G118.4+37.0 (Calvera’s SNR): Multiwavelength Study

• G118.4+37.0 (Calvera’s SNR) is a candidate supernova remnant at high Galactic latitude characterized by a nearly circular radio shell and multiwavelength emissions from radio to X-rays.
• The remnant displays a non-thermal synchrotron radio spectrum with a spectral index of about 0.71 and ambiguous gamma-ray signatures that suggest both leptonic and hadronic emission mechanisms.
• Recent multiwavelength analyses using LOFAR, Fermi-LAT, and XMM-Newton indicate a middle-aged SNR interacting with a structured ambient medium, offering a unique laboratory for studying particle acceleration.

Searching arXiv for papers on G118.4+37.0 / Calvera’s SNR to ground the article in the current literature. G118.4+37.0, commonly referred to as Calvera’s SNR, is a candidate supernova remnant at high Galactic latitude, spatially associated with the X-ray pulsar Calvera, 1RXS J141256.0+792204. It was first reported as a faint, nearly circular radio ring in LOFAR Two-metre Sky Survey data, then identified as an extended GeV source in Fermi-LAT observations, and subsequently examined through deeper XMM-Newton and optical H$\alpha$ follow-up. Across these studies, the source is characterized by a $\sim 1^\circ$-scale shell-like morphology, non-thermal radio emission, extended $\gamma$-ray emission, and diffuse thermal X-rays interior to the ring, although its detailed emission mechanism and evolutionary state remain under active discussion (Arias et al., 2022, Araya, 2022, Xin et al., 2022, Greco et al., 17 Jul 2025).

1. Discovery and basic identification

G118.4+37.0 was reported as a ring of low surface brightness radio emission around the Calvera pulsar in the LOFAR Two-metre Sky Survey (LoTSS) (Arias et al., 2022). The ring is centered at $\alpha=14\mathrm{h}11\mathrm{m}12.6\mathrm{s}$, $\delta=+79^\circ23'15''$, corresponding to Galactic coordinates $l = 118.41^\circ$, $b = +37.03^\circ$, with inner and outer radii of $14.2'$ and $28.4'$ (Arias et al., 2022). Its high latitude is central to its interpretation, since a confirmed association with Calvera would place the remnant far above the Galactic plane and make it one of the few known Galactic halo SNRs (Arias et al., 2022).

The designation G118.4+37.0 encodes the approximate Galactic coordinates and is used in the literature interchangeably with “Calvera’s SNR” (Araya, 2022, Xin et al., 2022). The radio ring is very round and of relatively uniform brightness, apart from a brighter, straight filament in the north-west (Arias et al., 2022). Its angular diameter is approximately $0.95^\circ$, consistent with the later GeV extension measured with Fermi-LAT (Xin et al., 2022).

The discovery paper considered three interpretations for the ring: an H II region, a supernova remnant, or an Odd Radio Circle (Arias et al., 2022). The authors preferred the SNR interpretation because of the positional coincidence of the ring, the pulsar, and an X-ray-emitting non-equilibrium ionisation plasma previously detected in the field (Arias et al., 2022). Later GeV detection of extended emission consistent with the radio morphology strengthened that interpretation by confirming the presence of relativistic particles (Araya, 2022).

2. Radio properties and early multi-wavelength diagnostics

The radio ring was detected at 144 MHz in LoTSS with an integrated flux density of $\sim 1^\circ$0 Jy, and also measured in WENSS at 325 MHz with $\sim 1^\circ$1 Jy and in NVSS at 1.4 GHz with $\sim 1^\circ$2 Jy (Arias et al., 2022). A power-law fit to these three flux densities gave a radio spectral index $\sim 1^\circ$3 for $\sim 1^\circ$4 (Arias et al., 2022). In the later $\sim 1^\circ$5-ray modeling paper, the same radio behavior is described as a “flat” integrated radio spectrum with $\sim 1^\circ$6 under the convention $\sim 1^\circ$7 (Xin et al., 2022). Both notations encode the same non-thermal synchrotron slope.

