Spectroscopic identification of water emission from a main-belt comet

The study investigates whether the recurrent activity observed in main-belt comets (MBCs) is driven by volatile sublimation, specifically testing for direct gas emission from MBC 238P/Read. Prior to this work, no volatile gas had been detected around MBCs despite strong circumstantial evidence for sublimation from recurring dust comae and tails near perihelion. Using the James Webb Space Telescope (JWST), the authors aimed to obtain sensitive near-infrared spectroscopy capable of detecting key cometary volatiles, notably H2O and CO2, which are typically the most prominent in near-infrared spectra of comets. The observations were timed near Read’s expected peak brightness to maximize detectability. Detecting H2O would confirm sublimation as the driver of activity, while constraining CO2 would contextualize MBC volatile inventories relative to classical comets and inform their origins and thermal histories.

• Composition of comets commonly includes abundant H2O, CO2, and CO; water and CO2 are readily detectable in near-IR spectra (prior surveys and reviews cited).
• Previous attempts to detect volatiles in main-belt comets resulted in non-detections, including sensitive limits from Herschel observations of water vapor and upper limits inferred indirectly from CN non-detections using assumed CN/H2O ratios typical of classical comets.
• Comparative spectra from comet 103P/Hartley 2 (Deep Impact) show typical strong H2O (2.7 µm) and CO2 (4.3 µm) bands; MBC 238P/Read deviates by lacking detectable CO2.
• Prior main-belt comet studies suggested CN/H2O ratios of classical comets may not apply to MBCs, potentially biasing inferred water production limits.
• 3 µm absorption features in small bodies have various proposed origins: water ice, water-ice coatings on grains, irradiated/processed organics, ammonium salts; prior work on asteroids (e.g., Themis) and comet 67P informs interpretation of Read’s rounded 3 µm band.
• Dynamical studies indicate some MBCs are native to the main belt, not recent implants from the outer Solar System; thermal models predict depletion of highly volatile species like CO2 for objects residing in the outer main belt over Myr timescales.

• Target and timing: JWST observations of 238P/Read on 2022-09-08 16:30 UTC, 95 days post-perihelion, near peak activity. Geometry: heliocentric distance r = 2.428 AU; observer distance Δ = 2.086 AU; phase angle α = 24.3°; true anomaly v = 28.3°.
• Instruments and configurations:
  • NIRSpec IFU with prism disperser covering 0.6–5.2 µm; resolving power varying from ~100 at 0.6 µm, down to ~30 near 1.2 µm, up to ~300 near 5.2 µm. Total exposure 3,210 s across four integrations with ~0.1″ dithers; observatory tracked non-sidereal rates.
  • NIRCam imaging in F200W and F277W (effective wavelengths 1.97 and 2.74 µm; 24% bandwidth). Five dithered exposures per filter; total 1,020 s per filter; pixel scales 0.031″/pix (short-λ), 0.063″/pix (long-λ).
• Data reduction:
  • NIRSpec: JWST pipeline v1.9.4, CRDS context 1041. Off-target sky frames 42″ away used for background subtraction; no comet signal in sky. Spectra extracted within 0.3″ radius aperture centered on inner coma. In-scene background subtraction tested; due to strong H2O gas in background, gas-band analyses used spectra without in-scene subtraction. Four spectra averaged with outlier rejection. Absolute spectrophotometric uncertainty adopted as 10%.
  • NIRCam: pipeline v1.6.2, CRDS 969; photometric calibration updated (2022-10-06). Aperture photometry in 0.3″ radius apertures with WebbPSF aperture corrections. Measured AB magnitudes: F200W = 22.84 ± 0.03; F277W = 23.22 ± 0.05. Synthetic photometry from NIRSpec spectrum consistent with NIRCam color (m(F200W) − m(F277W) ≈ −0.39 mag).
• Continuum and reflectance analysis:
  • Reflectance spectrum produced by dividing by a solar spectrum; coma shows red slope of 2.18 ± 0.02% per 100 nm between 1.0–2.55 µm (normalized at 2.0 µm). Thermal contribution modeled as scaled Planck function added to linear reflectance; fits indicate thermal emission contributes ~3–5% at 3.7 µm, not significantly affecting 3 µm band shape analysis.
  • Nucleus contribution estimated using effective radius R = 0.24 ± 0.05 km and nominal thermal model: nucleus dominates thermal emission at 5.0 µm (98 ± 37% of flux); contributes ~21 ± 8% of reflected flux at 2.0 µm.
• Gas fluorescence modeling:
  • Synthetic fluorescence spectra for H2O (ν2 band at ~2.7 µm) and CO2 (ν2 band at ~4.3 µm) compared to continuum-subtracted data to derive rotational temperatures and production rates. Water models with Trot = 15 and 25 K fit the observed band. CO2 model used to set an upper limit.
• Imaging products:
  • NIRCam images show coma and tail; NIRSpec IFU maps built for reflected light, H2O emission, and a continuum temperature proxy using the ratio of thermal (4.1–5.2 µm) to scattered (0.7–2.5 µm) light. Water emission appears asymmetric and predominantly sunward; temperature peak offset by ~0.1″ from nucleus.
• Active area estimation:
  • Water-ice sublimation models of nucleus used to infer active area from measured Q(H2O) under slow- and rapid-rotator assumptions, yielding 0.03–0.11 km2; slow-rotator favored due to sunward asymmetry.
• Comparative spectral analysis:
  • Read’s spectra compared with comet 103P/Hartley 2 (coma) and with surface spectra of comet 67P and asteroid (24) Themis after detrending and scaling band depths to examine the 2.8–3.7 µm absorption feature morphology.

