Enceladus is a prime target in the search for life in our solar system, having an active plume likely connected to a large liquid water subsurface ocean. Using the sensitive NIRSpec instrument onboard JWST, we searched for organic compounds and characterized the plume’s composition and structure. The observations directly sample the fluorescence emissions of H2O and reveal an extraordinarily extensive plume (up to 10,000 km or 40 Enceladus radii) at cryogenic temperatures (25 K) embedded in a large bath of emission originating from Enceladus' torus. Intriguingly, the observed outgassing rate (300 kg/s) is similar to that derived from closeup observations with Cassini 15 years ago, and the torus density is consistent with previous spatially unresolved measurements with Herschel 13 years ago, suggesting that the vigor of gas eruption from Enceladus has been relatively stable over decadal timescales. This level of activity is sufficient to maintain a derived column density of 4.5×1017 m-2 for the embedding equatorial torus, and establishes Enceladus as the prime source of water across the Saturnian system. We performed searches for several non-water gases (CO2, CO, CH4, C2H6, CH3OH), but none were identified in the spectra. On the surface of the trailing hemisphere, we observe strong H2O ice features, including its crystalline form, yet we do not recover CO2, CO nor NH3 ice signatures from these observations. As we prepare to send new spacecraft into the outer solar system, these observations demonstrate the unique ability of JWST in providing critical support to the exploration of distant icy bodies and cryovolcanic plumes.

Enceladus is likely the largest source of water within the Saturnian system1, with H2O and other materials jetted into Saturn orbit by localized geological activity2,3. Early hints of geological activity on Enceladus were provided by Voyager and telescopic observations in the 1980s and 1990s, finding a close association between Saturn’s E ring and Enceladus’ orbit4–8. In 2005, multiple instruments onboard the Cassini spacecraft discovered a plume of gases (predominately water vapor) and ice grains emerging from fissures in the south polar region of Enceladus9–14, while a torus of water along Enceladus’ orbit was most recently observed via sub-millimeter spectroscopy with the Herschel Observatory15. The Cassini measurements of the plume gas were made using in situ mass spectrometry along specific flyby trajectories16,17 and via stellar occultation in the inner region of the plume (<200 km)9,18. In contrast, the sub-millimeter measurements of the torus were not spatially resolved, but they indicated the presence of H2O gas widely throughout the Saturnian system. While the plume’s flux of icy grains varies on multiple timescales19, the variations in the vapor flux are much less well understood, together with how these affect the structure and evolution of the torus. By analyzing the molecular emissions across large distances from Enceladus with JWST20, we were able to map the distribution of outgassed water, compare the level of activity to that determined by Cassini measurements, and establish a direct connection of the plume to the extended cloud of material beyond the plume that likely accumulated over multiple orbits.

On November 9th 2022 UT, we observed Enceladus’ trailing hemisphere with JWST as part of the Solar System Guaranteed-Time-Observations (GTO) program 1250. The JWST/NIRSpec observations were made with the Integral Field Unit (IFU)21, delivering a data-cube across three high-resolving power gratings (G140H, G235H, G395H), two detectors per grating (NRS1, NRS2), with a uniform spaxel size of 0.1ʺx0.1ʺ across a 3ʺx3ʺ field-of-view (FOV) – Enceladus was 0.07ʺ in diameter at the time of the observations. To minimize saturation, we employed the NRSRAPID readout and short integration times per frame, totaling 215-270 seconds of integration per grating. The data were processed employing the latest version of the JWST Science Calibration Pipeline (v1.9), and we developed ad-hoc algorithms to analyze the frames, combine dithering images, and clean bad pixels (scripts publicly available at github.com/nasapsg). In Figure 1, we present flux-calibrated spectra for the integrated signal across the Enceladus disk and molecular residuals along several regions of the extended plume.

The Enceladus disk spectrum as observed with JWST is dominated by water (H2O) ice, including its crystalline form (features at 1.65 and 3.1 μm), with the main features similar to those identified using Cassini’s Visible and Infrared Mapping Spectrometer (VIMS)2 and from groundbased observations22,23. Carbon dioxide (CO2) ice features were previously identified near 2.7 μm and 4.3 μm on the surface of Enceladus and primarily at southern polar latitudes (80-90°S) from VIMS2,24, but in the JWST spectrum we do not identify either of these CO2 ice bands. The non-detection of CO2 on Enceladus with JWST is likely due to the observing geometry at the time of the observations (centered at 16°N, 270°W), which did not sufficiently sample Enceladus’ southern polar regions, further establishing that CO2 ice is likely confined to the south polar terrain and probably replenished by plume CO22,25. We do observe a potential signature near 4.5 μm, which could be related to CN compounds, yet this identification is only tentative from these data. More information about this and other possible ice identifications is given in the methods section (M2).

