PUBLICATIONS Geophysical Research Letters RESEARCH LETTER 10.1002/2016GL071853 Key Points: • Supersonic gas jets in Enceladus’ water vapor plume loft more particles when Enceladus is at apoapsis • The bulk quantity of gas from fissures across Enceladus’ south pole is relatively unaffected by the position of Enceladus in its orbit Correspondence to: C. J. Hansen, cjhansen@psi.edu Citation: Hansen, C. J., L. W. Esposito, K.-M. Aye, J. E. Colwell, A. R. Hendrix, G. Portyankina, and D. Shemansky (2017), Investigation of diurnal variability of water vapor in Enceladus’ plume by the Cassini ultraviolet imaging spectrograph, Geophys. Res. Lett., 44, doi:10.1002/2016GL071853. Received 11 NOV 2016 Accepted 3 JAN 2017 Accepted article online 5 JAN 2017 Investigation of diurnal variability of water vapor in Enceladus’ plume by the Cassini ultraviolet imaging spectrograph C. J. Hansen1 , L. W. Esposito2, K.-M. Aye2 D. Shemansky4 , J. E. Colwell3, A. R. Hendrix1 , G. Portyankina2 , and 1 Planetary Science Institute, Tucson, Arizona, USA, 2Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, Colorado, USA, 3Department of Physics, University of Central Florida, Orlando, Florida, USA, 4Space Environment Technologies, Altadena, California, USA Abstract An occultation of ε Orionis by Enceladus’ plume was observed with Enceladus at an orbital longitude near apoapsis in order to investigate whether water vapor flow is modulated diurnally, similar to ice particles. The occultation showed that the bulk water vapor emanating from Enceladus changes little with orbital position. The amount of gas in at least one supersonic jet increased significantly, implying that the increase in the number of particles lofted at apoapsis could be due to more gas coming from the supersonic jets and not the overall gas flux from the tiger stripe fissures that cross Enceladus’ south polar region. 1. Introduction Saturn’s moon Enceladus spews water vapor and ice particles out of four fissures (dubbed “tiger stripes”) that cross its south pole [Dougherty et al., 2006; Hansen et al., 2006; Porco et al., 2006; Spahn et al., 2006; Spencer et al., 2006; Waite et al., 2006]. The payload on the Cassini spacecraft, in orbit around Saturn since 2004, has been used to investigate Enceladus’ eruptive activity since its discovery in 2005. Tidal flexing powers the plume (summarized in Spencer and Nimmo, 2013) and appears to affect the source rate for ice particles. The brightness of the ice particle jets, a proxy for the amount of solid material streaming out, varies with Enceladus’ orbital longitude [Hedman et al., 2013; Nimmo et al., 2014; Porco et al., 2014]. This variation with orbital position implicates tidal stresses, which likewise vary with orbital longitude. The brightness of the plume when Enceladus is at apokrone (orbital longitude 180°) is greater than when it is at perikrone (periapsis) by about a factor of 3 [Hedman et al., 2013]. The subject of this paper is whether or not the gas flux varies with Enceladus’ position in its orbit (i.e., diurnally since 1 day is the same as one orbit) in a manner similar to the particles. Cassini’s Ultraviolet Imaging Spectrograph (UVIS) observes occultations of stars by Enceladus’ plume to learn about the composition of the gas (primarily water vapor) and the detailed structure of the plume [Hansen et al., 2006]. In 2007 the UVIS observation of an occultation of ζ Orionis revealed six supersonic gas jets imbedded in the vapor plume, further characterized with a solar occultation in 2010 [Hansen et al., 2008; Hansen et al., 2011]. Observations by Cassini’s cameras show ~100 ice particle jets [Porco et al., 2014], presumably lofted by gas jets. The UVIS resolution precludes detection of such narrow jets so it may be that the collimated supersonic gas jets detected by UVIS correspond to more than one particle jet [Portyankina et al., 2016]. Throughout this paper we use the term “plume” to refer to the bulk gas emanating from the south pole and “jet” to refer to a collimated supersonic gas stream. ©2017. American Geophysical Union. All Rights Reserved. HANSEN ET AL. Occultations observed by UVIS from 2005 to 2011 show only ~15% variability in the total mass of water vapor escaping from Enceladus, in contrast to the particles’ orbital variation. However, in those years UVIS did not observe any occultations near apokrone. An occultation of ε Orionis on 11 March 2016 offered an opportunity to investigate gas flux at a mean anomaly of 208°, closer to apokrone than any previous occultation. This allowed us to test the hypothesis that the gas behaves similarly