- The SDSS-V SPIDERS hemisphere survey
- The SDSS-V/eFEDS pilot survey
- Clarification of SPIDERS+eFEDS scope and timeline
- Observational goals and requirements
- SPIDERS Hemisphere survey
- SDSS-V/eFEDS pathfinder survey
- Target selection and survey implementation
- Targeting generation ‘v0.5.3’
- Targeting generation ‘eFEDS plates’
- Spectroscopic data released in DR18
‘SPectroscopic IDentfication of ERosita Sources’
The SDSS-V SPIDERS project aims to characterize a large sample of X-ray sources through an ambitious program of optical spectroscopic follow up.
The SPIDERS project has two main science themes – the follow up of X-ray point-like sources (‘SPIDERS-AGN’), and the follow up of X-ray/optical cluster candidates ( ‘SPIDERS-Clusters’).
The SDSS-V/SPIDERS project will explore the X-ray source population through the wide and deep X-ray surveys performed with the SRG/eROSITA X-ray telescope (Merloni et al., 2012). In particular we focus on the sky hemisphere for which the Max Planck Institute for Extraterrestrial Physics (MPE) has data rights (approx. 180 < l < 360 deg ), and on the eROSITA performance validation observations of a smaller deeper pilot field (a.k.a. ‘eFEDS’).
The SDSS-V SPIDERS hemisphere survey
The hemisphere survey is the main component of the SPIDERS program. We seek to obtain highly complete (>90%) optical spectroscopic information (redshifts, classifications) for a large sample (>250,000) of point-like X-ray selected targets over >10,000 deg2 of high Galactic latitude sky. The SPIDERS hemisphere sample will reach to X-ray flux limits of 2×10-14 erg/cm2/s and to optical magnitudes of i ~ 21.3, r~21.5 AB. In addition, over the same sky area, we will obtain redshifts for >50,000 cluster galaxies, selected via their X-ray and optical properties.
The SDSS-V/eFEDS pilot survey
The SDSS-V/eFEDS survey is a pathfinder component of the BHM survey program, led by the SPIDERS team. The SDSS-V/eFEDS survey exploits early performance validation observations from the SRG/eROSITA X-ray telescope in the ‘eROSITA Final Equatorial Depth Survey’ field (eFEDS, Brunner et al., 2022). The eFEDS X-ray footprint covers around 140 sq. degrees centered near RA,Dec = 135,+1 deg, and includes the ‘GAMA09’ Field (Driver et al. 2008). In SDSS-V we used pre-FPS (i.e. plate mode) observations at the Apache Point Observatory SDSS/2.5m telescope to collect optical spectroscopy for counterparts to point-like and extended X-ray sources (Salvato et al., 2022; Klein et al., 2022). The SDSS-V/eFEDS sample reaches to an X-ray flux limit approximately twice as deep as that pursued by the Hemisphere survey, whilst the optical magnitude limit is comparable. The SDSS-V/eFEDS program builds upon earlier efforts (by our team and others) to obtain spectroscopy for galaxies, AGN and stars in this well-known survey field.
The full eFEDS spectroscopic sample, including more than 12k X-ray sources with secure spectroscopic redshifts and classifications, will be presented by Merloni et al (in prep.).
Clarification of SPIDERS+eFEDS scope and timeline
The SPIDERS project was originally conceived in response to the call for ideas for a program to follow on from SDSS-III. At that time (c. 2013), it was expected that SRG/eROSITA would shortly be launched and start providing targets for observation within/alongside the SDSS-IV/eBOSS survey. Before the launch of SRG, initial SDSS-IV/SPIDERS target selection was therefore based on ROSAT and XMM-Newton Slew Survey X-ray catalogues (Dwelly et al., 2007; Clerc et al., 2016) with an expectation that these would be superseded by new eROSITA data . However, due to delays, eROSITA only started gathering all-sky survey data in Late 2019, a few months before the nominal end of SDSS-IV/eBOSS operations (Early 2020). In order to recover some of the SPIDERS science goals, we carried out a mini-survey of the eFEDS field towards the end of the SDSS-IV survey (Northern Spring 2020). That project was interrupted by the COVID-19 pandemic, and only a small fraction of the expected data set was collected. The SPIDERS project has been continued and updated for SDSS-V, forming one pillar of the SDSS-V/Black Hole Mapper project. SDSS-V/SPIDERS includes early plate-based observations to recover and expand upon the SDSS coverage of the eFEDS field (Dec 2020-May 2021), followed by very wide area coverage of the high-latitude eROSITA sky using the new twin SDSS-V/FPS robotic fiber positioner systems (2022-).
Observational goals and requirements
SPIDERS Hemisphere survey
The primary observational goal of the SPIDERS hemisphere program is to obtain a highly complete set of reliable spectroscopic redshifts and classifications for optical/IR counterparts to eROSITA X-ray sources.
