Modeling the Enceladus plume–plasma interaction

Authors


Abstract

[1] We investigate the chemical interaction between Saturn's corotating plasma and Enceladus' volcanic plumes. We evolve a parcel of ambient plasma as it passes through a prescribed H2O plume using a physical chemistry model adapted for water-group reactions. The flow field is assumed to be that of a plasma around an electrically-conducting obstacle centered on Enceladus and aligned with Saturn's magnetic field, consistent with Cassini magnetometer data. We explore the effects on the physical chemistry due to: (1) a small population of hot electrons, (2) a plasma flow decelerated in response to the pickup of fresh ions, and (3) the source rate of neutral H2O. The model confirms that charge exchange dominates the local chemistry and that H3O+ dominates the water-group composition downstream of the Enceladus plumes. We also find that the amount of fresh pickup ions depends heavily on both the neutral source strength and on the presence of a persistent population of hot electrons.

1. Introduction

[2] Early Cassini encounters with Enceladus revealed surprising evidence of a significant source of water (with trace percentages of other neutrals, including CO2) from geysers located at the moon's southern pole [Hansen et al., 2006; Porco et al., 2006; Spencer et al., 2006; Waite et al., 2006]. The H2O cloud reacts with Saturn's corotating plasma torus, loading Saturn's magnetosphere with fresh ions. The pickup rate equation image quantifies the amount of fresh ions added to the magnetosphere from charge exchange and impact/photoionization. Respective contributions to pickup from charge exchange and impact/photoionization are fundamentally different in that charge exchange does not contribute to ion production because one ion replaces another. Both processes however introduce slow-moving ions which must subsequently be accelerated by Saturn's magnetosphere.

[3] Early identification of the interaction between the local water source and Saturn's corotating plasma was made by Dougherty et al., 2006]. Based on the Cassini Plasma Spectrometer (CAPS [Young et al., 2004]) analysis by Tokar et al. [2006], Pontius and Hill [2006] modeled the interaction and derived a pickup rate of equation image ≈ 100 kg s−1. Khurana et al. [2007] (hereafter K07) and Saur et al. [2008] (hereafter S08) discovered a range in equation image (0.2–3 kg s−1) from Cassini magnetometer data in the three earliest Cassini Enceladus flybys (E0, 17 February 2005; E1, 09 March 2005; E2, 14 July 2005). Constrained by Ion Neutral Mass Spectrometer (INMS [Waite et al., 2006]) and Ultraviolet Imaging Spectrograph (UVIS [Hansen et al., 2006]) observations, Burger et al. [2007] estimated a pickup rate of equation image ≈ 2–3 kg s−1 from a neutral cloud model. The large discrepancy between the pickup rates derived from CAPS and magnetometer data is due not only to the fact that the region considered by Pontius and Hill [2006] is much larger than that considered by K07 and S08, but also because the Pontius and Hill [2006] result depends on the poorly-constrained value of Saturn's Pederson conductivity.

[4] In this paper, we use a physical chemistry model to investigate the chemical interaction between the corotating plasma and the Enceladus plumes. Charge exchange dominates the local chemistry and leads to an H3O+-dominated plasma downstream of Enceladus. We find that pickup increases when hot electrons are present—more so with a high neutral source rate.

2. Model

[5] We use a physical chemistry model developed to study the Enceladus torus [Delamere and Bagenal, 2003; Delamere et al., 2007; Fleshman et al., 2009] to investigate the composition of plasma traveling along prescribed flow lines. The model evaluates mass and energy rate equations for water-group ions (W+ ≡ O+ + OH+ + H2O+ + H3O+), protons, and thermal electrons in a parcel of plasma transiting the simulation. Neutrals are assumed to be cold, and in this study neutral abundances are fixed. The full set of reactions includes charge exchange, photoionization, ionization by electron impact, radiative excitation, recombination, and molecular dissociation by both electron impact and recombination. All species have isotropic Maxwellian speed distributions, and energy is transferred between species via Coulomb collisions. The simulation spans a rectangular domain extending 5 RE from Enceladus in all directions except south, where the simulation extends to 15 RE (RE = 252 km is the radius of Enceladus).

[6] A second population of supra-thermal ‘hot’ electrons is imposed with a fixed density (0.3 cm−3) and temperature (160 eV). Hot electrons near Enceladus have been reported by Tokar et al. [2009] and have been observed throughout the torus by CAPS and the Radio and Plasma Wave Science Instrument [Moncuquet et al., 2005; Young et al., 2005]. Fleshman et al. [2009] showed that a small amount of hot electrons is necessary to obtain the ambient ionization. Here we investigate the importance of hot electrons near Enceladus itself.

