scheduled nights (2/3 usable) in g and 105 nights in r to cover 5000 deg^2.
• z-band survey: 180 nights for 5000 deg2, using 90prime+red dewar, m_lim = 23.4 (5-σ). The observing efficiency is estimated following the same procedure as for g and r bands, expect now we are scaling the CCD response curve from the current Bok CCDs (blue sensitive) to those of the LBL thick devices, which are the same as those installed in DECam.

We also reserve 10% contingency for instrument issues and guard against atypically bad weather. If not needed, we will use this contingency to conduct a pilot survey in u and i bands over ~10% of the survey towards the end of the survey. This will enable a full five-band SDSS coverage of ~500 deg^2 at 2 magnitude deeper, essentially double the SDSS stripe82 data over the same band, and allow us to study a possible expansion of BASS into u and i bands. Altogether, we are requesting 400 nights on Bok to complete BASS.

The above table compares BASS with SDSS, PS1, DECam, VST, and SkyMapper. When completed, BASS will be the deepest and most complete survey of the northern high galactic latitude sky. The survey area also overlaps with UKIDSS (YJHK, with YAB ~21.5) northern stripe (~2000 deg2).

BASS will only be observing in the spring semester. BASS fields will begin to be visible in Dec until telescope summer shutdown. The fact that BASS focuses on high DEC helps with scheduling. BASS will use 400 nights (200 dark/grey nights for g and r bands and 200 grey/bright nights in z band). Assuming a four-year survey, we will need 100 nights per year, or about half of the available nights between Jan and June. The BASS team will operate the telescope from Dec to June for dark/grey time, with the exception of 1 week around full moon, which will be devoted to IR projects on Bok. A letter from Steward Observatory Director Dr. Buell Jannuzi (who is a CoI of this project) confirms that these nights are available.

The Steward Observatory will be upgrading the Bok telescope control from 2015 aiming at completion in 2016. In the first one-two years, we will continue to run classically staffed observing similar to the mode used in SCUSS, with a mixture of Chinese and US-based observers. We aim at transitioning to mostly remote observing, with a Steward night assistant on-site, towards the later part of the survey.

Timeline of the survey: • 2014: survey planning, starting construction of the new red dewar • 2015: BASS survey starts with g, r bands • 2016: completion of new red Dewar, observing in all three bands • 2017: routine survey observation, data release I for the first two years • 2018: completion of BASS observing • 2019: data release of the entire BASS survey

Raw data reduction and detrending will be based on image processing pipeline developed in SCUSS (Zou et al. 2014) for the same Bok data. Image combination, measurements and photometric calibration will use the Tractor photometric system developed by David Schlegel and Dustin Lang, with further modifications by the BASS team. There will also be other data processing needs including image subtraction, variability studies, and proper motion measurements. For these needs, we plan to work with LSST developed, publicly available codes to fit BASS needs. This will have added benefit of getting us familiar with LSST data and science. Xiaohui Fan and Peter Nugent are Co-Chairs of the DESI imaging working group. The BASS team will be working closely with the DESI team on data reduction pipeline development and data distribution. BASS will release its data after the first two years and then after the entire survey for the entire Chinese astronomical community. Chinese astronomers are also encouraged to join observing/data analysis effort at any time.

We require 400 nights on Steward Observatory’s 2.3m Bok Telescope. As indicated in the supporting letter from Steward Observatory, the nominal cost on the telescope is 5k (dollar) per night. As a joint China-Arizona project, Steward is providing the night at a 50% discount, for $2.5k per night. So the total nominal cost for acquiring 400 nights is $1M. Steward Observatory also indicates that it will consider the Chinese investment in survey operation and data reduction when discussing the final cost to the TAP, should the project be selected. So the final telescope time cost could be lower than $1M after final negotiation.

This multi-year survey requires significant investment in observing support and data processing support, including costs associated with travel, classical observing, data transfer, processing, storage and distribution. It will also be a public survey. If BASS is selected, the survey team will work with the Pilot B project office and Steward Observatory to develop a detailed plan to support observing and data release.

