Probing the environment around GW170817 with DESI: insights on galaxy group peculiar velocities for standard siren measurements

TL;DR

This work addresses the $H_0$ tension by leveraging DESI spectroscopy to map the environment around GW170817's host NGC 4993 and directly measure the group peculiar velocity using Fundamental Plane and Tully-Fisher calibrations. It builds a Bayesian standard siren posterior that combines GW luminosity distance with group PV measurements, carefully evaluating group membership and virialization to control systematics. The fiducial FP analysis yields $H_0 = 70.9^{+6.4}_{-8.5}$ km s$^{-1}$ Mpc$^{-1}$, with SBF-based PV delivering a tighter constraint of $H_0 = 73.4^{+3.3}_{-3.9}$ km s$^{-1}$ Mpc$^{-1}$, while TF-based results are less robust due to calibration issues; overall, the DESI results are consistent with Planck and SH0ES at the 2σ level. The study demonstrates the value of multiplexed spectroscopy for probing local dynamics in standard siren cosmology and projects that dozens of future GW events could push $H_0$ precision toward the 2% level, aided by DESI DR2 PV calibrations and complementary distance indicators.

Abstract

We present a new measurement of the Hubble constant, $H_0$, following the gravitational wave event GW170817 and Dark Energy Spectroscopic Instrument (DESI) observations. A standard siren measurement with a nearby (luminosity distance $\sim 40 $ Mpc) event such as GW170817 is typically sensitive to the peculiar motion of the host galaxy due to local dynamics. Previous measurements from this event have taken advantage of peculiar velocity measurements of nearby galaxies, including a handful of objects in the galaxy group that the host of the event, NGC 4993, has been associated with. Still, the group's properties and NGC 4993's membership were debated. We present DESI observations of thousands of galaxies in the vicinity of NGC 4993, resulting in 39 group galaxies and a five-fold increase in galaxies compared to previous observations with many of these galaxies contributing to a peculiar velocity measurement. Examining the local dynamics, our observations support the presence of a galaxy group of which NGC 4993 is part with a halo mass of order $\sim$$10^{13}~M_\odot$. Using peculiar velocity measurements from our Fundamental Plane galaxies observations, we find $H_0 =70.9^{+6.4}_{-8.5}$ km s$^{-1}$ Mpc$^{-1}$. In addition, using a peculiar velocity measurement for NGC 4993 from Surface Brightness Fluctuations in Cosmicflows-4 we find $H_0 =73.4^{+3.3}_{-3.9}$ km s$^{-1}$ Mpc$^{-1}$. We study the impact of different galaxy selection criteria on the determination of the peculiar velocity and, in turn, on the $H_0$ measurement. Our results highlight the importance of multiplexed spectroscopic observations of the environments of gravitational wave events to probe local dynamics, which can ultimately affect standard siren measurements.

Figures (6)

• Figure 1: On-sky depiction of DESI observations. Each blue circle indicates one of the 47,071 observations -- as targeted by our PV Survey and the DESI LOW-Z Survey -- in the on-sky vicinity of NGC 4993. Each red circle indicates a galaxy in the same galaxy group as NGC 4993 that is neither an FP nor a TF galaxy. Our TF group targets are demarcated by green squares, and the FP group targets are demarcated by yellow diamonds. One of the FP galaxies is NGC 4993 itself, indicated by a black cross. The approximate center of the galaxy group is designated by a black star.
• Figure 2: Peculiar velocity phase-space diagram of group galaxies. The x-axis displays each galaxy's projected distance from the brightest group galaxy (an approximation of the group center) where all galaxies are assumed to be at the same comoving distance. The y-axis indicates each galaxy's peculiar velocity offset by the weighted average peculiar velocity. We also display peculiar velocities from group galaxies in CF4 for comparison. We demarcate escape velocity envelopes for galaxy groups of differing halo masses along with their corresponding projected virial radii. Additionally, we delineate the projected virial radii measured from galaxy samples in kourkchi_ngc4993_pv and this work.
• Figure 3: Velocity- and redshift-space of observations. Left: For each galaxy, we plot its declination versus its heliocentric velocity (i.e. the speed of light multiplied by the heliocentric redshift), distinguishing our TF (green squares) and FP (yellow diamonds) targets from the DESI Low-Redshift Survey targets (blue circles). Previous work hjorth_2017 had observed a gap in redshifts from $3005$ km/s to $3169$ km/s. Our observations demonstrate that this gap does not exist. Right: A comparison between the chosen redshift bounds of our work (dashed) and previous work hjorth_2017 (dotted). Given our observations, we select boundary edges at the points where a gap first appear in redshift space.
• Figure 4: $H_0$ posteriors with varying PV samples. Lower panel: $H_0$ posteriors for a number of DESI galaxy sub-samples. See Section \ref{['subsection:pv_res']} for details about the selection criteria of each sub-sample. Upper panel: For each sub-sample, the posterior group-average log-distance ratio is plotted against the posterior $H_0$.
• Figure 5: $H_0$ posteriors with varying methods. Lower and upper panels are similar to those in Figure \ref{['fig:H0']}. The different posteriors derive from the fiducial posterior characterization ("Fiducial"), a parameterization that accounts for the comoving distance between NGC 4993 and the group center ("$d_L$ perturbation"), a posterior that incorporates a smoothing kernel into the log-distance ratio averaging ("Smoothing Kernel"), a posterior with the group centered on the coordinates of NGC 4993 ("NGC 4993 Center"), and a posterior with a smoothing kernel centered on NGC 4993 ("NGC 4993 Center & Smoothing Kernel"). The vertical bands represent the 1 and $2$ constraints from SH0ES and Planck, respectively.