The mean surface brightness at 144 MHz was reported as $\sim 1^\circ$8, and the corresponding 1 GHz surface brightness as $\sim 1^\circ$9 (Arias et al., 2022). These values place the object among low-surface-brightness remnants and were used for comparison with empirical $\gamma$0–$\gamma$1 relations (Arias et al., 2022). The mean brightness temperature at 144 MHz was estimated as $\gamma$2 K (Arias et al., 2022).

No polarized emission associated with the ring was detected in LoTSS RM-synthesis analysis over Faraday depths of $\gamma$3, with a polarized-intensity limit of $\gamma$4 at $\gamma$5 resolution (Arias et al., 2022). The resulting fractional polarization constraint is $\gamma$6 on those scales, although strong low-frequency depolarization is expected (Arias et al., 2022). This non-detection did not rule out an SNR interpretation.

The early optical and X-ray context was mixed but suggestive. No shell-like H$\gamma$7 emission was detected around the radio ring in INT/WFC imaging, with a $\gamma$8 surface-brightness limit of $\gamma$9 (Arias et al., 2022). A small V-shaped H$\alpha=14\mathrm{h}11\mathrm{m}12.6\mathrm{s}$0 “smudge,” about $\alpha=14\mathrm{h}11\mathrm{m}12.6\mathrm{s}$1 long, was detected inside the ring at $\alpha=14\mathrm{h}11\mathrm{m}12.6\mathrm{s}$2, $\alpha=14\mathrm{h}11\mathrm{m}12.6\mathrm{s}$3, but its relation to the radio shell was uncertain (Arias et al., 2022). In X-rays, no diffuse shell emission was seen in the ROSAT All-Sky Survey, yet an interior extended source, 1RXS J140818.1+792113, had previously been modeled as a non-equilibrium ionization plasma with oxygen overabundance, a property described as a strong SNR diagnostic (Arias et al., 2022).

These combined diagnostics disfavored the H II region and Odd Radio Circle interpretations. The H II hypothesis was found inconsistent with the steep radio spectrum, the stringent H$\alpha=14\mathrm{h}11\mathrm{m}12.6\mathrm{s}$4 non-detection, the lack of WISE emission, and the absence of a plausible ionizing star (Arias et al., 2022). The ORC interpretation was disfavored by the $\alpha=14\mathrm{h}11\mathrm{m}12.6\mathrm{s}$5 angular size, which is much larger than the arcminute scales typical of known ORCs, and by the lack of an obvious host galaxy (Arias et al., 2022).

3. GeV detection and spatial morphology

Extended GeV emission from the region was reported independently in two 2022 Fermi-LAT studies, both based on long integrations of Pass 8 data (Araya, 2022, Xin et al., 2022). One study analyzed almost 14 years of observations and detected extended GeV emission consistent with the size and location of the radio source, arguing that these features and its similarities to other isolated SNRs establish the source as the remnant of a supernova (Araya, 2022). The other used 13.6 years of Pass 8 data and provided detailed localization and extension tests (Xin et al., 2022).

In the 13.6-year analysis, the preferred morphology was a uniform disk (Xin et al., 2022). The point-source localization in the 1–1000 GeV band gave $\alpha=14\mathrm{h}11\mathrm{m}12.6\mathrm{s}$6 deg and $\alpha=14\mathrm{h}11\mathrm{m}12.6\mathrm{s}$7 deg, while the best-fit disk had centroid $\alpha=14\mathrm{h}11\mathrm{m}12.6\mathrm{s}$8 deg, $\alpha=14\mathrm{h}11\mathrm{m}12.6\mathrm{s}$9 deg, with disk radius $\delta=+79^\circ23'15''$0 deg and corresponding $\delta=+79^\circ23'15''$1 deg (Xin et al., 2022). A 2D Gaussian fit yielded $\delta=+79^\circ23'15''$2 deg, and a uniform ring model fixed to the radio shell used inner radius $\delta=+79^\circ23'15''$3 and outer radius $\delta=+79^\circ23'15''$4 (Xin et al., 2022). The uniform disk provided the best likelihood among the tested templates (Xin et al., 2022).