We also searched for ammonia (NH3) ice, which can strongly depress the freezing point of water and could indicate a sub-surface liquid water reservoir that feeds the plume17, but we did not identify it in the spectrum (either amorphous or crystalline). Spectroscopic searches for crystalline NH3 ice in the near infrared have mostly focused on a feature near 2.24 μm, attributed to a combination mode. Ground-based observations of the trailing side of Enceladus detected a subtle band centered near 2.25 μm that was tentatively attributed to NH3 ice, but other observations obtained in 1995 and 1998 did not display this feature22. We did notice a weak and broad feature spanning ~2.15 to 2.2 μm that has been attributed to NH3-H2O mixtures and ammonium (NH4) bearing species on other icy moons26. However, this absorption was inconclusive between dithers and not consistent with the 2.24 μm expected location for NH3 ice, nor the previously detected 2.25 μm feature on Enceladus. Additionally, there is no clear evidence of the ~2.0 μm absorption band due to solid ammonia and its hydrates27 and expected to overlap with the much stronger 2.0 μm water ice band. We also searched for the stronger NH3 ice band at 2.96 μm28 but the NIRSpec spectra do not show any absorption at this wavelength. Our non-detection of NH3 ice agrees with the non-detection of surface NH3 on a global scale by Cassini VIMS2.

The uniqueness of JWST for exploring Enceladus is most evident when probing with unparalleled sensitivity the narrow infrared emissions emanating from the plume. Such probing is possible thanks to JWST’s very low operational temperatures (~40 K), broad wavelength coverage (0.6 to 28.3 μm), large collecting area (25.4 m2), high spectral resolving power (up to l/dl~3000) and advanced infrared instrumentation (e.g., IFU, multi-grating). Infrared gas emissions at these wavelengths are primarily due to solar-pumped fluorescence, so they are particularly weak at these large heliocentric distances. Furthermore, the molecular features are narrow and confined to at most twice the expansion velocity (vexp ~ 540 m/s)1 requiring highresolution spectroscopy, which is sensitively done with JWST since we can also probe the strong molecular fundamentals of H2O and CO2 (not accessible to ground-based observatories due to the lack of atmospheric transparency). Across the Enceladus plume, we detect several strong H2O molecular emissions (Figure 1) – also at the spaxel level (Figure 2) – revealing a highly localized source oriented southward and extending out to at least 10,000 km (40 RE). We retrieved column densities from the measured line fluxes by employing the publicly available Planetary Spectrum Generator (PSG) tool at psg.gsfc.nasa.gov29,30, which integrates non LTE (Local Thermodynamic Equilibrium) fluorescence radiative-transfer models31–33 (see S1). From the relative intensities of the water ro-vibrational lines, we determined that the water molecules are at a rotational temperature of Trot = 25 ± 3 K, and this temperature is consistent across the different regions explored (i.e., local plume, extended plume, background emission) by our measurements (Figure 1). These observed temperatures refer to the rotational excitation temperature, not to the ambient kinetic temperature of H2O vapor. Due to the large extension of the plume and low local densities, this sparse collisional regime dictates that the molecules are no longer in LTE but in an equilibrium state defined by the insolation rate and the intrinsic probability of absorption or emission of the molecule.