The project will be managed by the PIs Xu Zhou and Xiaohui Fan, with deputy-PI Hu Zou. The project team includes key people from the NAOC, PKU, THU, Steward Observatory as well as participations from the DESI collaboration. We are: (a) experienced in successfully carrying out large survey programs; (b) knowledgeable about the Bok telescope and instrument and operational software; © experienced in the creation of pipelines and in reducing, analyzing and calibrating imaging data; (d) have a proven track record of delivering publicly useful data products in a timely manner; (e) experienced observers; and (f) outstanding scientists experienced in doing research with large and multi-wavelength datasets.

PIs Xu Zhou and Xiaohui Fan will each devote 1/2 of their research time to this project to manage the execution and data analysis and monitor the science products. Both are involved in the successful SCUSS project. Deputy PI. Hu Zou will lead the observing effort. CoIs Jun Ma and Zhimin Zhou will lead the extra-galaxy science, Chao Liu and Jundan Nie(NAOC) will lead the galactic science project. CoIs Linhua Jiang and Xuebing Wu (KIAA) will lead the AGN science project. CoI Xiaofeng Wang (Tsinghua) and Tianmeng Zhang (NAOC) will lead the time-domain project. On the US side, CoI Mike Lesser (Arizona) will lead the camera and telescope upgrade and CoI Peter Nugent (LBNL) and BASS Chinese team will work with PI. Xiaohui Fan to provide imaging data processing support. We emphasize again that BASS, if selected, we would strongly encourage additional Chinese astronomers to join and contribute to the survey.

BASS is a natural extension and expansion from the highly successful SCUSS survey, which is also a collaboration between China and US. This survey will have broad science impact, high profile legacy value and high visibility. In addition to the science produced by the survey, other benefits to the Chinese community include:

• Legacy data product with impact similar to that of SDSS imaging survey (2 mag deeper), raising the international profile of Chinese astronomy;
• Synergy with LAMOST/GAIA in Galactic structure studies;
• Synergy with future wide-field survey projects in China, in particular the space station survey project and Dome-A project, in development of software, observing tools and training;
• Preparing Chinese astronomers in the area of LSST science.

In addition, the DESI survey have decided that BASS data will be used in DESI target selection, therefore the Chinese contributions to the BASS survey will be used as in-kind contribution for China to join DESI, with a value about $1M. If BASS is supported by TAP, we will aim at forming a Chinese participation group for DESI membership and data right, that includes Chinese astronomers interested in BASS/DESI science.

• Aird, I.A. et al. 2012, ApJ, 746, 90
• Becker, M.R. et al. 2007, ApJ, 669, 905
• Bell, E.F. et al. 2008, 680, 295
• Brusa, M. 2010, ApJ, 716, 348
• Dawson, K. et al. 2013, AJ, 145, 10
• Fan, X. et al. 2006, AJ, 131, 1203
• Grillmair, C.J. 2009, ApJ, 693, 1118
• Holmes, R. et al. 2012, PASP, 921, 1219
• Hopkins, P.F. et al. 2006, ApJS, 163, 1
• Hopkins, P. H. et al. 2007, ApJ, 662, 110
• Jiang, T. et al. 2012, ApJ, 759, 140
• Juric, M. et al. 2008, ApJ, 673, 684
• Klypin, A. et al. 1999, ApJ, 522, 82
• Moore, B. et al. 1999, ApJ, 524, L29
• Newberg, H.J. 2002, ApJ, 569, 245
• Pilachowski, C. et al. 2012, arXiv:1211.0285
• Ross, N. et al. 2013. ApJ, 773, 14
• Ruhland, C. et al. 2011, ApJ, 731, 119
• Schlegel, D. J. et al. 2011, arXiv:1106. 1706
• Willman, B. 2010, AdAst, 21
• Yanny, B. et al. 2013, ApJ, 588, 824
• Yoon, J.H. et al. 2011, ApJ, 731, 58
• Zou, H. et al. 2014