The Milky Way has a magnetic field, but no camera can step outside the Galaxy and photograph its shape. Astronomers have to infer that field from inside it, using radio waves that have crossed magnetised gas on their way to Earth.

Those measurements had already established an unusual feature between the Sun and the Sagittarius Arm. The large-scale field near us points mainly one way around the Galaxy, while the field farther inward points the other. Somewhere between the two is a reversal.

The usual diagrams made that reversal look almost architectural: two magnetic regions separated by a boundary standing through the Galactic disc. A model published in 2026 replaces that wall with something much less intuitive. The boundary behaves like a tilted plane slicing diagonally through the disc and passing above the Sun in Galactic coordinates.

This is not a direct image of a physical sheet, and “above” does not mean just beyond the planets. The model’s central-plane intercept lies an estimated 0.25 to 0.55 kiloparsecs toward the Galactic centre, roughly 800 to 1,800 light-years away. What has changed is the best simple geometry for explaining the radio pattern, not the discovery of a solid surface near the Solar System.

How to map a field that cannot be seen

The raw information comes from Faraday rotation. A linearly polarised radio wave has a measurable orientation. As it travels through free electrons threaded by a magnetic field, that orientation rotates. The amount of rotation changes with the square of the wavelength and with the electron density, field strength and distance along the line of sight.

The sign is especially useful. A positive or negative Faraday depth indicates whether the average line-of-sight component of the magnetic field points toward or away from the observer. Gather enough measurements in enough directions and broad magnetic patterns emerge.

There is a complication. One sightline may pass through several emitting and rotating regions at different distances. Diffuse synchrotron radiation can be generated throughout the same plasma that rotates it. The signal then becomes “Faraday complex,” with multiple contributions mixed together rather than one source shining through a simple foreground screen.

This is one reason radio astronomy can turn an apparently empty sky into a map of invisible structure. SpaceDaily recently revisited Karl Jansky’s accidental discovery of Galactic radio emission. Less than a century later, frequency-resolved polarisation lets astronomers use that emission to probe the direction of magnetic fields within the interstellar medium.

The reversal itself was not new

Near the Sun, within roughly one kiloparsec, the coherent large-scale field is usually described as clockwise when the Milky Way is viewed from its north Galactic pole. Toward and beyond the Sagittarius Arm, the field becomes counterclockwise. Pulsars and polarised sources outside the Galaxy had supplied evidence for that switch over decades.

What remained poorly known was the boundary’s physical location and three-dimensional form. A flat diagram of the Milky Way encourages a radial interpretation: the field changes direction at a certain distance from the Galactic centre, producing a current-carrying boundary roughly perpendicular to the disc.

Yet the observations had begun refusing that tidy picture. In a 2017 study of the Sagittarius-Carina reversal, Anna Ordog and colleagues found a diagonal gradient across a radio rotation-measure map. The boundary ran from about Galactic longitude 67 degrees and latitude 4 degrees to longitude 56 degrees and latitude minus 2 degrees.

A later analysis of pulsar rotation measures added distance information. It suggested that the reversal lies nearer at positive Galactic latitude, around the boundary between the Local and Sagittarius arms, while at negative latitude it extends farther into the Sagittarius Arm. A plane tilted toward us in the northern Galactic hemisphere would naturally produce that arrangement.

DRAGONS turned the hint into a model

The 2026 work was led by Rebecca Booth of the University of Calgary and published in The Astrophysical Journal. Its foundation was the Dominion Radio Astrophysical Observatory Global Magneto-Ionic Medium Survey of the Northern Sky, shortened to DRAGONS.

Using the 15-metre DRAO telescope in British Columbia, the survey measured full radio polarisation across the northern sky from 350 to 1,030 megahertz. Booth’s analysis used the upper part of that band, from 500 to 1,030 megahertz, after contaminated frequencies were removed. The broad wavelength coverage enabled Faraday synthesis, which estimates how much polarised emission appears at different Faraday depths.

The published DRAGONS survey paper describes a dataset designed to retain both useful Faraday-depth resolution and sensitivity to broad structures. That combination matters because a survey can otherwise resolve fine features while filtering out the large-scale pattern, or recover the broad pattern without separating its depth structure.

Booth and colleagues reduced each Faraday-depth spectrum to a polarised-intensity-weighted mean, then examined how that value changed across Galactic longitude and latitude. South of the Galactic plane, the sign broadly alternated once around the sky. In northern mid-latitudes it alternated twice. The team asked whether both patterns could be produced by one nearby reversal viewed from different sides.