2.1. Neutral Source

[7] Following S08, the plume is prescribed as

equation image

where Hθ = 12° and Hd = 948 km (4 × the Hill radius). S08 offset the plume from Enceladus' southern pole by 8° and considered more than one source with the form of (1). We consider a single source whose origin coincides with Enceladus' south pole. In the nominal case, n0 is set to 2.5 × 109 cm−3, corresponding to a neutral source rate of ≈200 kg s−1 (S08). S08 found a much stronger source for E0, so we also investigate the implications of a source with n0 = 2.2 × 1010 cm−3, corresponding to a neutral source rate of ≈1600 kg s−1.

2.2. Plasma Flow Field

[8] Because of the low Alfvén Mach number at Enceladus (MA ≈ 0.1 [Sittler et al., 2008]), perturbations travel rapidly along the magnetic field so that the source region presents a cylindrical obstacle to the corotating plasma. We adopt the flow field used by Dols et al. [2008] to study the plasma interaction with Jupiter's moon Io:

equation image

where vamb ≈ 0.8 × vcor (vcor ≈ 26 km s−1) is the ambient plasma speed far from Enceladus [Wilson et al., 2009]. The magnetic field defines the z-axis and ϕ is measured from the flow direction.

[9] Along each flow line, parcels of plasma were started 5 RE upstream of Enceladus with the steady-state composition given by Fleshman et al. [2009]. The plasma was moved in the direction of the plasma flow to 5 RE downstream, and the chemistry was updated at the associated time steps. The pickup energy—determined by the relative speed between neutrals and plasma flow—was also updated. Ions far from Enceladus are picked up at vamb, while those near Enceladus are picked up more slowly up- and downstream and more rapidly on the flanks.

[10] To investigate the effect of ion pickup, we slowed the flow near the obstacle. We followed Dols et al. [2008] by decreasing the component of the plasma velocity in the flow direction: ux was replaced by γ(ρ)ux, where γ(1 RE) = 0.5, increasing linearly to γ(2 RE) = 1. We find however that a stronger slowing factor (γ(1 RE) = 0.1) does not qualitatively change our results. In section 3, we compare the pickup rates both for when the plasma has (nominal case) and has not been slowed.

[11] Two effects are due directly to the slowing of the flow. First, the pickup energy is reduced, affecting the plasma temperature because fresh ions are picked up at the local plasma speed. Second, impact ionization increases because plasma transits more slowly. Impact ionization contributes directly to pickup, as well as indirectly, by seeding multiple charge exchanges.

[12] We neglect gyromotion on the basis of scale. For example, an H3O+ ion picked up at vamb has a gyroradius of only 0.1 RE. More important is that (in the frame of Enceladus) ions oscillate between zero and twice their pickup speed and we ignore the velocity dependence of charge exchange. Including gyromotion would enhance the effects we report in this paper.

2.3. Hot Electrons

[13] We estimate that hot electrons cool rapidly near the dense plume via impact ionization, and thus imposed a discontinuity at ρ = 3 RE:

equation image

[14] However, Dols et al. [2008] showed that field-aligned electron beams (perhaps associated with the Io auroral footprint) are necessary to model the high plasma density in Io's wake. At Enceladus, hot electrons may also be related to weak UV auroral spots recently observed by UVIS (W. Pryor, personal communication, 2009). To investigate the implication of hot electrons at Enceladus, we consider the additional case where neh is held at 0.3 cm−3 throughout the simulation. The pickup rates for each case are compared in section 3.

2.4. Pickup Rate Calculation

[15] Fresh ions are added to the magnetosphere by both charge exchange (equation imageexch) and impact/photoionization (equation imageioz):

equation image
equation image

The reaction rates [Fleshman et al., 2009] are represented by κimp, κexch [cm3 s−1], and κphot [s−1]; the ion masses by m; and n(1), n(2) are the charge-exchanging neutral and ion densities. Summations are carried out over processes involving the creation of fresh ions.

[16] We calculated the time-averages of (4) and (5) to find the average pickup rates for plasma parcels migrating along each flow line and multiplied by the flow line volume to find the pickup rate for each flow line. The total pickup rate for each process was determined by summing the contribution from all flow lines throughout the simulation.

