The DESI DR1 Peculiar Velocity Survey: The Tully-Fisher Distance Catalog

TL;DR

This work calibrates the Tully-Fisher relation using 10,262 DESI DR1 PV galaxies by measuring $V(0.4R_{26})$ and performing a joint, multi-bin fit of the TF relation in 14 redshift slices from $0.03<z<0.10$, yielding a slope of $a=-7.22\pm0.01$ AB mag and intrinsic scatter $\sigma=0.466\pm0.001$ AB mag in the $r$-band. The calibrated TF is then used to derive redshift-independent distances and peculiar velocities, producing a DESI DR1 TF PV catalog and a clustering catalog with $N=6{,}807$ galaxies (plus random mocks) for large-scale structure analyses. The zero-point of the TF scale is anchored to external distance calibrators via a joint fit with SH0ES/Pantheon+ distances and galaxy groups, enabling robust PV and $H_0$ constraints. The DR1 TF catalog expands the TF galaxy sample to higher redshift than previous surveys while maintaining comparable scatter, providing substantial leverage for growth-rate studies and cosmological insights in the local universe.

Abstract

We calibrate the Tully-Fisher relation (TFR) using observations of spiral galaxies taken during the first year (DR1) of the DESI galaxy redshift survey. The rotational velocities of 10,262 galaxies are measured at 0.4 R26 by comparing the redshifts at 0.4 R26 with those at the galaxy centers of spatially-resolved galaxies targeted as part of the DESI Peculiar Velocity Survey. The DESI DR1 TFR slope is calibrated by separating the spiral galaxies into redshift bins of width dz = 0.005 from 0.03 < z < 0.1 and jointly fitting the TFR across all bins. We find a slope of -7.22+/-0.01 AB mag in the r-band for the TFR, with an intrinsic scatter of 0.466+/-0.001 AB mag. We present a catalog of the distances and peculiar velocities to these 10,262 galaxies using our calibrated TFR. For cosmological analyses, we also present a clustering catalog and associated random catalogs using a subset of 6807 of the DESI DR1 TF galaxies.

Figures (13)

• Figure 1: Observed correlation between $r$-band apparent magnitude and axis ratio. A linear fit to the median magnitudes in each bin (dark blue crosses) is shown in the black line.
• Figure 2: Left: Pull distribution of the difference in $V(0.4R_{26})$ for pairs of observations on the same galaxy (on the same or opposite side of the galaxy center). Right: Distribution in $\Delta V / V_{\rm min}$ for galaxies with multiple measures of $V(0.4R_{26})$. For a given galaxy, all its pairs must fall to the left of the dotted orange line for the galaxy to be in the final sample; those that do are shown in the black hashed histograms.
• Figure 3: Difference in the velocity at $0.4R_{26}$ as measured in the DESI PV Survey and calculated from the SDSS MaNGA velocity field fits from Ravi2023 as a function of the difference in the photometric position angle from the SGA and the kinematic position angle from the velocity field fits from Ravi2023.
• Figure 4: Rotational velocities at $0.4R_{26}$ as measured in the DESI PV Survey and calculated from the the SDSS MaNGA velocity field fits from Ravi2023 corrected for the differences in inclination and position angle between the two measurements. The left figure compares the two velocities with the black dotted line indicating where the velocities are equal. The center histogram shows the perpendicular distance of each galaxy to $y = x$, $\Delta V(0.4R_{26})$, normalized by the average of the two velocity measurements, the expected value on $y = x$. The right histogram shows the pull distribution of $\Delta V(0.4R_{26})$. The black dashed line shown a Gaussian fit to the pull distribution with a mean of $-0.07$ and standard deviation of 1.25.
• Figure 5: Uniform Manifold Approximation and Projection (UMAP) of the DESI DR1 TF sample. Top: We show the cutouts of a random subset of 1,718 galaxies from the full sample of 10,262 DESI DR1 TF galaxies + 4000 anchor galaxies, to demonstrate how the classifier sorts the galaxies by similar appearance. This subset of galaxies are a mix of galaxies with unknown classifications and anchor galaxies of Elliptical, Spiral, Lenticular, or Irregular classifications, as shown in the bottom left plot. Bottom left: Complete UMAP of the full 10,262 DESI DR1 TF sample (shown as black points) along with the 4000 anchor galaxies with known morphological classifications (Irregulars in pink, Lenticulars in blue, spirals in olive, and ellipticals in brown). A random subset of this projection is shown on top with cutouts of the galaxy images to show how the classifier shorts the images based on their appearance. We see that the anchor ellipticals are found grouped together towards the top right, anchor lenticulars in the top left, anchor irregulars in the bottom left, and anchor spirals are spread across the middle and right. The galaxies being classified (black points) sit primarily across the central and right regions of this projection, where the anchor spirals also sit, so the vast majority of the DESI DR1 TF sample are spiral galaxies. Bottom right: Result after the first classification iteration, showing the anchor galaxies and the newly classified galaxies in their corresponding colors. Those galaxies which have multiple potential classifications (because they sit in the overlap regions between the anchor points) are shown in black and will be evaluated in the second iteration.