GPM J1839-10: Ultra-Long Radio Transient

• GPM J1839-10 is a Galactic long-period radio transient characterized by a 22-minute periodicity, strong polarization, and persistent activity spanning over three decades.
• Radio observations reveal complex sub-pulse structures, frequency drifting, and orthogonal polarization mode switches that shed light on its emission processes.
• Current models favor either a magnetar-like neutron star or a magnetic white-dwarf binary, with constraints provided by precise timing, DM, RM, and spectral analysis.

GPM J1839−10 is a Galactic long-period radio transient, also described as an ultra-long period object, distinguished by periodic coherent radio emission on a timescale of about 22 minutes, sustained activity over at least three decades, and unusually strong polarization and sub-pulse structure. Later observational and theoretical work has fixed the commonly used radio modulation at $P = 1318.1957(2)\,\mathrm{s}$, measured $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$, found $\mathrm{RM} \approx 531\,\mathrm{rad\,m^{-2}}$, and placed an upper limit of $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$. These properties place the source at the center of an active debate over whether it is best understood as a magnetar-like neutron star, a neutron star shaped by fallback-disk evolution, or a magnetic white-dwarf binary (Men et al., 17 Jan 2025, Horváth et al., 21 Jul 2025).

1. Observational identity

GPM J1839−10 is one of the prototypical members of the emerging class of long-period radio transients. In the current literature it is treated interchangeably as an LPT, an LPRT, or a ULPO, reflecting the fact that its periodicity is orders of magnitude longer than that of ordinary radio pulsars. Timing-based summaries give the source a radio period of $1318.1957(2)\,\mathrm{s}$, or about 22 minutes, and only an upper limit on the period derivative, $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$. From that limit, one evolutionary study adopts a characteristic-age bound of $\tau_{\rm c} > 1.2\times10^{8}\,\mathrm{yr}$ as a rough maximum timescale for viable tracks (Fan et al., 2024).

Propagation measurements are comparatively stable. The source has $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$, while radio polarimetry has consistently found $\mathrm{RM}$ near $531\,\mathrm{rad\,m^{-2}}$; one white-dwarf-binary analysis adopts $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$0, and later FAST observations reported RMs consistent with earlier values rather than strong epoch-to-epoch changes (Horváth et al., 21 Jul 2025, Men et al., 14 Feb 2026). A distance estimate of $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$1, derived from the electron-density model used in later summaries, remains uncertain and propagates into essentially all luminosity and counterpart arguments (Men et al., 17 Jan 2025).

The source is also unusual in persistence. Several later studies emphasize that its radio activity extends back to archival detections from 1988, making it active for at least three decades. That longevity sharply distinguishes it from one-off transients and constrains models that would imply rapid secular evolution on decade timescales (Horváth et al., 21 Jul 2025, Pelisoli et al., 24 Sep 2025).

2. Radio phenomenology

Follow-up radio observations have turned GPM J1839−10 from a timing anomaly into a detailed polarimetric and time–frequency laboratory. A MeerKAT campaign in the UHF band detected pulses in 3 of 15 observed rotation periods, implying an “on” duty in periods of about 20%. Individual emission windows lasted $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$2–$\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$3, and the narrowest burst subcomponents reached $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$4. The same study reported peak fluxes of $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$5–$\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$6, a quasi-periodic modulation in one pulse at $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$7, frequent orthogonal polarization mode switches, large and rapidly varying linear and circular polarization fractions, and drifting substructures closely resembling those of repeating FRBs. In one fitted pulse complex, the lower-frequency component drifted at $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$8, while a multi-component comparison gave $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$9 (Men et al., 17 Jan 2025).

