🔭 astrophysics

LISA's view of the Galactic Halo: forecasts for the Galactic double white dwarf population using Gaia data

This paper uses Gaia-informed Galactic models to forecast that while incorporating the metal-rich Gaia-Sausage-Enceladus population significantly alters the chirp mass and distance distributions of double white dwarfs resolvable by LISA, it does not affect the overall strength or height of the gravitational wave foreground.

Original authors: Ann-Marsha Alexis, Katelyn Breivik

Published 2026-07-14

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ann-Marsha Alexis, Katelyn Breivik

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the Milky Way as a giant, bustling city of stars. For years, astronomers have been trying to map out the "ghosts" of this city: the double white dwarfs. These are pairs of dead stars, the dense, cooling embers of stars that used to shine, now locked in a tight, invisible dance. They are so close they whisper to each other through gravitational waves, ripples in the fabric of space-time.

In the 2030s, a space mission called LISA (Laser Interferometer Space Antenna) will launch to listen to these whispers. But to know what LISA will hear, scientists first have to guess where these ghostly couples are hiding and how heavy they are.

The Old Map vs. The New Map

Until now, the "old map" of our galaxy's history assumed the halo (the giant, fuzzy sphere of stars surrounding the city) was made of one big, ancient, and very metal-poor family. Think of it like a neighborhood where everyone is from the same very old, very poor village.

But recently, the Gaia satellite gave us a new set of blueprints. It revealed that the halo isn't just one family; it's a mix. A huge chunk of it comes from a "Gaia-Sausage-Enceladus" (GSE) event—a massive galaxy that crashed into the Milky Way billions of years ago. This GSE family is different: they are slightly younger and, crucially, they have more "metals" (in astronomy, that means elements heavier than hydrogen and helium, like the stuff that makes up planets and people).

The Big Simulation

In this study, the authors, Alexis and Breivik, ran a massive computer simulation to see how this new "GSE family" changes the picture for LISA. They didn't just look at the stars; they simulated the entire life story of billions of stars, turning them into white dwarfs, and then calculated how LISA would hear them.

Here is what their simulation suggests:

1. The "Background Noise" Stays the Same
Imagine LISA is trying to hear a single violin in a stadium full of people humming. The "humming" is the combined sound of all the double white dwarfs, called the "foreground."
The authors found that even with the new GSE family added to the mix, the overall volume and shape of this humming noise remains unchanged. The "strength and height" of the gravitational wave foreground is the same whether you use the old map or the new map. The crash of the GSE galaxy didn't turn up the volume on the background noise.

2. The "Soloists" Look Different
While the background hum is the same, the specific couples LISA can pick out as individual "soloists" (about 40,000 of them) do change.

The Heavyweights Vanish: In the new model, the GSE family is younger. Because they are younger, the heavy double white dwarfs (specifically the Carbon-Oxygen ones) haven't had enough time to spiral close enough to be heard clearly. The simulation suggests LISA will see fewer of these heavy couples in the halo, especially those with a "chirp mass" above 0.7 solar masses.
The Lightweights Shine: On the flip side, the lighter couples (made of Helium) are doing just fine. In fact, because the GSE family is a bit richer in metals, the simulation suggests there might be more of these light Helium couples in the thick disk of the galaxy.
The "Neon" Ghosts Disappear: The simulation predicts that in the new model, LISA won't detect any of the Oxygen-Neon white dwarfs in the halo at all. They are too far away, and the higher metallicity of the GSE family seems to stop them from forming the tight orbits needed to be heard.

3. A New Shape to the Halo
The most exciting visual change is the shape of the crowd. The old map saw the halo as a perfect sphere. The new map, with the GSE family, sees it as a triaxial ellipsoid—a bit like a squashed football or a rugby ball.
The simulation shows that the couples LISA can resolve will stretch out further, reaching distances of up to 40 kpc (kiloparsecs). If LISA spots a double white dwarf that far away, the authors suggest it's almost certainly part of this GSE "crash" family. The distribution of these resolved stars will look more stretched out (prolate) than the old, perfectly round model predicted.

What This Means
The authors are careful to note that these are results from a simulation, not a direct measurement. They haven't heard the waves yet; they've just calculated what the waves should look like based on our new understanding of the galaxy's history.

They explicitly rule out the idea that the GSE crash would drastically change the total background noise LISA hears. The "hum" is steady. However, they suggest that the "soloists" LISA picks out will tell a different story. If LISA sees a halo full of heavy white dwarfs, our new model might need a tweak. If it sees a halo dominated by lighter couples stretching out to 40 kpc in a rugby-ball shape, it will be a strong hint that the GSE crash really did happen just as the Gaia satellite suggested.

In short: The background noise is the same, but the specific stars LISA will be able to point at and say, "There you are!" will look younger, lighter, and shaped more like a squashed ball than a perfect sphere.