A tilted plane above the Sun

In the three-dimensional reversal paper, the boundary is represented as a plane. Below it, where the Sun is placed, the coherent field runs clockwise when seen from the north Galactic pole. Above it, the field runs counterclockwise.

The plane’s normal points along Galactic longitude 168.5 degrees and latitude minus 60 degrees. Its vertical orientation was chosen to reproduce the slope of the diagonal identified in the earlier observations, while its orientation within the disc was constrained so that the boundary lay between the Local and Sagittarius arms without cutting through either one.

Comparisons with three-dimensional dust maps helped set the distance. Two conspicuous Faraday structures aligned morphologically with dust at about 395 and 497 parsecs. From those associations, the researchers placed the reversal’s intercept toward the Galactic centre between 0.25 and 0.55 kiloparsecs.

That is the sense in which the plane passes just above our region of space. The Sun is on the clockwise side, below the model boundary, while northern lines of sight cross into the counterclockwise region sooner than southern ones. “Above” refers to height relative to the Galactic mid-plane, not north in Earth’s sky and not a direction fixed by our planet’s axis.

One simple plane reproduced much of the sky

The researchers tested two limiting ways that radio emission and Faraday rotation might be arranged. In a screen, most of the polarised emission sits behind the rotating material. In a slab, emission and rotation are mixed along the path. Despite that difference, the best-fit all-sky maps were nearly indistinguishable by eye.

Across the DRAGONS map, the model reached a pixel-by-pixel correlation of 0.6 and matched the observed sign over 74 percent of the fitted sky area. Within most of the selected three-degree-wide latitude bands, correlations exceeded 0.7. The plane reproduced the number and spacing of the large positive and negative peaks without requiring extra large-scale magnetic components.

Those figures are evidence that the geometry is useful, not proof that the Milky Way contains a limitless flat sheet. The authors explicitly say the true reversal cannot be an infinite plane. It probably curves roughly with the spiral arms, and the planar description is an approximation over the local region, defined here as within one kiloparsec of the Sun.

The model also simplifies the interstellar medium by treating quantities such as electron density and field strength more uniformly than they are in reality. Some fitted path lengths came out longer than the expected distance over which the DRAGONS band remains sensitive to diffuse polarised emission. The paper argues that a non-uniform medium could produce the same Faraday pattern over shorter paths.

Why local geometry can masquerade as a Galactic pattern

One of the study’s more consequential suggestions is that much of the large-scale Faraday sky may be controlled by the magnetic configuration relatively close to us. A pattern spanning a large fraction of the sky does not necessarily arise from a structure spanning the whole Milky Way.

This line-of-sight problem appears repeatedly in Galactic astronomy. SpaceDaily reported how a radio feature once interpreted as a tower rising from the Galactic centre was reclassified as a much nearer stellar bubble in the foreground. The objects are different, but the interpretive risk is similar: a two-dimensional sky position does not supply a distance.

A better local magnetic model matters for more than drawing the spiral arms correctly. Galactic fields guide charged cosmic rays, interact with turbulent gas and influence how matter is organised in the interstellar medium. Their radio signature is also a foreground that must be separated from attempts to measure fainter signals behind the Milky Way.

If the reversal is tilted, calculations that assume a vertical arm-to-arm wall can assign the wrong field direction or path length to a given sightline. Pulsar distances, cosmic-ray transport models and studies of the Galactic halo can all inherit that geometric error.

The shape is clearer than the cause

The model does not determine why the field reverses. Galactic dynamo theories can generate reversals as turbulence, differential rotation and other motions amplify and organise magnetic fields. Magnetohydrodynamic simulations also show that spiral-arm density shocks and changing gas flows can flip a field’s direction.

Whether the Milky Way contains only this one prominent local reversal or several remains disputed. Magnetic reversals this obvious are also difficult to identify in external spiral galaxies, which raises the possibility that our inside view and local surroundings make the Milky Way appear unusual.

Future progress will come from more sightlines with reliable distances. Pulsars are especially valuable because their rotation measures can be placed at finite points within the Galaxy, rather than integrated all the way through it. Denser measurements above and below the disc can test whether the inferred plane bends, thickens, fragments or follows a spiral structure beyond the local volume.

For now, the advance is careful but real. Astronomers did not discover the reversal in 2026. They found that a familiar flat map had encouraged the wrong three-dimensional intuition. The boundary near Sagittarius looks less like a wall between spiral arms and more like a tilted magnetic seam passing across our portion of the Galactic disc.