The same MeerKAT analysis identified phase-dependent linear-to-circular polarization conversion. Fits using generalized Faraday conversion yielded $\mathrm{RM} \approx 531\,\mathrm{rad\,m^{-2}}$0 indices of $\mathrm{RM} \approx 531\,\mathrm{rad\,m^{-2}}$1, $\mathrm{RM} \approx 531\,\mathrm{rad\,m^{-2}}$2, and $\mathrm{RM} \approx 531\,\mathrm{rad\,m^{-2}}$3 in three selected sub-pulses, with corresponding generalized rotation measures of $\mathrm{RM} \approx 531\,\mathrm{rad\,m^{-2}}$4, $\mathrm{RM} \approx 531\,\mathrm{rad\,m^{-2}}$5, and $\mathrm{RM} \approx 531\,\mathrm{rad\,m^{-2}}$6. The same paper also reported a “down-drifting polarization conversion” band, in which circular polarization increased, linear polarization decreased, total intensity dropped by about 50%, and the drift was fitted with $\mathrm{RM} \approx 531\,\mathrm{rad\,m^{-2}}$7 and $\mathrm{RM} \approx 531\,\mathrm{rad\,m^{-2}}$8, corresponding to an effective drift rate of $\mathrm{RM} \approx 531\,\mathrm{rad\,m^{-2}}$9 in the UHF band (Men et al., 17 Jan 2025).

FAST observations at $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$0–$\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$1 extended this picture. Across 5 sessions totaling 4 hours, FAST detected 7 distinct pulses with durations of tens of seconds. Those data showed that orthogonal polarization mode switches coincide with dips in polarized intensity, supporting an interpretation in terms of incoherent summation of orthogonal modes, and they yielded no statistically significant secular variation in DM or RM. Most notably, one pulse showed a narrow, drifting cyclotron absorption feature in which total and linear intensities decreased while circular polarization increased. The paper interpreted this as requiring a magnetic field at the absorption site with a lower limit of tens of Gauss, and in oblique geometries of order $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$2 (Men et al., 14 Feb 2026).

Instrumental studies have independently confirmed the source’s detectability via its sub-pulse structure. In a 3-hour ASKAP/CRACO observation, pulses from GPM J1839−10 were detected in an untargeted search at 110.592 ms resolution. The pipeline returned $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$3, offset from the known value because of the wide pulse profile, low time resolution, and 120 MHz bandwidth, but re-dedispersion at the known DM recovered a broadband pulse in the expected form (Wang et al., 2024).

3. Neutron-star and fallback-disk interpretations

A large fraction of the literature treats GPM J1839−10 as an extreme neutron-star system, but not as an ordinary rotation-powered pulsar. Several papers make the same starting point: standard magnetic-dipole braking does not reach such periods. In one fallback-disk study, full magnetic-field-decay simulations gave a maximum period from dipole braking alone of only $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$4, far below the $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$5–$\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$6 range occupied by GPM J1839−10 and related objects (Xu et al., 2024).

That mismatch motivates disk-assisted spin-down. In the “magnetar + fallback disk” framework, an early interacting disk can spin a strong-field neutron star down to ultralong periods, but a presently active disk is strongly constrained by the observed upper limit on $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$7. One evolutionary calculation concluded that J1839 is impossible to reconcile with the observed $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$8 limit during an active fallback-disk phase and therefore proposed that the source is now in a “second ejector phase” after the disk became inactive. In that model, possible progenitors require $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$9 and an initial fallback-disk accretion rate of $1318.1957(2)\,\mathrm{s}$0. An illustrative track with $1318.1957(2)\,\mathrm{s}$1, $1318.1957(2)\,\mathrm{s}$2, and disk activity until $1318.1957(2)\,\mathrm{s}$3 reaches $1318.1957(2)\,\mathrm{s}$4 at disk death, then evolves to $1318.1957(2)\,\mathrm{s}$5 and $1318.1957(2)\,\mathrm{s}$6 at $1318.1957(2)\,\mathrm{s}$7 (Fan et al., 2024).

A related population study of fallback disks emphasizes timescale separation. It distinguishes an interacting lifetime $1318.1957(2)\,\mathrm{s}$8 from a much longer existence lifetime $1318.1957(2)\,\mathrm{s}$9, finding that nearly all interacting lifetimes are shorter than $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$0, more than half are shorter than 1 year, and only a tiny fraction approach Myr scales, whereas many disks survive in noninteracting form for $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$1–$\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$2. Applied to GPM J1839−10, this supports the interpretation that an early disk could have supplied the required torque, while any surviving disk is now cold, faint, noninteracting, or disrupted, naturally explaining the lack of active X-ray or infrared signatures (Xu et al., 2024).

Other neutron-star analyses sharpen the objection to a standard ejector interpretation. One study of evolutionary stage argued that GPM J1839−10 can remain an ejector in a typical interstellar medium only for unrealistically large dipolar fields $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$3; otherwise it should be in a propeller-like state rather than an ordinary radio-pulsar regime (Afonina et al., 2023). Another death-line and pulse-width analysis treated the $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$4 pulse window as $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$5 and concluded that GPM J1839−10 lies far below the fiducial neutron-star death line and is unlikely to be a normal radio pulsar, leaving magnetar or white-dwarf possibilities open (Tong, 2023).