to the particles, with an increase in flux near apokrone. At a mean anomaly of 180° the visible and infrared mapping spectrometer (VIMS) and imaging science subsystem (ISS) results show a factor of >3 increase in plume brightness over that at perikrone, but the peak is narrow in orbital longitude [Hedman et al., 2013; Ingersoll and Ewald, 2017]. The occultation observed by UVIS was at a mean anomaly of 208°, away from the peak at 180°. A rough estimate for the increase ENCELADUS’ WATER VAPOR PLUME VARIABILITY 1 Geophysical Research Letters 10.1002/2016GL071853 Figure 1. Occultation Geometry. (a) The diagonal track of the star crossing behind the plume and going behind the limb is shown. The position of ε Orionis is indicated by a circle at the time the ray to the star pierced the Baghdad I gas jet, at an altitude of ~40 km. The red line shows the position of the terminator. (b) The ground track, shown in orange, is computed as the intercept of a perpendicular dropped from the ray between the UVIS boresight and the star to the surface of Enceladus. The line of sight is thus perpendicular to the ground track shown. “Ingress” is labeled at the time that the star signal began to dip, “A” corresponds to the Baghdad I gas jet, “end” is the end of data collection, and at “impact” the star went behind the limb. (c) Ground tracks for all the occultations observed by UVIS are shown. The occultations in 2007, 2010, and 2011 made a horizontal cut across the entire plume. From those three cuts the dimensions of the bulk plume can be estimated. The base map is courtesy of Spitale and Porco [2007]. in flux that UVIS would have seen if the gas behaves like the particles is 50% to 100% more, or roughly 300 to 400 kg/s, based on the proximity to this peak, rather than the average of 210 kg/s seen in previous occultations. 2. ε Orionis Occultation Observation The occultation was observed with the far ultraviolet (FUV), extreme ultraviolet, and high-speed photometer (HSP) channels of UVIS [Esposito et al., 2004]. The FUV integration time was set to 1 s to maximize temporal resolution. The spectra were binned to give 512 samples from 111.5 nm to 190 nm, using the low-resolution slit, which has a spectral resolution of 4.8 nm. The geometry and ground track for the 11 March occultation are illustrated in Figure 1. In this occultation the line of sight to the star made a diagonal cut through the plume, illustrated in Figure 1a. Starlight began to be extinguished at 11:52:10.6, at an altitude of 91 km, at the location shown in Figure 1b. Sixteen seconds later, the ray to the star pierced the Baghdad I gas jet. (Although it is now clear that there are ~100 ice particle jets [Porco et al., 2014], we retain the older terminology [Spitale and Porco, 2007] for the gas jets. The gas jets imbedded in the broader plume of water vapor discovered in the 2007 ζ Orionis occultation [Hansen et al., 2008] were thought to correspond to the particle jets identified in Spitale and Porco [2007]. We thus refer to the gas jets with the names shown in Figure 1, with the understanding that there is not a 1:1 correspondence to a particular ice particle jet.) UVIS data collection ended ~6 km above the Cairo fissure, at a latitude/longitude of 82.0/151.8, shortly before the star went behind the limb of Enceladus. As expected, no gas was detected when the star reemerged from behind the limb near the equator. 3. Water Vapor Column Density and Mass Flux The FUV data from the occultation are shown in Figure 2, which shows the occulted (I) spectrum of ε Orionis divided by the unocculted spectrum (I0) for the average of the 1 s time records 242 to 246. During this 5 s period the star crossed from an altitude of 32 to 22 km. Using the occulted time records 242–246, the overall integrated column density for the plume was computed. The best fit water column density was determined to be 1.5 × 1016 molecules per square centimeter. We performed the same calculation as for previous occultations (described in Hansen et al., 2006): n (the column density) and Y are computed from UVIS data, and Vth is the thermal velocity of the gas assuming a HANSEN ET AL. ENCELADUS’ WATER VAPOR PLUME VARIABILITY 2 Geophysical Research Letters 10.1002/2016GL071853 Figure 2. The occulted spectrum (I) divided by the unocculted spectrum (I0), labeled I/Inot, of ε Orionis shows the distinctive deep absorption bands of water vapor. The spectrum is from a summation of all spatial channels, smoothed by 3, for the average of time records 242 to 246, thus representative of the bulk gas in the plume. The smooth curve is a theo16 2 retical