1 of 12

What do the Milky Way’s stellar streams teach us about dark matter?

Vedant Chandra

Background: The Aquarius Project

2 of 12

The Underlying Physics of Dark Matter

cold dark matter

subhalos down to < 1 M scales

warm dark matter

subhalos suppressed below ~107 M

Bullock & Boylan-Kolchin (2017)

3 of 12

in the galaxy

on-sky

in radial velocities

see Erkal+16, Bonaca+19, and many more

4 of 12

I. DM from Stream Gaps

  • Reconstructing individual subhalo impacts from stream morphology �
  • Photometry + astrometry can be sufficient, spectroscopy helps with membership

Bonaca et al (2020)

see also Banik+21

5 of 12

II. DM from Stream Kinematics

  • The velocity distribution of stream stars encodes the history of all past encounters �
  • Requires a good understanding of the ‘unperturbed’ velocity distribution (see Anya’s talk)�
  • Stable velocities are required to detect binary motion and measure individual impact signatures

Nibauer, Bonaca �et al (2026)

6 of 12

III. Anatomy of Individual Impacts

(from the Via Data Simulator)

7 of 12

Modeling the MW’s Streams and Subhalos

simulating the growth and evolution of every known stellar stream

simulating a realistic population of MW subhalos with semi-analytic models

8 of 12

grey: DM subhalos

colors: stream stars

red: ‘detected’ subhalos

9 of 12

The Most Promising DM Detectors

# of subhalo encounters

# of subhalo encounters

10 of 12

Simulating ‘Perturbed’ Streams

Chandra+ in prep

black: unperturbed

red: with CDM subhalos

On-Sky Positions

Radial Velocities

11 of 12

Simulating Stellar Streams as DM Detectors

Chandra+ in prep

🡨 ‘dark’ subhalos 🡪

12 of 12

Summary

  • The detailed structure and velocities of stellar streams encode the abundance and character of dark substructure in the Milky Way�
  • The Via Project is developing a new instrument and survey to map dark matter substructure around the Milky Way�
  • We are developing end-to-end forecasts of stream-subhalo interactions, generating mock Via data, and developing frameworks to infer the nature of DM