The magnetar reading is strengthened by radio phenomenology rather than only by spin evolution. The MeerKAT study argued that the complex PA structure, orthogonal modes, linear-to-circular conversion, and FRB-like drifting substructures are more naturally explained by a magnetospheric origin than by a simple shock model, and treated these properties as evidence in support of a long-period magnetar interpretation and of a possible magnetar–FRB connection (Men et al., 17 Jan 2025).

4. White-dwarf binary interpretation

A distinct and increasingly elaborate alternative treats GPM J1839−10 as a magnetic white dwarf in a binary with a low-mass companion. In the most developed version, a 36-year timing baseline combined with a 36-hour multi-telescope radio campaign yields an orbital period inferred solely from the radio data, $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$6. Within that framework, the previously reported $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$7 is reinterpreted as a spin–orbit beat period, while the true white-dwarf spin is $\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$8, related through

$\dot P < 3.6\times10^{-13}\,\mathrm{s\,s^{-1}}$9

The model attributes radio emission to episodes when the magnetic pole of the rotating white dwarf intersects the companion’s wind in the orbital plane (Horváth et al., 21 Jul 2025).

The same work argues that this geometry reproduces several otherwise disconnected observations: intermittent activity, double pulse groups, stable $\tau_{\rm c} > 1.2\times10^{8}\,\mathrm{yr}$0 substructure, and the separation of orthogonal polarization modes between the two pulse groups. Its fitted parameters are $\tau_{\rm c} > 1.2\times10^{8}\,\mathrm{yr}$1, $\tau_{\rm c} > 1.2\times10^{8}\,\mathrm{yr}$2, $\tau_{\rm c} > 1.2\times10^{8}\,\mathrm{yr}$3, and $\tau_{\rm c} > 1.2\times10^{8}\,\mathrm{yr}$4, with characteristic angular widths $\tau_{\rm c} > 1.2\times10^{8}\,\mathrm{yr}$5 and $\tau_{\rm c} > 1.2\times10^{8}\,\mathrm{yr}$6. The assumed companion is an M dwarf with $\tau_{\rm c} > 1.2\times10^{8}\,\mathrm{yr}$7; for the fitted orbital period, the derived orbital separation is $\tau_{\rm c} > 1.2\times10^{8}\,\mathrm{yr}$8, while the companion does not fill its Roche lobe, so the system is not in Roche-lobe-overflow accretion (Horváth et al., 21 Jul 2025).

Optical follow-up does not exclude this picture. A 2-hour HiPERCAM observation detected a source at the radio position in $\tau_{\rm c} > 1.2\times10^{8}\,\mathrm{yr}$9 with $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$0 AB mag and obtained $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$1 limits of $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$2, $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$3, $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$4, and $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$5. Because the field is crowded, the probability of a chance alignment within the $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$6 radio/astrometric error circle is about 7%. Even so, white-dwarf-plus-companion models are not ruled out for plausible combinations of distance and reddening: distances $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$7 are excluded for a WD+M binary, but at the likely $\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$8–$\mathrm{DM} = 273.5 \pm 2.5\,\mathrm{pc\,cm^{-3}}$9 range and extinctions up to $\mathrm{RM}$0, even hot white dwarfs remain allowed. The same dataset also showed suggestive, but not definitive, periodic behavior at harmonics of the radio period, including a Fourier peak above $\mathrm{RM}$1 at the third harmonic and folded-light-curve sinusoid preferences at the 95% level, although the BIC remained slightly in favor of a constant model (Pelisoli et al., 24 Sep 2025).

From a broader population perspective, rotating white-dwarf scenarios are easier to reconcile with the existence of ultra-long periods than rotating neutron-star dipoles. A population-synthesis study found that white-dwarf spin-down can readily accommodate a large population of long-period radio emitters, whereas neutron-star dipole evolution does not. That study nevertheless stressed that no mechanism easily explains the production of such bright coherent radio emission in either isolated dipole picture, so the white-dwarf interpretation alleviates the period problem more effectively than the emission-physics problem (Rea et al., 2023).