water spectrum computed using cross sections from Mota et al. [2005] with a column density of 1.5 × 10 cm . Units of wavelength are angstroms. temperature of 170 K [Spencer et al., 2011]. To estimate Y for this occultation since it did not traverse the entire plume, we relied on the three previous horizontal cuts through the plume for the bulk plume dimensions, shown in Figure 1c for 2007, 2010, and 2011. We assume that the gross physical dimensions have not changed. The results from all the occultations are listed in Table 1. While 250 kg/s is the highest flux value detected to date in a UVIS occultation, it is only slightly higher than in 2005, and is not higher by the expected factor of 1.5 to 2 times the average, as illustrated in Figure 3. (Given the uncertainties in the calculation of the mass flux, one could argue that UVIS has not seen any variability at all in the 2005 to 2016 time period. The most significant simplification and source of uncertainty is that the temperature from which Vth is calculated does not change.) Brightness variability of the plume on a decadal timescale has been detected in Cassini images acquired from 2005 to 2015 [Ingersoll and Ewald, 2017], decreasing overall by a factor of ~2. Since UVIS has observed just a small number of occultations over the decade, it is difficult to separate mean anomaly from decadal variability, except at a mean anomaly of 236, where we observed occultations in 2007 and 2011. As shown in Table 1, the vapor mass flux in those years was very similar. Ingersoll and Ewald [2017] hypothesize that the trends are due to the decadal variability of Enceladus’ orbital eccentricity. Interestingly, our results are consistent with this hypothesis, referring to Figure 17 of Ingersoll and Ewald [2017], illustrating that the eccentricity in 2007 and 2011 was similar. The occultations in 2005 and 2010 show a larger difference in vapor mass flux, also qualitatively consistent with the wide gap in eccentricity illustrated in Figure 17 of Ingersoll and Ewald [2017]. 4. Supersonic Gas Jets and Flow Direction The character of the occulted signal as the ray passed through the Baghdad I jet (Figure 1b) showed distinct differences relative to previous occultations. In a manner similar to that illustrated in Figure 2 the occulted (I) a Table 1. Water Vapor Flux for the Bulk Plume Computed From Column Density for Occultations Observed in 2005, 2007, 2010, 2011, and 2016 Year n (× 10 2005 2007 2010 2011-e 2011-z 2016 a 16 cm 1.5 1.4 0.9 1.35 1.2 1.5 2 ) 16 5 27 Uncertainty (+/ , × 10 ) Y (× 10 cm) Vth (cm/s) H2O (molecules/s, × 10 ) H2O (kg/s) Mean Anomaly 0.15 0.14 0.23 0.15 0.2 0.15 120 110 150 120 135 125 45,000 45,000 45,000 45,000 45,000 45,000 8.1 6.9 6.0 7.3 7.3 8.4 240 210 180 220 220 250 117 236 98 237 237 208 Note that the data acquired in 2011 were from a simultaneous occultation of ε and ζ Orionis. HANSEN ET AL. ENCELADUS’ WATER VAPOR PLUME VARIABILITY 3 Geophysical Research Letters 10.1002/2016GL071853 Figure 3. UVIS occultation data have been acquired at four orbital longitudes, shown with colored boxes, compared to VIMS plume brightness data [Hedman et al., 2013, base plot kindly provided by M. Hedman]. Green boxes are UVIS data from 2005, 2007, 2010, and 2011, tabulated in Table 1. The red box shows the new 2016 results. spectrum of ε Orionis was divided by the unocculted spectrum (I0) for time record 239. This 1 s integration was acquired as the ray to the star pierced the Baghdad I jet at an altitude of ~40 km. Water vapor with a column density of 2.0 × 1016 molecules per square centimeter is the best fit. In 2011, ζ Orionis was occulted by this jet at a similar altitude (~40 km), and we observed a column density of 1.6 × 1016 molecules per square centimeter. Comparing 2011 to 2016, there were 25% more molecules in the line of sight to the star in the jet in 2016. The portional amount of gas in this jet relative to the combined contribution from the rest of the tiger stripes is also larger, with the Baghdad I gas jet contributing ~8% of the total molecules seen, compared to other orbital longitudes when its contribution is just 2%. Thus, the supersonic jet contribution is 4 times larger. Since we do not find the predicted increase in the total number of water molecules in the bulk plume, we infer that the primary difference that occurs near apokrone is the relative contribution between the plume and the jets, with the jets that lift the observed ice grains much stronger at this orbital longitude. The Baghdad I jet was 86 km wide at ~40 km altitude, from which we compute a Mach number of 1.6 +/ 