The observed map of water column densities can be relatively well modeled30 (see Figure 2) when considering an ejecta plume with an outgassing rate of (1.0±0.1)×1028 molecules per second (300 kg/s), a 3D ejecta cone half-width of 40° as inferred from these data, and an expansion velocity of 540 m/s (an intermediate value between the ~400 m/s thermal velocity3, the vertical velocity of ~600 m/s as inferred from near-surface measurements34, and the ~700 m/s flow velocities inferred from an outgassing model16). This observed level of activity is similar to that inferred from measurements1,18 made in Saturn orbit by Cassini 6-19 years ago (1×1028 molecules per second), although some other analyses suggest large temporal variability in outgassing from Enceladus16,35. We also observe a relatively constant background water emission across the whole FOV with an average column density of (1.7±0.1)×1018 m-2. Such emission most likely comes from water molecules within the torus, which is being observed at the ansa of the orbital path of Enceladus with an inclination of 15.2°. If we assume a constant average density within the torus and zero density outside, and take into account that the torus minor radius or scale height is considerably smaller than its major radius (~237,000 km, centered on Enceladus’ orbit), then the equatorial column density cutting perpendicular through the torus center can be estimated to be NT ~ 4.5×1017 m-2 (1.7×1018 m-2 × sin 15.2°). Our measurement is remarkably consistent with that inferred from sub-millimeter observations15 obtained 13 years ago, which indicated an equatorial column density of 4×1017 m-2 and a torus scale height of HT ~ 25,000 km.

As Enceladus orbits rapidly around Saturn with a period of only 1.37 Earth days, the ejected water vapor is spread along and around its orbit, forming a large torus around Saturn. Considering that the photochemical lifetime36 of water near Saturn is relatively long (~94 days) and combining this value with the derived production rate of the plume, we estimate that up to 8×1034 molecules are available at a given time. Alternatively, and from our derived torus equatorial column density (NT) and the inferred torus scale height (HT), we estimate that a total of 2.5×1034 molecules are confined within the torus, equivalent to 32% of the ejected molecules. This would mean that a large fraction of the ejected H2O molecules (and its OH and O products) are spread beyond the torus and across the Saturnian system. These results generally agree with the findings by Cassidy and Johnson1, establishing Enceladus as the dominant source of exogenous H2O / OH / O species in the Saturnian system. In addition to H2O vapor, we searched for CO2, CO, CH4, C2H6 and CH3OH molecular emissions across the plume, but none were detected (see extended Figure 1 and the methods section M1). Upper limits (3s) on their abundances are respectively: <1%, <10%, <4%, <6% and <20% relative to water. These limits are within the abundances reported from Cassini INMS measurements25,37 of the dense plume region of Enceladus (CO2:0.3-0.8%, CO<0.05%, CH4:0.1-0.3%, C2H6<0.2%, CH3OH<0.01%). UVIS occultations also showed no evidence of CO (<1%) in the inner regions of the plume18. Our upper limit on the CO2/H2O ratio provides additional support for the idea38 that extensive CO2 sequestration in Enceladus’ rocky core is probably needed to explain why its plume is significantly depleted in CO2 compared with cometary observations39,40.

These first observations with JWST (only a few minutes of integration time) demonstrate the power of this observatory for sensitively characterizing this ocean world, opening a new window into the exploration of Enceladus’ ongoing plume activity while preparing for future missions41. More generally, JWST can provide detailed quantitative insights into H2O vapor-dominated geological and cryovolcanic activity elsewhere in the solar system. This work was supported by NASA’s Goddard Astrobiology Program, Goddard’s Fundamental Laboratory Research (FLaRe), and the Sellers Exoplanet Environments Collaboration (SEEC). This work is based on observations made with the NASA/ESA/CSA James Webb Space Telescope. The data were obtained from the Mikulski Archive for Space Telescopes at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-03127 for JWST. These observations are associated with program #1250. SNM, HBH and NRG acknowledge support from NASA JWST Interdisciplinary Scientist grant 21-SMDSS21-0013. CRG was supported by Southwest Research Institute Internal Research & Development grant 15-R6248 and NASA Astrobiology Institute grant NNN13D485T. KPH acknowledges support from the NASA Astrobiology Program (award #80NSSC19K1427) and the Europa Lander Pre-Project, managed by the Jet Propulsion Laboratory, California Institute of Technology, under a contract with NASA.

The data used in this analysis are publicly available at the Space Telescope Science Institute (STScI) JWST archive (https://mast.stsci.edu/), program #1250.

The retrieval software package used in this study is the Planetary Spectrum Generator, which is free and available online at https://psg.gsfc.nasa.gov29,30, with the data-reduction scripts available at https://github.com/nasapsg. Figures were made with Matplotlib version 3.2.142, available under the Matplotlib license at https://matplotlib.org/.

The authors declare no competing interests.