Technical Summary: LISA's view of the Galactic Halo: forecasts for the Galactic double white dwarf population using Gaia data

Problem Statement
The Laser Interferometer Space Antenna (LISA), scheduled for launch in the 2030s, will detect gravitational waves (GWs) in the frequency range of $10^{-4}$ to $10^{-1}$ Hz. The dominant source population in this band consists of close double white dwarfs (DWDs) within the Milky Way. While a subset of these ( $\mathcal{O}(10^4)$ ) will be resolved as individual binaries, the majority will form an unresolved foreground. Previous population synthesis studies forecasting the LISA DWD population have largely relied on classical Galactic models that treat the stellar halo as a single, extremely metal-poor component. However, recent astrometric data from Gaia has revealed a triaxial, accreted component of the Galactic halo known as the Gaia-Sausage-Enceladus (GSE). This component contains stars that are more metal-rich and younger than the traditional in-situ halo population. The problem addressed in this work is how incorporating these empirically motivated, metal-rich, and younger stellar populations (specifically the GSE and an updated thick disk) alters the predicted properties, spatial distribution, and detectability of the DWD population for LISA.

Methodology
The authors simulate a realistic present-day Galactic DWD population using two distinct Star Formation History (SFH) models:

Fiducial Model: A classical model based on previous population synthesis forecasts (e.g., McMillan 2011; Robin et al. 2003; Ruiter et al. 2009). It assumes a single, extremely metal-poor halo ($[Fe/H] = -2.3$) formed 14 Gyr ago and a thick disk formed 11 Gyr ago with low metallicity.
Empirical Model: A model incorporating recent Gaia findings. It splits the halo into two components: the accreted GSE (42% of halo mass, $[Fe/H] = -1.2$, age $\sim$ 10 Gyr) and the in-situ halo (58% of halo mass, $[Fe/H] = -0.54$, age 14 Gyr). The thick disk is also updated to a "high- $\alpha$ " population with $[Fe/H] = -0.54$.

The simulation pipeline proceeds as follows:

Binary Evolution: The authors utilize the COSMIC binary population synthesis suite to generate DWD populations across a grid of metallicities. They consider four DWD types: HeHe, HeCO, COCO, and ONeX.
Formation Efficiency: They calculate the formation efficiency ( $\eta_{form}$ ) of DWDs per unit stellar mass for each component, scaling the simulated populations to the total mass of each Galactic component (thin disk, bulge, thick disk, halo).
Population Construction: DWDs are sampled according to the SFH of their respective components. Systems are filtered to ensure they have evolved into DWDs by the present day and have not merged or filled their Roche lobes.
GW Signal Calculation: The LEGWORK code is used to evolve the orbits via GW emission (following Peters 1964) and calculate the Signal-to-Noise Ratio (SNR) for a 4-year observation. Sources with SNR $> 7$ are considered resolved.

Key Results

Formation Efficiency and Population Counts: The empirical model, characterized by higher metallicities in the halo and thick disk, results in a lower overall DWD formation efficiency, particularly for He-hosting binaries. Consequently, the total number of detached DWDs in the empirical model decreases by approximately $4.7 \times 10^7$ compared to the fiducial model. The reduction is most pronounced in the thick disk and the GSE component.
LISA Foreground: Despite the changes in the underlying stellar population, the authors find that the strength and shape of the unresolved GW foreground remain qualitatively unchanged. The irreducible foreground is nearly indistinguishable between the two models.
Resolved Population Statistics: The total number of resolved sources decreases by less than 5% in the empirical model. However, the composition and distribution of these sources change significantly:
- Thick Disk: The empirical thick disk shows an increase in resolved COCO DWDs with chirp masses between 0.6 and 0.8 $M_\odot$ . These systems form at higher frequencies due to the higher metallicity progenitors undergoing common envelope (CE) evolution differently than low-metallicity counterparts.
- Halo: The inclusion of the younger GSE component reduces the population of COCO and ONeX DWDs in the halo. Specifically, the empirical model predicts a near-total absence of resolved ONeX DWDs in the halo, as higher metallicity inhibits their formation via enhanced line-driven winds and wider initial separations.
- Spatial Distribution: The resolved DWDs in the empirical model extend to distances of $\sim$ 40 kpc, driven by the triaxial GSE component. DWDs beyond $\sim$ 30 kpc are predicted to be predominantly He-hosting binaries with low chirp masses.
- Morphology: The 3D spatial distribution of resolved sources in the empirical model appears more prolate and less concentrated toward the Galactic center compared to the spherical fiducial halo, reflecting the triaxial nature of the GSE.

Significance and Claims
The paper claims that while the inclusion of the GSE and updated thick disk parameters does not significantly alter the total GW foreground or the total count of resolved sources, it fundamentally modifies the observable properties of the resolved DWD population. Specifically, the authors predict that:

Metallicity Dependence: Higher metallicity in the thick disk and halo leads to a shift in the chirp mass distribution of resolved COCO binaries and a suppression of ONeX binaries in the halo.
Halo Structure: LISA-resolved sources can serve as tracers of the Galactic halo's structure. The triaxiality of the GSE is imprinted on the 3D positions of resolved DWDs, particularly those at distances $>30$ kpc, which are predicted to be dominated by the GSE component.
Future Forecasts: The study suggests that refining Galactic formation history models using Gaia data is essential for accurate forecasts of the specific sub-populations LISA will resolve, even if the global foreground statistics remain robust.

The authors conclude that future observations will benefit from these updated models to better interpret the stellar-origin binaries LISA detects, particularly regarding the spatial and chemical history of the Milky Way's halo.

More like this