pop-cosmos: Scaleable inference of galaxy properties and redshifts with a data-driven population model

TL;DR

This paper introduces pop-cosmos, a scalable Bayesian method using a data-driven population model to accurately estimate galaxy redshifts and properties from photometric data, enabling analysis of large galaxy catalogs efficiently.

Contribution

The paper presents a novel, GPU-accelerated Bayesian inference framework with a score-based diffusion prior for galaxy properties, significantly improving scalability and accuracy over previous methods.

Findings

Achieves minimal bias and high accuracy in redshift estimation for COSMOS2020 galaxies.

Demonstrates efficient inference on over 290,000 galaxies with GPU acceleration.

Generalizes well to fainter galaxies beyond the training set.

Abstract

We present an efficient Bayesian method for estimating individual photometric redshifts and galaxy properties under a pre-trained population model (pop-cosmos) that was calibrated using purely photometric data. This model specifies a prior distribution over 16 stellar population synthesis (SPS) parameters using a score-based diffusion model, and includes a data model with detailed treatment of nebular emission. We use a GPU-accelerated affine invariant ensemble sampler to achieve fast posterior sampling under this model for 292,300 individual galaxies in the COSMOS2020 catalog, leveraging a neural network emulator (Speculator) to speed up the SPS calculations. We apply both the pop-cosmos population model and a baseline prior inspired by Prospector-$\alpha$, and compare these results to published COSMOS2020 redshift estimates from the widely-used EAZY and LePhare codes. For the $\sim 12,000$ galaxies with spectroscopic redshifts, we find that pop-cosmos yields redshift estimates that have minimal bias ($\sim10^{-4}$), high accuracy ($\sigma_\text{MAD}=7\times10^{-3}$), and a low outlier rate ($1.6\%$). We show that the pop-cosmos population model generalizes well to galaxies fainter than its $r<25$ mag training set. The sample we have analyzed is $\gtrsim3\times$ larger than has previously been possible via posterior sampling with a full SPS model, with average throughput of 15 GPU-sec per galaxy under the pop-cosmos prior, and 0.6 GPU-sec per galaxy under the Prospector prior. This paves the way for principled modeling of the huge catalogs expected from upcoming Stage IV galaxy surveys.