The same study reported a point-source TS of 23.8 and a uniform-disk TS of 54.1 in the 1–1000 GeV band, corresponding to an approximately $\delta=+79^\circ23'15''$5 detection (Xin et al., 2022). The extension significance was quantified as

$\delta=+79^\circ23'15''$6

equivalent to an approximately $\delta=+79^\circ23'15''$7 preference for an extended source (Xin et al., 2022). The GeV extension was described as closely matching the diffuse radio ring seen by LoTSS, WENSS, and NVSS (Xin et al., 2022).

A later reanalysis with 17 years of LAT data shifted the emphasis from single-template fitting to morphology on sub-degree scales (Greco et al., 17 Jul 2025). Using a $\delta=+79^\circ23'15''$8 radius ROI centered at $\delta=+79^\circ23'15''$9, $l = 118.41^\circ$0, with 0.05° pixels and 10 energy bins per decade, that study found patchy extended residuals over $l = 118.41^\circ$1, with TS peaks at $l = 118.41^\circ$2–$l = 118.41^\circ$3 ($l = 118.41^\circ$4–20) after removing 4FGL J1409.8+7921 from the background model (Greco et al., 17 Jul 2025). The highest TS peak was reported as colocated with the brightest soft X-rays and the isolated H$l = 118.41^\circ$5 filament (Greco et al., 17 Jul 2025). This suggests spatial non-uniformity in the high-energy emissivity rather than a perfectly uniform shell.

4. Spectral properties and emission scenarios

The Fermi-LAT spectrum between 100 MeV and 1 TeV has been modeled with both a single power law and a log-parabola (Xin et al., 2022). In that analysis, the single power law,

$l = 118.41^\circ$6

gave $l = 118.41^\circ$7, photon index $l = 118.41^\circ$8, and integrated photon flux

$l = 118.41^\circ$9

over 100 MeV–1 TeV (Xin et al., 2022). The log-parabola,

$b = +37.03^\circ$0

gave $b = +37.03^\circ$1, with $b = +37.03^\circ$2, $b = +37.03^\circ$3, and integrated photon flux

$b = +37.03^\circ$4

(Xin et al., 2022). The improvement $b = +37.03^\circ$5 relative to the power law corresponds to $b = +37.03^\circ$6 evidence for spectral curvature concentrated at several tens of GeV (Xin et al., 2022).

The spectral energy distribution was constructed in six logarithmically spaced bins between 0.1 and 1000 GeV, with 95% upper limits assigned where $b = +37.03^\circ$7 (Xin et al., 2022). The highest-energy bins were reported as upper limits, reinforcing a turnover or softening at hundreds of GeV (Xin et al., 2022). Assuming $b = +37.03^\circ$8 kpc, the 10–316 GeV luminosity was estimated as $b = +37.03^\circ$9–$14.2'$0 (Xin et al., 2022).

Broadband modeling has not yielded a unique emission mechanism. In the leptonic scenario, synchrotron emission explains the radio flux while inverse Compton scattering on the cosmic microwave background, plus infrared and optical fields, accounts for the $14.2'$1 rays; bremsstrahlung is negligible at the adopted halo density (Xin et al., 2022). The electron distribution was parameterized as

$14.2'$2

with target photon fields including the CMB, an infrared component with $14.2'$3 K and $14.2'$4, and an optical component with $14.2'$5 K and $14.2'$6 (Xin et al., 2022).