5. Alternative proposals and exclusions

Not all proposed models survive current constraints. A self-lensed pulsar–black-hole scenario was put forward in which the observed 22-minute period is orbital, and bursts are magnified conjunction events from an otherwise faint pulsar. Applied to GPM J1839−10, the observed period and $\mathrm{RM}$2–$\mathrm{RM}$3 burst durations imply a lensing black-hole mass of $\mathrm{RM}$4–$\mathrm{RM}$5, a coalescence time of $\mathrm{RM}$6 to $\mathrm{RM}$7, and an orbital-period derivative of $\mathrm{RM}$8 to $\mathrm{RM}$9. The same model then implies a merger rate of $531\,\mathrm{rad\,m^{-2}}$0. Because those $531\,\mathrm{rad\,m^{-2}}$1 values are far larger than the observational upper limit and the implied rates are correspondingly extreme, the self-lensing interpretation is regarded as disfavored (Xiao et al., 2024).

Other neutron-star alternatives remain more weakly constrained. Optical follow-up summarized the continuing viability of double-neutron-star binary models, but also noted that a simple precessing-pulsar interpretation is disfavored because microsecond-resolution radio data have not revealed a shorter underlying spin period. The same study argued that standard magnetar interpretations are challenged by the absence of detected X-rays at levels typical of classical magnetars, while emphasizing that the optical data alone do not formally exclude neutron-star scenarios (Pelisoli et al., 24 Sep 2025).

A broader conceptual point, emphasized in population work, is that identifying the compact-object class does not by itself solve the radio-emission problem. In neutron-star dipole models, GPM J1839−10 occupies a region where radio activity is not expected in standard pair-creation language; in isolated white-dwarf dipole models, the period distribution is more natural but the origin of bright coherent radio emission remains similarly unsettled (Rea et al., 2023).

6. Present status and discriminating tests

The present literature converges on a negative statement more strongly than on a unique positive identification: GPM J1839−10 is not comfortably explained as an ordinary rotation-powered radio pulsar. The main live interpretations are instead an old or disk-processed magnetar-like neutron star, and a magnetic white-dwarf binary in which the 22-minute modulation is a beat period rather than the true spin (Fan et al., 2024, Horváth et al., 21 Jul 2025).

Several observational tests have been identified as decisive. In fallback-disk neutron-star models, the expectation is that the source is now in a dipole-dominated late stage with a true $531\,\mathrm{rad\,m^{-2}}$2 well below the current upper limit, rather than in an actively torqued disk phase; more precise long-term timing is therefore central (Fan et al., 2024). In the white-dwarf-binary picture, spectroscopy or radial-velocity work at the proposed $531\,\mathrm{rad\,m^{-2}}$3 orbital period would directly test the binary interpretation, while deeper and longer-baseline optical photometry could confirm or refute the harmonic signals seen tentatively in $531\,\mathrm{rad\,m^{-2}}$4 (Pelisoli et al., 24 Sep 2025).

Multiwavelength searches remain equally important. The short interacting lifetimes predicted for fallback disks imply that, if a disk once existed, it is now most likely cold, faint, and infrared-dominated rather than X-ray luminous, making deep infrared searches a direct test of the inactive-disk hypothesis (Xu et al., 2024). Conversely, the white-dwarf scenario expects an optical or infrared stellar counterpart to be more likely than in isolated neutron-star models, though distance and extinction currently prevent a clean exclusion (Pelisoli et al., 24 Sep 2025). On the radio side, higher-time-resolution polarimetry can refine the interpretation of orthogonal modes, drifting substructure, and cyclotron absorption, while new survey backends such as CRACO are explicitly being developed to discover similar systems through their sub-pulses rather than their minute-scale periodicity (Men et al., 14 Feb 2026, Wang et al., 2024).

GPM J1839−10 therefore occupies a dual role. Empirically, it is among the best-studied long-period radio transients, with a radio phenomenology now extending from minute-scale periodicity to millisecond-scale substructure and polarization transport effects. Interpretively, it remains a contested object whose most important contribution may be methodological: it forces any successful model to account simultaneously for an ultralong recurrence timescale, persistent activity over decades, a very small $531\,\mathrm{rad\,m^{-2}}$5, strong polarization structure, FRB-like drifting behavior, and the absence of a straightforward high-energy or optical identification.