• First clear detection of gas from a main-belt comet: strong H2O ν2 emission at 2.7 µm detected from 238P/Read.
  • Water production rate Q_H2O = (9.9 ± 1.0) × 10^24 molecules s^−1, equivalent to 0.30 ± 0.03 kg s^−1. Water coma is asymmetric and predominantly sunward.
• No significant CO2 detected:
  • Upper limit Q_CO2 < 7 × 10^22 molecules s^−1 (99.7% confidence), equivalent to <5 g s^−1.
  • Coma abundance ratio CO2/H2O < 0.7% (99.7% CL), about ten times lower than typical comets at similar heliocentric distances and three times lower than the lowest previous spectroscopic measurement overall.
• Sublimating area and activity level:
  • Active area estimated at 0.03–0.11 km^2 depending on thermal/rotation state; slow-rotator solution preferred. With nucleus radius R = 0.24 ± 0.05 km, active fraction ~4–15%, comparable to typical comets when scaled to its small size.
• Reflectance and 3 µm band:
  • Broad, rounded absorption from ~2.8 to 3.7 µm with minimum near ~3.2 µm; lacks accompanying 1.5 and 2.0 µm water-ice bands. Band shape resembles processed organics/ammonium-bearing materials more than pure water-ice signatures; not a perfect match to 103P, 67P, or Themis analogs.
• Dynamical and evolutionary context:
  • Read’s orbit is consistent with outer main-belt origin, not a recent Jupiter-family comet implant; strong CO2 depletion aligns with thermal models for long residence in the main belt (≥1 Myr). Read is associated with the low-albedo Gorchakov asteroid family.
• Mass-loss characteristics and evolution:
  • Dust-to-ice mass-loss rate ratio ~0.3. Analysis suggests subsurface ice retreats faster than surface, implying gradual quenching of activity; observed activity appears to be declining from orbit to orbit.
• Impact-driven localized source scenario requiring a ~10 m impactor to expose ~100 m-radius patch is considered but appears unlikely to avoid catastrophic disruption for a nucleus as small as Read (though sub-catastrophic parameters may be tuned).

The detection of H2O vapor directly confirms that the recurrent activity of main-belt comet 238P/Read is driven by water-ice sublimation, resolving a long-standing question posed by earlier dust-only detections. The stringent upper limit on CO2 reveals an exceptionally CO2-depleted coma, distinguishing Read—and plausibly MBCs more broadly—from classical comets, which typically display both H2O and CO2 bands at comparable heliocentric distances. This chemical distinction suggests different formation conditions and/or prolonged thermal evolution within the main belt rather than recent implantation from outer reservoirs. The inferred active area and active fraction are consistent with sublimation from modest surface patches on a small nucleus, in line with typical cometary surface behaviors when scaled by size. The sunward asymmetry of the H2O coma supports a slow-rotator thermal regime. The broad, rounded 3 µm absorption feature without corresponding 1.5 and 2.0 µm ice bands points to complex surface or coma grain composition (processed organics and/or ammonium-bearing materials) rather than exposed pure water ice, aligning with interpretations from 67P and certain low-albedo asteroids. The low CO2/H2O ratio further implies that previous indirect water constraints for MBCs based on CN non-detections may have been underestimated if MBCs are depleted in those tracer species as well. Dynamical analyses corroborate a main-belt residency and predict CO2 loss over Myr timescales, matching the observations. The measured dust-to-ice ratio and declining activity indicate ongoing devolatilization and an evolutionary trajectory toward quiescence unless surface renewal processes (e.g., impacts or YORP-driven mass movement) intermittently rejuvenate activity.

This study provides the first unambiguous spectroscopic detection of water vapor from a main-belt comet and sets a stringent upper limit on CO2, demonstrating that 238P/Read’s activity is driven by H2O sublimation while being strongly depleted in CO2 compared to classical comets. These results show that MBCs likely represent a distinct volatile reservoir, preserving materials and evolutionary pathways not captured by traditional comet or meteoritic samples, and they are crucial for reconstructing the Solar System’s volatile inventory and processing history. Future work should include systematic JWST (and complementary) spectroscopy of additional MBCs to map compositional diversity, time-resolved monitoring across orbits to quantify activity evolution and dust-to-gas ratios, higher S/N studies of the 3 µm band to constrain grain composition and processing, and refined thermophysical and dynamical modeling to link observed volatile depletion to residence times and surface renewal mechanisms.

• CO2 was constrained only by an upper limit; actual abundance may be below detection threshold.
• Absolute spectrophotometric calibration uncertainty of ~10% for NIRSpec; potential systematic effects from in-scene background subtraction choices.
• Spatial calibration and temperature mapping from IFU data show an unexpected offset; improvements in instrument calibration may refine interpretations.
• Active area estimates depend on unknown thermal inertia, emissivity, and rotation state (slow vs rapid rotator), introducing model-dependent uncertainty.
• Nucleus contribution estimates assume spherical shape, nominal thermal properties, and coma-nucleus color similarity.
• The 3 µm absorption feature interpretation is non-unique; multiple compositions and grain properties can reproduce rounded band shapes.
• Results are based on a single-epoch observation near peak activity; temporal variability across the active arc is not directly sampled.
• Impact scenario assessment relies on scaled simulations and assumptions about material strength and porosity, with uncertainties for a small nucleus like Read.