Two representative leptonic fits were highlighted (Xin et al., 2022). A fit with $14.2'$7, $14.2'$8 TeV, $14.2'$9, and $28.4'$0 reproduces the radio data but overpredicts low-energy $28.4'$1 rays unless finely tuned (Xin et al., 2022). A harder fit with $28.4'$2, $28.4'$3 TeV, $28.4'$4, and $28.4'$5 better matches the hard GeV spectrum and respects the X-ray upper limits, but is in tension with the radio slope, since $28.4'$6 would imply $28.4'$7 rather than the observed $28.4'$8 (Xin et al., 2022). The same study emphasized that the combination of a flat radio spectrum and a hard GeV spectrum is difficult to reconcile with a simple one-zone leptonic model (Xin et al., 2022).

The hadronic scenario parameterized the proton population as

$28.4'$9

with $0.95^\circ$0 rays produced by neutral-pion decay in $0.95^\circ$1 interactions (Xin et al., 2022). To reproduce the hard GeV spectrum, the proton index had to be very hard, $0.95^\circ$2, with best fit $0.95^\circ$3 and $0.95^\circ$4 TeV (Xin et al., 2022). For the halo density estimate $0.95^\circ$5, the required proton energy was

$0.95^\circ$6

more than an order of magnitude above a canonical supernova kinetic energy of $0.95^\circ$7 erg (Xin et al., 2022). A broken-power-law proton spectrum still required $0.95^\circ$8 (Xin et al., 2022). On that basis, the hadronic model was judged energetically challenging unless the ambient density had been underestimated (Xin et al., 2022).

The 2025 XMM-Newton and LAT study altered this balance by inferring much larger local densities from thermal X-rays (Greco et al., 17 Jul 2025). That work argued that the measured ambient density, together with the patchy morphology of the $0.95^\circ$9-ray emission and the detection of H$\sim 1^\circ$00 filaments, indicates that a hadronic origin is compatible with the $\sim 1^\circ$01-ray flux, though a mixed leptonic-hadronic scenario cannot be excluded (Greco et al., 17 Jul 2025). This suggests that the viability of hadronic emission depends strongly on whether the remnant is expanding into a smooth halo medium or into a structured circumstellar environment.

5. X-ray plasma, optical filaments, and environmental structure

New XMM-Newton observations obtained in 2024 provided the first detailed characterization of the diffuse X-ray emission inside the ring (Greco et al., 17 Jul 2025). Count-rate, O VII equivalent-width, and median photon energy maps isolated three diffuse soft regions—Central, North, and South—plus a harder “Clump” northeast of the Central region (Greco et al., 17 Jul 2025). The diffuse emission was fitted in XSPEC with TBabs$\sim 1^\circ$02vcie and TBabs$\sim 1^\circ$03vnei models, grouped to 25 counts per bin, with two local background choices to bracket systematics (Greco et al., 17 Jul 2025).

For the fiducial Central region, the CIE fits gave $\sim 1^\circ$04 keV for one background choice and $\sim 1^\circ$05 keV for the other; across the bright region, $\sim 1^\circ$06–0.16 keV was found to be robust (Greco et al., 17 Jul 2025). Oxygen abundance was mildly sub-solar, with $\sim 1^\circ$07 or $\sim 1^\circ$08 depending on background, while nitrogen was sub-solar but weakly constrained (Greco et al., 17 Jul 2025). The best-fit thermal normalization was $\sim 1^\circ$09 or $\sim 1^\circ$10, and the absorbed 0.3–1 keV flux was $\sim 1^\circ$11 (Greco et al., 17 Jul 2025). NEI models did not improve the fits, although NEI could not be excluded because of a $\sim 1^\circ$12–$\sim 1^\circ$13 degeneracy (Greco et al., 17 Jul 2025).