0.28. This compares to a previous estimate of Mach 2.3 from the 2011 occultation. Thus, in 2016, we see a larger number of molecules moving at a lower speed in the Baghdad I jet compared to previous occultations. This is similar to the VIMS results, which showed a lower speed at apokrone compared to perikrone for the particles [Hedman et al., 2013]. Another interesting feature is that the amount of gas close to the surface between the tiger stripes is less than any other occultation. Figure 4 shows the extinction profile for the 2005 γ Orionis occultation compared to this 2016 ε Orionis occultation. The HSP data are plotted as a function of altitude. Both occultations made a diagonal cut across the plume and intercepted the surface (other occultations have been horizontal cuts across the plume without the star being occulted by Enceladus). Figure 4. Two extinction profiles from the HSP are shown, for 2005 and 2016. They look similar at 20–30 km altitude; however, the signal continues to decrease as the line of sight approaches the surface in 2005, while in 2016, the signal almost returns to its unocculted value between the fissures. HANSEN ET AL. ENCELADUS’ WATER VAPOR PLUME VARIABILITY At 20–30 km altitude the extinction is similar in both occultations. However, as the star neared the surface in the 2016 occultation, it returned to almost 100% transmission, in contrast to the optical depth in 2005, which continued to increase as the star approached the surface. The 2016 profile is consistent with the interpretation that there are supersonic jets, not just a curtain of gas, coming from the fissure. A comparison of the profiles suggests that, in general, in 2016, molecules 4 Geophysical Research Letters 10.1002/2016GL071853 streaming from the tiger stripes were exiting Enceladus’ tiger stripes with more vertical trajectories. 5. Summary The higher-column density of gas in the Baghdad I jet allows the jet to loft more and/or larger particles at orbital longitude 208° compared to when Enceladus is closer to perikrone. The larger contribution of the jets to the plume at longitude 208° is significant but does not result in a substantially higher overall column density for the plume, so these two results are consistent. Tidal energy had been hypothesized as the driver for Enceladus’ eruptions early on [Hurford et al., 2007], but it was not until the VIMS results were published [Hedman et al., 2013] that there was indisputable evidence for tidal forces. However, the variation in brightness of the plume correlated with Enceladus’ position in its orbit does not answer the question of the actual physical mechanism. Are the fissures simply widening and narrowing in response to changes in the stress field? UVIS results describing gas flux, jets, and general structure of the plume, the observables above the surface, have been key to understanding what is going on below Enceladus’ surface. The case has been made that channels connecting the subsurface body of liquid water to the surface act as nozzles when their width varies, accelerating the gas to supersonic speeds and locally enhancing condensation [Schmidt et al., 2008]. The supersonic jets may be responsible for the size and composition stratification observed, by lofting the smallest pure ice particles to escape velocity to the E ring, while the larger saltier particles fall back to the surface [Postberg et al., 2009, 2011; Hansen et al., 2011]. The new UVIS data are likewise a means for testing hypotheses of what is happening below Enceladus’ surface in response to tidal forces. For example, a recent work by Kite and Rubin [2016] describes water-filled slots at the base of the tiger stripe fissures that are sustained against freezing by tidally modulated turbulent flow of the water in the slots. Diurnal flexing of the slots modulates the vapor outflow, but it is never stopped completely, consistent with UVIS observations of gas mass flux persisting over an orbit. This model may need to be modified to agree with our new result that the jets are the predominant source of the particle flux variability seen by VIMS. Perhaps the nozzle width varies along a tiger stripe. Then the major contribution would be where the fracture width matches the resonant condition stated by Kite and Rubin [2016] (about 3 m), which may correlate with the individual jet sources. If tidal stresses during the orbit vary the fracture width, these would also be the locations of maximum variability. The jet locations are where the flux is thus high enough to loft the small dust grains seen by VIMS. UVIS will observe one more stellar