The choice of cadence is driven by the total expected BOSS exposure time that is required to achieve reliable redshifts. Brighter targets have lower exposure time requests.
SPIDERS will spectroscopically target >250,000 optically bright (r < 21.5 or i < 21.3 AB) counterparts of X-ray sources, distributed over a high Galactic latitude survey footprint >10,000 deg2. SPIDERS will use the part of the eROSITA All Sky Survey data set which lies in the ‘MPE’ hemisphere (approx. 180<l<360 deg). Initial SPIDERS Hemisphere target selection is based on ‘eRASS1’ depth data (first 6-months eROSITA scanning data), but will be extended to ‘eRASS:3’ data (18 months depth), reaching to a X-ray flux limit of ∼2×10-14 erg cm-2 s-1 (in the 0.5-2 keV band). We expect X-ray AGN to be the dominant source of counterparts. However, we will also obtain optical spectra of ∼104 X-ray emitting clusters of galaxies (typically targeting 3-10 member galaxies per cluster), for cluster astrophysics and cosmology studies.
Target priorities are allocated within the SPIDERS ‘AGN-centric’ samples according to several factors, including: the X-ray detection likelihood, the cross-match method and reliability, the availability of an earlier good-quality SDSS spectrum, and Galactic latitude. SPIDERS clusters galaxy targets are prioritized according to the relative probability of them being a cluster member, with candidate BCGs given first priority.
SPIDERS targets need to reach a sufficient SNR to guarantee a very high redshift success rate (spectroscopic completeness). Based on SDSS-III and -IV experience, we estimate a requirement of a median SNR/pixel> (2.9, 3.9) in the r, i-bands respectively, per target (to guarantee >97% redshift success for AGN). Simple scaling from BOSS performance suggest that nominal exposure times of (3-4)x15 minutes in dark-time will be sufficient for SPIDERS targets, given the magnitude limits of (r < 21.5 or i < 21.3 AB).
SDSS-V/eFEDS pathfinder survey
The primary observational goal for SDSS-V observations in the eFEDS field was to achieve a high spectroscopic completeness (i.e reliable redshifts, classifications) for counterparts to X-ray sources detected as part of the eROSITA/eFEDS performance verification survey, particularly for the eFEDS X-ray sources having counterparts in the magnitude range 16 < r < 22 AB . This goal was constrained by the number of hours of dark observing time expected to be available to the project, a general desire to minimize the total number of drilled plug plates, and the overall capabilities of the SDSS-V plate system to deliver fibers to naturally clustered targets.
The goal of high completeness was assisted by the wealth of previous spectroscopic survey data in the eFEDS field, both from previous SDSS generations and from other projects. However, these existing data did make the plate design process more complex. Our strategy was therefore to prioritize targets which did not have existing high quality spectroscopic observations. However, targets with existing spectroscopy were still eligible to receive a fiber, but at a lower priority level to fresh targets.
The SDSS-V/eFEDS observations used the new joint BOSS+APOGEE configuration. In this mode, the focal plate is populated with 500 fibers feeding a single BOSS optical spectrograph, and 300 fibers feeding the near-infrared APOGEE spectrograph. New procedures and software were put in place to manage target selection, fiber assignment and plate design for this mode, these will be described in more detail in future work. Here we provide a summary of special considerations that apply specifically to the SDSS-V/eFEDS plates.
The design process for the SDSS-V/eFEDS plates is described here.
Target selection and survey implementation
As part of the SDSS-V DR18 targeting release, we are providing lists of SPIDERS hemisphere targets (optical counterparts to eRASS:1 X-ray catalogs), but not the X-ray data from which they were selected. The DR18 SPIDERS-hemisphere targeting release consists of several target cartons, in which X-ray targets were associated with longer wavelength counterparts via several supporting optical/IR catalogs. The optical/catalogs considered were: legacysurvey.org/dr8 (lsdr8, Dey et al., 2018), Pan-STARRS1-dr2 (ps1dr2, Chambers et al., 2016), Gaia dr2 (Gaia collaboration, 2018), SkyMapper dr2 (Onken et al., 2019) and SuperCosmos (Hambly et al., 2001). For the ps1dr2, gdr2, skymapperdr2 and SuperCosmos samples, the mid-IR CatWISE2020 catalog (Marocco et al., 2021) provides a stepping stone to get from X-ray source positions to optical counterparts. There is a high degree of overlap between the targets selected via these independently performed target selection algorithms.
The SPIDERS clusters hemisphere target cartons rely on either legacysurvey.org/dr8 or Pan-STARRS1-dr2 photometric catalogues.
Reminder. The SPIDERS hemisphere survey is initially based on eRASS:1 depth eROSITA data, but will evolve to use eRASS:3 in the next update of targeting.