Villanueva, Hammel, Milam, Hand, Paganini, Spencer, Stansberry and Strazzulla designed the observations and prepared the observational plans. Villanueva, Kofman, Faggi, Cartwright, Stansberry, Holler, Protopapa, Liuzzi, Hedman and Denny analyzed the data, extracted calibrated spectra, produced maps and performed retrievals. Glein, Roth, Rowe-Gurney, Cruz-Mermy, and El-Moutamid assisted with the interpretation of the results, and provided context to related mission and astronomical investigations. All authors contributed to the preparation, writing and edition of the manuscript.

In order to search for narrow molecular features, we analyzed the residual spectra which were derived by subtracting a continuum model that included solar Fraunhofer lines from the observed Enceladus spectra. We performed integrations across several regions throughout the environment of Enceladus (see Figure 1) and determined the number of molecules and corresponding column densities. We detect signatures of water vapor in all regions, with the most prominent molecular features and detections across the plume region (7<r<30 RE, 0.2 to 1.0”, 1765 to 7563 km). This region also has a low intrinsic continuum signature from Enceladus’ disk which simplifies the removal of the non-gas signatures. We therefore used this region to search for other gases beyond H2O. The fluorescence models are based on non-LTE radiative-transfer modeling31–33. Broad non-molecular features were removed by fitting a polynomial function to the continuum shape over the spectral regions presented in Extended Figure 1. Retrievals and the statistical analysis were performed using PSG, in which the retrieval algorithm is based on the Optimal Estimation method 43. After each iteration of the retrieval algorithm, a new model was constructed, and numerical derivatives were computed for each parameter. This process was repeated until convergence was achieved, and the differences between data and model were minimized. The mean statistical variation of the residual spectra (RMS or chi-square) was used to quantify the uncertainty (sigma) in the retrieved column densities.

The trailing hemispheric spectrum of Enceladus measured with NIRSpec does show some interesting features, which could provide additional information regarding the composition and physical properties of Enceladus’ surface. Yet these findings are not conclusive at this stage and additional laboratory experiments, observations, analysis, and modeling would be required to further establish the significance of these findings.

Origin of the water crystalline features: The presence of the 1.65 μm and 3.11 μm features that we observe in the Enceladus spectrum (Figure 1), clearly testifies to the crystallinity of the H2O ice44,45. Interestingly, laboratory spectra46 (1.3 to 2.5 μm) of a thin film of crystalline ice deposited at 150 K and cooled down to 16 K show this 1.65 μm crystalline band. On the other hand, crystalline ice formed at higher temperatures does not exhibit this feature, which is also absent in amorphous ice44. From the relative intensities of these features and their observed central wavelengths one could obtain constraints on the formation and current temperature of the observed ices, yet this would also need detailed modeling of the other strong nearby features that impact the shape of the nearby continuum.

Hydrogen peroxide-bearing ice: We observe a subtle “plateau” at 3.5 μm (see Figure 1), which could be attributed to hydrogen peroxide (H2O2), yet proper recovery of this feature would require accurate modeling of the nearby strong water bands. It is well known that the icy satellites within the Saturnian and Jovian magnetospheres are subjected to intense fluxes of energetic particles that alter their surficial properties and induce many physical and chemical effects47,48. Although modeling how efficiently radiation processes affect Enceladus' surface quantitatively will need to account for the relatively low energetic proton fluxes49,50, and relatively high particle deposition rates of plume fallout51 experienced by the Saturnian moon. Among the several effects studied in the laboratory is the formation of H2O2 evidenced from the appearance of a band at about 3.5 μm in the spectrum of water ice irradiated with energetic particles52–54. This 3.5 μm H2O2 feature has been found on the surface of Europa55,56, while ultraviolet observations made by the Galileo Ultraviolet Spectrometer suggest that H2O2 may be present on Ganymede and Callisto as well57. Consequently, the possible 3.5 μm feature on Enceladus could be consistent with radiolytic generation of H2O2 from its H2O ice rich surface. CN compounds: A possible absorption is observed near 4.5 μm (see Figure 1) and could be associated to CN and/or (iso)cyanate compounds, yet baseline issues for these low flux continuum levels makes this identification only tentative. Radiolytic and photochemical processing of C-containing ices results in the formation of an organic material that, formed at low temperatures, evolves during the heating of the samples and yields an organic refractory material that is stable at room temperature and above58. When the original ice contains both C and N atoms, the stable residue exhibits a strong and clear feature centered at about 4.6 μm that effectively reproduces the features observed in some ultra-carbonaceous Antarctic meteorites59. That feature is attributed to cyanate and isocyanate bonds and is considered evidence of the energetic processing (by photons, electrons or ions) of the ices. Similarly, a feature centered near 4.57 μm was detected on Callisto and attributed to CN-bearing organics60. Such ices on Enceladus could be deposited on the surface by micrometeoritic bombardment or by energetic processing of reduced C- and N-bearing materials. Another possibility is that CN compounds might already be present in Enceladus’ plume37, and could then be deposited over the surface. 31. Villanueva, G. L. et al. Water in planetary and cometary atmospheres: H2O/HDO transmittance and fluorescence models. Journal of Quantitative Spectroscopy and Radiative Transfer 113, 202–220 (2012). 32. Villanueva, G. L., Disanti, M. A., Mumma, M. J. & Xu, L.-H. A Quantum Band Model of the ν3 Fundamental of Methanol (CH3OH) and its Application to Fluorescence Spectra of Comets. The Astrophysical Journal 747, 1–11 (2012). 33. Villanueva, G. L., Mumma, M. J. & Magee-Sauer, K. Ethane in planetary and cometary atmospheres: Transmittance and fluorescence models of the ν7 band at 3.3 μm. Journal of Geophysical Research 116, 1–23 (2011). 34. Hansen, C. J. et al. Water vapour jets inside the plume of gas leaving Enceladus. Nature 456, 477–479 (2008). 35. Teolis, B. D. et al. Enceladus Plume Structure and Time Variability: Comparison of Cassini

