Reading the Milky Way’s gravity in a spiral of stellar motions

A spiral in motion, not a spiral arm

The pattern astronomers call a “phase snail” is not a structure you could photograph across the sky. It appears in a diagram pairing each star’s height above or below the Galactic plane with its velocity perpendicular to that plane.

After a disturbance, stars with different vertical oscillation periods gradually move out of step. Their distribution winds into a spiral in this position–velocity diagram. The process is called phase mixing: initially coordinated motions spread out because different orbits run at different rates.

Those rates depend on the Galaxy’s gravitational potential—the landscape of gravitational energy that determines how stars move. In their preprint, Rui Guo and colleagues investigate how much of that landscape can be recovered from the spiral’s shape alone. They do not establish what caused the disturbance.

Turning a curved pattern into a straight-line test

The researchers used Gaia’s third data release, selecting stars with the measurements needed to reconstruct their motions. After quality cuts, their sample contained about 27.2 million stars.

They grouped stars by guiding-center radius: the radius of a circular orbit with the same angular momentum as the star. This made the spirals clearer than grouping stars simply by their current distance from the Galactic center.

The analysis covered eight groups with guiding-center radii between 8 and 11 kiloparsecs. A kiloparsec is about 3,260 light-years. The researchers traced the density ridge running along each spiral, producing 321 points in total, rather than fitting the entire distribution of stars.

For each candidate gravitational model, they calculated two quantities at those points: the frequency of vertical oscillation and the phase angle, which describes progress through an oscillation. Under the method’s assumptions, the spiral should become a straight line in this new diagram. Its slope indicates the elapsed time since the disturbance.

They then explored different combinations of disk and dark halo parameters to find models that made these transformed curves approximately straight. An empirical correction accounted for the difference between guiding-center radius and the radius whose gravity a grouped spiral represents.

What the fits found

First, the team tested the procedure using simulated stars moving in a prescribed gravitational field. The stars received a downward velocity kick and then evolved for 500 million years. The method recovered the input vertical potential to within about 5% over the region studied and estimated an elapsed time of about 491 million years.

Applied to Gaia, the spiral-only fit produced vertical potentials broadly consistent with an established Milky Way model. But it predicted a Galactic rotation curve that was flatter than the observed one. A rotation curve describes how circular orbital speed changes with distance from the Galactic center.

The team therefore fitted the spirals together with published rotation measurements. These supply complementary information: the spirals probe gravity perpendicular to the disk, while rotation constrains its radial pull.

The combined analysis gave a median dark halo mass of about 870 billion solar masses, with reported uncertainties extending roughly 330 billion below and 310 billion above that value. This is the mass within a conventionally defined halo boundary, not a direct census of all Galactic matter.

Near the Sun, the inferred dark matter density was 0.0097 solar masses per cubic parsec, with reported uncertainties of +0.0002 and −0.0009. That estimate was substantially tighter than the spiral-only result. Across the radial groups, inferred disturbance times ranged from about 400 million to 750 million years, with a median of 564 million years.

Where the interpretation remains uncertain

The simulation test is encouraging, but it uses a fixed gravitational field. Neither that test nor the fitting method includes the changing gravity generated by the disturbed disk itself. This self-gravity can alter how a spiral winds and bias estimates of its age.

The method also approximates vertical motion as separable from motion within the disk. The authors restrict their analysis to heights below about 1.5 kiloparsecs, where their tests support that approximation. The outer radial groups have noisier spirals and fewer visible turns, making their interpretation less secure.

A straight transformed curve is not a unique guarantee of the correct potential. The paper’s mathematical tests show that some altered potential shapes can preserve approximate straightness while substantially changing the inferred disturbance time.

Finally, the results depend on the adopted mass models and radial correction. Adding rotation data tightened the local density estimate but left two favored regions of halo parameter space, increasing the uncertainty in halo mass. The study therefore demonstrates a useful gravitational probe—not a uniquely settled map of the Milky Way’s dark matter.