Targeting generation ‘v0.5.3’
This targeting generation was used during initial SDSS-V/FPS operations (2022-), below is a list of SPIDERS cartons that were considered:
These SPIDERS target catalogues can be found in the targeting release for SDSS DR18
Targeting generation ‘eFEDS plates’
This generation of targeting was used for eFEDS plates during one season of SDSS-V/plate operations (Dec 2020-May 2021). The parent catalogues from which the eFEDS targets were selected were derived from early reductions of the eROSITA/eFEDS X-ray data set, and on early attempts to match those X-ray sources to longer wavelength counterparts provided by several supporting optical/IR catalogs, including legacysurvey/dr8, SDSS photObj dr13, and HSC-SSC dr2 (Aihara et al., 2019). The eFEDS science target selection process concentrates on AGN candidates (one target carton) and galaxy cluster candidates (4 cartons).
Below we list the SPIDERS/eFEDS target cartons which were considered in SDSS-V/eFEDS plates:
These link to sections within: BHM Cartons
Targets from the following MWM-led cartons were also included in the eFEDS plates (please see MWM Programs for further details):
The following calibration cartons (blank sky locations and spectrophotometric standard stars) were also used in the eFEDS plates:
As part of the SDSS-V plate design process, it was necessary to label subsets of the targets within these cartons with “-boost1”, “-boost2” and “-boost3” suffixes. This re-labeling was required in order that the plate design software would correctly manage the order of target assignment in spatially overlapping plates. Those carton names are propagated into the data products.
Note that we are not releasing these eFEDS target catalogues in SDSS DR18 – they are expected to form part of the targeting documentation in SDSS DR19.
Spectroscopic data released in DR18
SDSS DR18 contains new optical (BOSS) spectroscopic data for (primarily) X-ray selected targets observed in the eFEDS field.
In SDSS DR18 we are releasing new SDSS-V plate-based optical spectroscopic data collected during the period Oct 2020-May 2021. These new SDSS-V/eFEDS observations seek to complete and expand upon the earlier SDSS-IV/eFEDS project, which was released in SDSS DR17 (Abdurro’uf et al., 2021). A total of 37 SDSS-V joint BOSS+APOGEE plates were observed during during the period Oct 2020-May 2021. Only the BOSS data is released in DR18, which we discuss below.
Our observing strategy was to prioritize coverage over depth – given the somewhat uncertain quantity of good quality observing time that would be available whilst the eFEDS field was visible from APO. By the end of the observing season, we had observed all plates to at least a total fiducial SN2_G of 7.5. However, most of the plates are significantly deeper than that , with 33/37 of the eFEDS plates having SN2_G > 10 (comparable to the the exposure depth in BOSS/eBOSS), and more than half the plates having SN2_G>20. The fiducial SN2 in the i-band is typically twice that measured in the g-band.
Several plates were observed on more than one MJD (and through more than one plugging); some targets were observed on more than one plate. The plates were reduced with the BOSS pipeline. The standard BOSS idlspec2d algorithm has been used to generate a set of spectral data products per Plate-MJD combination, with individual 1D spectra per Field-MJD-CATALOGID combination, following the usual conventions.
In addition, a specialized co-adding algorithm (within idlspec2d) was implemented for these eFEDS plates, which gathers data together across Plates and MJDs, in order to increase SNR in the derived spectral products. This routine creates a single stacked spectrum per astrophysical object, using all available (and suitable) SDSS-V spectroscopic data. A total of 16548 of these specially co-added spectra are released in DR18 (13269 science targets, 2608 skies and 671 spectrophotometric standards). Of the co-added spectra associated with science targets, 12446 are primarily associated with BHM/SPIDERS target cartons, and the remaining 823 are primarily associated with MWM/Compact Binary and MWM/White Dwarf cartons.
Redshifts, classifications and quality flags were computed automatically on the specially co-added spectra using the idlspec1d template fitting pipeline. A warning flag was set for 669/13269 science spectra, indicating a problem with the data or the fit. Of the 12600 science spectra without warnings, the main pipeline gives source classifications of ‘QSO’ for 6204, ‘GALAXY’ for 4782 and ‘STAR’ for 1614 spectra. Pipeline redshifts for non-flagged ‘QSO’ and ‘GALAXY’ classified spectra span the range 0.0 < z < 4.5, with 90% of these objects falling in the range 0.14 < z < 2.55 (median 0.55).
In order to increase the completeness and reliability of the redshifts/classifications derived from these spectra, the SPIDERS team has carried out visual inspections of a large fraction of the SDSS-V/eFEDS spectra. Those inspections have been combined with earlier SDSS spectroscopic redshift information, and non-SDSS redshift information gathered from the literature. Please see Merloni et al (in prep) and the associated Value Added Catalogue to take advantage of this compilation.