Observations. Astrobiology 17, 926–940 (2017). 36. Huebner, W. F., Keady, J. J. & Lyon, S. P. Solar photo rates for planetary atmospheres and atmospheric pollutants. Astrophysics and Space Science (ISSN 0004-640X) 195, 1–289 (1992). 37. Postberg, F. et al. Plume and Surface Composition of Enceladus. Enceladus and the Icy

Moons of Saturn 1 29 (2018 ). doi:10.2458/azu_uapress_9780816537075-ch007. 38. Glein, C. R. & Waite, J. H. The Carbonate Geochemistry of Enceladus’ Ocean. Geophysical

Research Letters 47, e2019GL085885 (2020). 39. Pinto, O. H., Womack, M., Fernandez, Y. & Bauer, J. A Survey of CO, CO2, and H2O in

Comets and Centaurs. Planet. Sci. J. 3, 247 (2022). 51. Southworth, B. S., Kempf, S. & Spitale, J. Surface deposition of the Enceladus plume and the zenith angle of emissions. Icarus 319, 33–42 (2019). 52. Moore, M. H. & Hudson, R. L. IR Detection of H2O2 at 80 K in Ion-Irradiated Laboratory

Ices Relevant to Europa. Icarus 145, 282–288 (2000). 53. Gomis, O., Satorre, M. A., Strazzulla, G. & Leto, G. Hydrogen peroxide formation by ion implantation in water ice and its relevance to the Galilean satellites. Planetary and Space Science 52, 371–378 (2004). 54. Loeffler, M. J., Raut, U., Vidal, R. A., Baragiola, R. A. & Carlson, R. W. Synthesis of hydrogen peroxide in water ice by ion irradiation. Icarus 180, 265–27 3 (2006 ). 55. Carlson, R. W. et al. Hydrogen Peroxide on the Surface of Europa. Science 283, 2062–2064 (1999). 56. Trumbo, S. K., Brown, M. E. & Hand, K. P. H2O2 within Chaos Terrain on Europa’s

Leading Hemisphere. AJ 158, 127 (2019). 57. Hendrix, A. R., Barth, C. A., Stewart, A. I. F., Hord, C. W. & Lane, A. L. Hydrogen

Peroxide on the Icy Galilean Satellites. 2043 (1999). 58. Accolla, M. et al. Combined IR and XPS characterization of organic refractory residues obtained by ion irradiation of simple icy mixtures. A&A 620, A123 (2018). 59. Baratta, G. A. et al. Organic samples produced by ion bombardment of ices for the EXPOSE-R2 mission on the International Space Station. Planetary and Space Science 118, 211–220 (2015). 60. McCord, T. B. et al. Organics and Other Molecules in the Surfaces of Callisto and Ganymede. Science 278, 271–275 (1997).