occultation by the plume in 2017 at a mean anomaly of 113°. We will compare it to the occultation in 2005 to further test the sensitivity of output to the eccentricity of Enceladus’ orbit. Acknowledgments This research was supported by a grant from the Cassini Project. Data will be delivered to the Planetary Data System (PDS) in January 2017. The authors thank the Cassini Project for the navigational work required to tweak the trajectory to optimize the occultation geometry. HANSEN ET AL. References Dougherty, M. K., K. K. Khurana, F. M. Neubauer, C. T. Russell, J. Saur, J. S. Leisner, and M. E. Burton (2006), Identification of a dynamic atmosphere at Enceladus with the Cassini magnetometer, Science, 311, 1406–1409. Esposito, L. W., et al. (2004), The Cassini ultraviolet imaging spectrograph investigation, Space Sci. Rev., 115, 299–361. Hansen, C. J., L. Esposito, A. I. F. Stewart, J. Colwell, A. Hendrix, W. Pryor, D. Shemansky, and R. West (2006), Enceladus’ water vapor plume, Science, 311, 1423–1425. Hansen, C. J., L. W. Esposito, A. I. F. Stewart, B. Meinke, B. Wallis, J. E. Colwell, A. R. Hendrix, K. Larsen, W. Pryor, and F. Tian (2008), Water vapour jets inside the plume of gas leaving Enceladus, Nature, 456, 477–479. Hansen, C. J., et al. (2011), The composition and structure of the Enceladus plume, Geophys. Res. Lett., 38, L11202, doi:10.1029/2011GL047415. Hedman, M. M., C. M. Gosmeyer, P. D. Nicholson, C. Sotin, R. H. Brown, R. N. Clark, K. H. Baines, B. J. Buratti, and M. R. Showalter (2013), An observed correlation between plume activity and tidal stresses on Enceladus, Nature, 500, 182–184. Hurford, T. A., P. Helfenstein, B. V. Hoppa, R. Greenberg, and B. Bills (2007), Eruptions arising from tidally controlled periodic openings of rifts on Enceladus, Nature, 447, 292–294. Ingersoll, A. P., and S. P. Ewald (2017), Decadal timescale variability of the Enceladus plumes inferred from Cassini images, Icarus, 282, 260–275. Kite, E. S., and A. M. Rubin (2016), Sustained eruptions on Enceladus explained by turbulent dissipation in tiger stripes, Proc. Natl. Acad. Sci. U.S.A., 113, 3972–3975. Mota, R., et al. (2005), Water VUV electronic state spectroscopy by synchrotron radiation, Chem. Phys. Lett., 416, 152–159. Nimmo, F., C. Porco, and C. Mitchell (2014), Tidally modulated eruptions on Enceladus: Cassini ISS observations and models, Astron. J., 148(46), 14. Porco, C., D. DiNino, and F. Nimmo (2014), How the geysers, tidal stresses, and thermal emission across the south polar terrain of Enceladus are related, Astron. J., 148(45), 24. Porco, C. C., et al. (2006), Cassini observes the active south pole of Enceladus, Science, 311, 1393–1401. ENCELADUS’ WATER VAPOR PLUME VARIABILITY 5 Geophysical Research Letters 10.1002/2016GL071853 Portyankina, G., L. W. Esposito, A. Ali, and C. J. Hansen (2016), Modeling of the Enceladus water vapor jets for interpreting UVIS star and solar occultation observations, Lunar Planet, Sci. Conf. 47, 2600. Postberg, F., S. Kempf, J. Schmidt, N. Brilliantov, A. Beinsen, B. Abel, U. Buck, and R. Srama (2009), Sodium salts in E-ring ice grains from an ocean below the surface of Enceladus, Nature, 459, 1098–1101. Postberg, F., J. Schmidt, J. Hillier, S. Kempf, and R. Srama (2011), A salt water reservoir as the source of a compositionally stratified plume on Enceladus, Nature, 474, 620–622. Schmidt, J., N. Brilliantov, F. Spahn, and S. Kempf (2008), Slow dust in Enceladus’ plume from condensation and wall collisions in tiger stripe fractures, Nature, 451, 685–688. Spahn, F., et al. (2006), Cassini dust measurements at Enceladus and implications for the origin of the E Ring, Science, 311, 1416–1418. Spencer, J., and F. Nimmo (2013), Enceladus: An active ice world in the Saturn system, Ann. Rev. Earth Planet. Sci., 41, 695–717. Spencer, J. R., C. J. A. Howett, A. J. Verbiscer, T. A. Hurford, M. E. Segura, J. C. Pearl (2011), Observations of thermal emission from the south pole of Enceladus in August 2010, EPSC-DPS 2011–1630. Spencer, J. R., J. C. Pearl, M. Segura, F. M. Flasar, A. Mamoutkine, P. Romani, B. J. Buratti, A. R. Hendrix, L. J. Spilker, and R. M. C. Lopes (2006), Cassini encounters Enceladus: Background and the discovery of a south polar hot spot, Science, 311, 1401–1405. Spitale, J. N., and C. C. Porco (2007), Association of the jets of Enceladus with the warmest regions on its south-polar fractures, Nature, 449, 695–697. Waite, J. H., et al. (2006), Cassini ion and neutral mass spectrometer: Enceladus plume composition and structure, Science, 311, 1419–1422. HANSEN ET AL. ENCELADUS’ WATER VAPOR PLUME VARIABILITY 6