Preliminary Target Selection for the DESI Emission Line Galaxy (ELG) Sample

DESI (DESI Collaboration et al. 2016) will measure spectroscopic redshifts for ∼35 million galaxies and quasars over ∼80% of cosmic history and one third of the sky. DESI will target multiple extragalactic tracers optimized for different redshift ranges, supplemented by significant stellar samples for calibration and Galactic science. Relaxed selections will be tested during a preliminary "Survey Validation" phase to validate and optimize targeting for the DESI "main" survey.

This note outlines preliminary targeting for an Emission Line Galaxy (ELG) sample in the redshift range 0.6 < z < 1.6, which will constitute approximately half of DESI extragalactic tracers. DESI exploits the abundance of star-forming galaxies at z ∼ 1–2 to target ELG tracers with a spectroscopic redshift that can be measured reasonably quickly. The high star formation rate at these redshifts produces identifiable emission lines without the need to detect a strong continuum. A key spectroscopic diagnostic is the [O ii] doublet at λλ3726,3729: the DESI spectrographs are designed to resolve this feature over the targeted ELG redshift range. ELGs, which have been used as tracers in previous surveys (e.g., WiggleZ and eBOSS), will also underpin future Baryon Acoustic Oscillations (BAO) surveys, such as PFS (Takada et al. 2014) and Euclid (Laureijs et al. 2011).

The target selection will use grz imaging from the Legacy Surveys (Dey et al. 2019). Results presented here are based on Data Release 8 of the Legacy Surveys. ²⁴ The DESI footprint is split into two regions; "north" (Galactic b > 0° and decl. > 32375) and "south"; the "north" has a slightly different photometric system and is shallower (0.5 magnitudes) in the g- and r-bands. We therefore define slightly different cuts for the "north" and "south."

First, we require a minimum photometric quality by enforcing at least one observation and a positive signal-to-noise ratio in each of g-, r-, and z-band. We also require that targets are not in corrupted pixels, nor near bright or medium-bright stars, globular clusters, or large galaxies (MASKBITS is not set for bits 1, 5, 6, 7, 11, 12 or 13).

Next, we apply the following cuts in grz (see Figure 1):

with (g_max, zpt) = (23.6, −0.35) for the "north" and (g_max, zpt) = (23.5, −0.15) for the "south." All magnitudes are corrected for Galactic extinction using the Schlegel et al. (1998) maps. Equations 1(b) and (c) select targets in the desired redshift range and Equation 1(d) favors star-forming galaxies. As the photometry is noisier in the "north," our selection box is farther from the low-redshift locus to avoid significant contamination from z < 0.6 galaxies. Equation 1(a) targets the requisite [O ii] flux (see e.g., Comparat et al. 2015) and also sets the density to ∼2400 deg⁻².

Zoom In Zoom Out Reset image size

As no spectroscopic truth table exists for a complete sample with g ≲ 23.5, we assess our selection using HSC/PDR2 DEmP photometric redshifts (z_phot; Aihara et al. 2019) for the redshift distribution, and DEEP2 spectroscopic data over 0.8 < z < 1.4 for the [O ii] flux (Newman et al. 2013). The HSC/PDR2 z_phot cover ∼100 deg² in the "north" and ∼200 deg² in the "south." The z_phot are estimated from deep grizy-photometry and are of exquisite quality for z < 1.6 ELGs when compared to spectroscopy from eBOSS and from DESI Pilot Observations with the MMT (Karim et al. 2020; Raichoor et al. 2020). Figure 1 shows the z_phot distribution of our ELG targets, demonstrating that z_phot > 1.0 objects are efficiently selected; overall, ∼80% of the selection has 0.6 < z_phot < 1.6 for both "north" and "south."

Finally, we characterize [O ii] flux for our selection using measurements from Comparat et al. (2015) in DEEP2 over 0.8 < z < 1.4, where DEEP2 is complete over all fields for our g < 23.5–23.6 ELG sample. We find that 76% ("north") and 83% ("south") of our targets have sufficient [O ii] flux for a secure spectroscopic redshift measurement given the expected DESI specifications.

This note outlines a preliminary DESI ELG selection based on a g-band cut and a (g − r) versus (r − z) color–color box that produces ∼2400 deg⁻² targets. Analyses using HSC/PDR2 z_phot and DEEP2 [O ii] flux show that ∼65% of resulting ELGs will provide a reliable spectroscopic redshift within 0.6 < z < 1.6, in accord with DESI Collaboration et al. (2016). Preliminary ELG targeting will be tested during DESI "Survey Validation" to inform a final selection for the DESI "main" survey. Target catalogs that use the selection described in this note are public. ²⁵

A.R. acknowledges support from the ERC advanced grant LIDA and from the SNF grant 200020_175751. This research is supported by the Director, Office of Science, Office of High Energy Physics of the U.S. Department of Energy under Contract No.DE–AC02–05CH1123, and by the National Energy Research Scientific Computing Center, a DOE Office of Science User Facility under the same contract; additional support for DESI is provided by the U.S. National Science Foundation, Division of Astronomical Sciences under Contract No. AST-0950945 to the NSF's National Optical-Infrared Astronomy Research Laboratory; the Science and Technologies Facilities Council of the United Kingdom; the Gordon and Betty Moore Foundation; the Heising-Simons Foundation; the French Alternative Energies and Atomic Energy Commission (CEA); the National Council of Science and Technology of Mexico; the Ministry of Economy of Spain, and by the DESI Member Institutions. The authors are honored to be permitted to conduct astronomical research on Iolkam Du'ag (Kitt Peak), a mountain with particular significance to the Tohono O'odham Nation.