The DREAMS Project: Disentangling the Impact of Halo-to-Halo Variance and Baryonic Feedback on Milky Way Dark Matter Speed Distributions
Ethan Lilie, Jonah C. Rose, Mariangela Lisanti, Alex M. Garcia, Paul Torrey, Kassidy E. Kollmann, Jiaxuan Li (李嘉轩), Olivia Mostow, Bonny Y. Wang, Stephanie O’Neil, Xuejian Shen, Alyson M. Brooks, Arya Farahi, Nitya Kallivayalil, Lina Necib, Andrew B. Pace, Mark Vogelsberger · 2025
Read The DREAMS Project: Disentangling the Impact of Halo-to-Halo Variance and Baryonic Feedback on Milky Way Dark Matter Speed Distributions by Ethan Lilie, Jonah C. Rose, Mariangela Lisanti, Alex M. Garcia, Paul Torrey, Kassidy E. Kollmann, Jiaxuan Li (李嘉轩), Olivia Mostow, Bonny Y. Wang, Stephanie O’Neil, Xuejian Shen, Alyson M. Brooks, Arya Farahi, Nitya Kallivayalil, Lina Necib, Andrew B. Pace, Mark Vogelsberger (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-04157.txt.
Introduction
Dark matter detectors deep underground listen for particles whose speeds near Earth decide what the detectors can hear, yet those speeds are uncertain. Ethan Lilie and a large team quantify that uncertainty in a 2025 astrophysics preprint from the DREAMS project. They simulate more than a thousand Milky Way-mass halos, varying initial conditions along with cosmological and astrophysical parameters.
A normalising-flows emulator trained on the suite lets the authors apportion blame among the unknowns. Halo-to-halo variance dominates the spread, with host halo mass second. Feedback from supernovae and black holes, often worried over, turns out negligible by comparison. Their prediction still contains the standard halo model within its error bars, though individual galaxies can stray from it. Applied to the XENON1T experiment, the astrophysical spread proves comparable to the experimental one, a calibration target for the whole direct detection field.
Catalog record
| CREATOR | Ethan Lilie, Jonah C. Rose, Mariangela Lisanti, Alex M. Garcia, Paul Torrey, Kassidy E. Kollmann, Jiaxuan Li (李嘉轩), Olivia Mostow, Bonny Y. Wang, Stephanie O’Neil, Xuejian Shen, Alyson M. Brooks, Arya Farahi, Nitya Kallivayalil, Lina Necib, Andrew B. Pace, Mark Vogelsberger |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 87 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-04157
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The DREAMS Project: Disentangling the Impact of Halo-to-Halo Variance and Baryonic Feedback on Milky Way Dark Matter Speed Distributions Ethan Lilie, Jonah C. Rose, Mariangela Lisanti, Alex M. Garcia, Paul Torrey, Kassidy E. Kollmann, Jiaxuan Li (李嘉轩), Olivia Mostow, Bonny Y. Wang, Stephanie O’Neil, Xuejian Shen, Alyson M. Brooks, Arya Farahi, Nitya Kallivayalil, Lina Necib, Andrew B. Pace, Mark Vogelsberger Abstract Direct detection experiments require information about the local dark matter speed distribution to produce constraints on dark matter candidates, or infer their properties in the event of a discovery. In this paper, we analyze how the uncertainty in the dark matter speed distribution near the Sun is affected by baryonic feedback, halo-to-halo variance, and halo mass. To do so, we harness the statistical power of the new DREAMS Cold Dark Matter simulation suite, which is comprised of 1024 zoom-in Milky Way-mass halos with varied initial conditions as well as cosmological and astrophysical parameters. Applying a normalizing flows emulator to these simulations, we find that the uncertainty in the local DM speed distribution is dominated by halo-to-halo variance and, to a lesser extent, uncertainty in host halo mass. Uncertainties in supernova and black hole feedback (from the IllustrisTNG model in this case) are negligible in comparison. Using the DREAMS suite, we present a state-of-the-art prediction for the DM speed distribution in the Milky Way. Although the Standard Halo Model is contained within the uncertainty of this prediction, individual galaxies may have distributions that differ from it. Lastly, we apply our DREAMS results to the XENON1T experiment and demonstrate that the astrophysical uncertainties are comparable to the experimental ones, solidifying previous results in the literature obtained with a smaller sample of simulated Milky Way-mass halos. \NewPageAfterKeywords I Introduction A central goal of modern astrophysics and particle physics is to identify the nature of dark matter (DM). Direct detection experiments represent a key strategy in this effort (see, e.g., ddreview for a review). These experiments search for DM particles interacting with Standard Model particles in a terrestrial target. The scattering rates depend on the accurate determination of the phase-space distribution of DM near the Sun, in particular its speed distribution. However, the speed distribution is sensitive to the detailed structure of the Galaxy, the host halo mass, its formation history, the location of the Solar circle, and the treatment of baryonic physics ( newdm1 ; lmc1 ; newdm4 ; newdm3 ; 1 ; lmc2 ; lmc3 ; Staudt_2024 ; folsom2025darkmattervelocitydistributions ) . The Standard Halo Model (SHM; see, e.g., shm1 ; shm2 ; shm3 ), which was first posited by vlsr1 , is often assumed for the local DM speeds. It is a Maxwell–Boltzmann distribution truncated at the Galactic escape speed ( found_shm ; found_shm2 ) . While a distribution like the SHM is expected for a virialized halo, deviations can arise depending on how a halo evolves in a cosmological setting. To this end, cosmological simulations that track the merger history of Milky Way (MW)-mass halos play an important role in updating our theoretical understanding for the speed distribution. Initial DM-only simulations of MW-mass halos predicted speed distributions that were typically slower than that expected from the SHM ( earlydm1 ; earlydm2 ; earlydm6 ; earlydm4 ; earlydm5 ; earlydm3 ) . This was largely due to the omission of a baryonic disk, which would otherwise deepen the gravitational potential and increase the DM speeds ( baryons_dm1 ; baryon1 ; baryon2 ; baryon3 ; baryon4 ; sales2022baryonicsolutionschallengescosmological ; Garcia_to_Appear ) . The next generation of fully hydrodynamic simulations did indeed find that the DM speed distributions shifted to higher values, closer to the local standard-of-rest speed in our Galaxy ( newdm2 ; newdm6 ; newdm7 ; newdm5 ) . That being said, there was still significant uncertainty comparing the high-speed tails of these distributions with the SHM expectation—especially with regards to the impact of different simulation models for baryonic feedback physics ( newdm2 ; baryon2 ; newdm8 ; newdm6 ; newdm7 ; newdm5 ; newdm1 ; newdm4 ; newdm9 ; newdm3 ; Lawrence_2022 ; baryon4 ) . A main challenge in robustly inferring speed distributions from simulations is the limited statistics, with studies often focusing on one or a few simulated galaxies due to computational cost ( newdm2 ; newdm8 ; newdm6 ; newdm7 ; newdm5 ; newdm1 ; newdm4 ; newdm9 ; newdm3 ) . folsom2025darkmattervelocitydistributions studied the largest sample to date, using 98 TNG50-1 galaxies ( Nelson_2019 ) . Their study was the first to quantify the degree to which intrinsic halo-to-halo variance—which arises from the unique merger history and evolution of a given halo—affects predictions for the local DM speed distribution. Moreover, choosing what is considered a MW analog is non-trivial. Previous works have attempted to account for the unique MW merger history by including a satellite like the Large Magellanic Cloud (LMC) or a merger like Gaia Sausage–Enceladus (GSE) ( lmc1 ; lmc2 ; lmc3 ; newdm4 ; folsom2025darkmattervelocitydistributions ) . While some find potential impacts on the high-speed tails of the distributions, others find less notable effects. At this stage, it is difficult to draw definitive conclusions regarding the impact of the merger history given the limited statistics involved in these studies. The likelihood of finding a MW-mass galaxy with both an LMC and GSE is \lesssim 1\% ( nadler ) , which underscores the computational challenge in generating large-enough samples that match specific details of the MW’s evolution. In addition to matching the merger history, one would also want to ensure that the morphology of the simulated galaxy is similar to our own. This is particularly important because the MW is compact compared to other galaxies of a similar mass ( compact1 ; compact2 ; compact3 ; compact4 ) —although see noncompact . Therefore, naively setting the solar circle in the simulation to its observed value of 8.3 kpc in the MW ( sun1 ; sun2 ) can potentially underestimate the DM speeds. newdm8 addressed this by only considering simulated galaxies with rotation curves that match our own, which further limited their sample size. In contrast, Staudt_2024 used an inference procedure to determine properties of the speed distribution consistent with the MW’s observed circular velocity and folsom2025darkmattervelocitydistributions set the solar circle by using an energy-conserving transformation to match the mass enclosed at 8.3 kpc to that of our Galaxy. These approaches are advantageous as they utilize all the simulations in a given suite. This paper uses the new DREAMS suite of 1024 MW-mass halos ( Rose_Centrals_to_Appear_A ) from the DREAMS project ( rose2024introducingdreamsprojectdark ) which varies over halo initial conditions and the IllustrisTNG (TNG; tng2 ) baryonic physics parameters. This is one of the largest simulation suites of high-resolution MW-mass galaxies in Cold Dark Matter (CDM) and spans a wide range of merger histories and galaxy morphologies. This unprecedented dataset allows us to robustly understand the interplay of different uncertainties on the local DM speed distribution. In particular, using a machine-learning-based approach, we isolate the effects of feedback parameters and halo mass on the DM speeds and also quantify the degree to which halo-to-halo variance affects the results. This approach allows us to learn the relationship between MW properties and the DM speed distribution, and