The authors derived the electron density from the XSPEC thermal normalization,

$\sim 1^\circ$14

assuming $\sim 1^\circ$15 and an ellipsoidal geometry for the Central region (Greco et al., 17 Jul 2025). For $\sim 1^\circ$16 kpc, filling factor $\sim 1^\circ$17, and $\sim 1^\circ$18, they obtained $\sim 1^\circ$19 (Greco et al., 17 Jul 2025). Allowing for the fitted normalization range, distances of 4–5 kpc, and $\sim 1^\circ$20 led to $\sim 1^\circ$21–$\sim 1^\circ$22, corresponding to pre-shock density $\sim 1^\circ$23–$\sim 1^\circ$24 for compression ratio 4 (Greco et al., 17 Jul 2025). These values are orders of magnitude larger than the smooth-halo estimate used in the earlier $\sim 1^\circ$25-ray modeling (Xin et al., 2022, Greco et al., 17 Jul 2025).

The “Clump” region is spectrally harder than the diffuse soft component (Greco et al., 17 Jul 2025). It can be fitted either by a thermal model with $\sim 1^\circ$26–1.7 keV or by a power law with $\sim 1^\circ$27, with statistically comparable fit quality (Greco et al., 17 Jul 2025). Its location close to the LAT TS maximum and to the isolated H$\sim 1^\circ$28 smudge was interpreted as suggestive of a site where the shock meets denser material or amplified turbulence (Greco et al., 17 Jul 2025).

Optical imaging with TNG/DOLORES reached a $\sim 1^\circ$29 H$\sim 1^\circ$30 surface-brightness sensitivity of $\sim 1^\circ$31 (Greco et al., 17 Jul 2025). No H$\sim 1^\circ$32 filaments were detected along the outer radio shell, but the previously noted smudge was recovered as an approximately $\sim 1^\circ$33-long filamentary complex with two bright heads and faint diffuse emission on the border of the X-ray bright patch (Greco et al., 17 Jul 2025). Additional faint filaments were present nearby, with no continuum counterparts (Greco et al., 17 Jul 2025). The morphology was described as clumpy rather than a smooth Balmer-dominated arc, which is consistent with radiative shocks in denser pockets, although spectroscopy is required for a definitive classification (Greco et al., 17 Jul 2025).

This X-ray and optical evidence supports a structured medium rather than a uniform halo. The 2025 study concluded that the remnant is expanding in a tenuous environment but has encountered a denser phase, likely the relic of wind activity from a massive progenitor star (Greco et al., 17 Jul 2025). A plausible implication is that different sectors of the shell may sample different densities, magnetic fields, and radiative regimes, which would naturally complicate single-zone broadband modeling.

6. Distance, age, relation to Calvera, and broader significance

The relation between G118.4+37.0 and the Calvera pulsar is central to the source’s astrophysical significance. In the discovery paper, the pulsar lay $\sim 1^\circ$34 from the geometric center of the radio ring (Arias et al., 2022). At the NICER-based distance estimate of $\sim 1^\circ$35 kpc for Calvera, this corresponds to a projected offset of about 4.7 pc, while the ring has outer radius about 27.3 pc and diameter about 54.5 pc (Arias et al., 2022). The pulsar has period $\sim 1^\circ$36 ms, spin-down rate $\sim 1^\circ$37, characteristic age $\sim 1^\circ$38 kyr, spin-down power $\sim 1^\circ$39, and surface dipole field $\sim 1^\circ$40 G; it is radio-quiet and no confirmed LAT pulsations have been detected (Xin et al., 2022).

The distance and age of the SNR have evolved across the literature. The early work adopted $\sim 1^\circ$41 kpc, inherited from Calvera’s X-ray atmosphere modeling, and used a Sedov scaling

$\sim 1^\circ$42

with $\sim 1^\circ$43 and $\sim 1^\circ$44 (Arias et al., 2022). Using what that paper stated as an “SNR radius of 54 pc,” the inferred age was $\sim 1^\circ$45 yr (Arias et al., 2022). The same synthesis also noted that elsewhere the paper quoted 54 pc as a diameter, so the 7.7 kyr estimate depends on treating 54 pc as the shock radius (Arias et al., 2022). In the 2022 $\sim 1^\circ$46-ray analysis, use of Sedov-Taylor arguments and low $\sim 1^\circ$47 led instead to $\sim 1^\circ$48 yr, explicitly noting sensitivity to the density assumption (Xin et al., 2022).

The 2025 XMM-Newton study derived a different physical picture (Greco et al., 17 Jul 2025). From the post-shock temperature relation

$\sim 1^\circ$49

the measured $\sim 1^\circ$50 keV implies $\sim 1^\circ$51, with range $\sim 1^\circ$52–370 km s$\sim 1^\circ$53 for $\sim 1^\circ$54–0.16 keV (Greco et al., 17 Jul 2025). Using the observed angular radius $\sim 1^\circ$55 and distance 4–5 kpc gives a physical radius $\sim 1^\circ$56–41 pc (Greco et al., 17 Jul 2025). A naïve Sedov age from $\sim 1^\circ$57 then yields $\sim 1^\circ$58–45 kyr, but the authors emphasized that the measured $\sim 1^\circ$59 pertains to a denser interior circumstellar patch rather than necessarily to the forward shock, so the true blast-wave speed is likely larger and the true age smaller (Greco et al., 17 Jul 2025). Solving the Sedov relation with the X-ray-derived density range and canonical explosion energy gave a favored age of $\sim 1^\circ$60–20 kyr and a preferred distance of 4–5 kpc (Greco et al., 17 Jul 2025).

This later distance estimate remains consistent within uncertainties with that inferred for Calvera (Greco et al., 17 Jul 2025). The pulsar’s proper motion was quoted as $\sim 1^\circ$61, directed away from the ring center (Greco et al., 17 Jul 2025). A minimal kinematic estimate using angular offset $\sim 1^\circ$62 and age 10–20 kyr gives

$\sim 1^\circ$63

which was described as compatible with observed neutron-star kicks (Greco et al., 17 Jul 2025). Using the directly measured proper motion gives

$\sim 1^\circ$64

so $\sim 1^\circ$65–1800 km s$\sim 1^\circ$66 for $\sim 1^\circ$67–5 kpc, large but not unprecedented (Greco et al., 17 Jul 2025). The association therefore remains plausible but not yet dynamically closed.

In the broader SNR context, G118.4+37.0 has been discussed as atypical. One 2022 study proposed that it and similar isolated remnants could belong to a radio-dim SNR population evolving in low-density environments and showing hard GeV emission of leptonic origin (Araya, 2022). Another emphasized that its evident GeV curvature and absence of non-thermal X-ray emission make it an interesting source bridging young-aged SNRs with bright non-thermal X-ray emission and old-aged SNRs interacting with molecular clouds (Xin et al., 2022). The 2025 multi-wavelength study instead classified it as a middle-aged remnant expanding in a tenuous medium but interacting with denser clumps, with hadronic $\sim 1^\circ$68-ray production becoming viable once the higher X-ray-derived density is taken into account (Greco et al., 17 Jul 2025).

The main unresolved issues follow directly from these differing inferences. The detailed extended-source $\sim 1^\circ$69-ray spectrum remains uncertain in the more recent LAT reanalysis, CIE and NEI X-ray descriptions are not decisively distinguished, and H$\sim 1^\circ$70 imaging alone cannot determine whether the filaments are Balmer-dominated or radiative (Greco et al., 17 Jul 2025). Even so, the combination of a $\sim 1^\circ$71 radio shell, extended GeV emission, soft thermal X-rays interior to the ring, and close association with Calvera establishes G118.4+37.0 as a significant laboratory for studying SNR evolution, particle acceleration, and neutron-star birth environments at high Galactic latitude (Arias et al., 2022, Araya, 2022, Xin et al., 2022, Greco et al., 17 Jul 2025).