3. Results

[17] We consider the xy plane 7.5 RE south of the center of Enceladus. For this, the nominal case, hot electrons exist throughout the domain, the neutral source rate is ≈200 kg s−1, and the plasma is slowed in the flow direction (section 2). The pickup energy, heavy-to-light ion abundance (W+/H+), H3O+/W+, and H3O+ temperature (equation image) in this plane are shown in Figure 1. The obstacle is plotted in black and 18 flow lines are over-plotted and labeled. Composition and temperatures along this line are plotted in Figure 2, where the corresponding flow lines are indicated.

Figure 1.

(a) The x component of the assumed local plasma flow speed (section 2) normalized to the ambient flow speed (80% of rigid corotation). (b) Water-group to proton abundance ratio. (c) H3O+/W+ abundance ratio. (d) H3O+ temperature. The plane represented here is 7.5 RE south of Enceladus. Model output along the dashed line is given in Figure 2.

Figure 2.

(a) Abundances and electron density from the simulation along the dashed line in Figure 1. (b) Electron, proton, and H3O+ temperatures along the same cut. The electron temperature is normalized to its ambient temperature (2 eV), and the ion temperatures are normalized to their ambient pickup energies—1.5 and 29 eV for H+ and H3O+, respectively.

[18] The H3O+ ion is the most abundant ion in the wake—a consequence explained by the importance of charge exchange and the fact that charge exchanges lead to H3O+ in the presence of an abundant water source. All charge exchanges with H2O in our model ultimately lead to either H3O+ or H2O+ by H2O+ + H2O → H2O + H2O+ and H2O+ + H2O → OH + H3O+. The former reaction supports H2O+ density somewhat, but the H2O+ products also feed into the latter, producing H3O+. The W+/H+ ratio increases rapidly because protons are removed by H+ + H2O → H + H2O+. The increase in electron density is due mainly to impact ionization of the plume by hot electrons.

[19] In Figure 2, the electron temperature has been normalized to the ambient electron temperature (2 eV), and the ion temperatures have been normalized to their respective ambient pickup energies (1.5 and 29 eV). Protons and H3O+ bear the signature of the cooler pickup temperature from where they were created by charge exchange. A factor of a few decrease in ion temperature through Enceladus' wake has recently been observed by Tokar et al. [2009]. The electron temperature has also been cooled by a factor of 2.

[20] We calculated individually (section 2) the pickup rates equation imageexch, and equation imageioz–the ratio of which illustrates the importance of charge exchange over impact ionization. The simulation was run for the eight cases shown in Table 1. The flow-field, hot-electron, and source-rate treatments are described in section 2. We discuss the cases corresponding to a ‘weak’ 200 kg s−1 neutral source (Cases 1a–4a) and a ‘strong’ 1600 kg s−1 neutral source (Cases 1b–4b) separately.

Table 1. Pickup Rates From Charge Exchange and Impact/Photoionization for the Eight Cases Discussed in the Texta
 Caseequation imageexchequation imageiozequation imageexch/equation imageioz
kg s−1(1025 H2O s−1)kg s−1(1025 H2O s−1)
  • a

    Charge exchange is equation imageexch, and impact/photoionization is equation imageioz. Only equation imageioz increases the plasma density (ne). The cases labeled with ‘a’ correspond to a neutral source rate of 200 kg s−1; those labeled with ‘b’ correspond to a neutral source rate of 1600 kg s−1. Case (1a), in bold, is the nominal case from which Figures 1 and 2 have been generated.

(1a)hot electrons + slowed flow0.25(0.84)0.038(0.13)6.7
(2a)hot electrons + un-slowed flow0.22(0.74)0.034(0.11)6.6
(3a)no hot electrons + slowed flow0.18(0.61)0.0091(0.030)20
(4a)no hot electrons + un-slowed flow0.17(0.58)0.0087(0.029)20
(1b)hot electrons + slowed flow2.2(7.3)0.33(1.1)7.0
(2b)hot electrons + un-slowed flow1.9(6.5)0.30(0.99)6.6
(3b)no hot electrons + slowed flow0.77(2.6)0.079(0.26)9.8
(4b)no hot electrons + un-slowed flow0.77(2.6)0.076(0.25)10

3.1. Weak Source (Cases 1a–4a)

[21] When hot electrons exist locally (1a/2a), the total pickup is roughly 0.3 kg s−1 with a 40% increase when the plasma is slowed. Because of the longer occupation time associated with the slowed flow, hot electrons increase seed ionization (equation imageioz), and in turn increase charge exchange (equation imageexch). When hot electrons are removed (3a/4a), equation imageioz decreases by a factor of 4, but equation imageexch drops by ≈30%, implying that much of the pickup is occurring outside the cut-off point at 3 RE. Slowing the plasma has less effect in 3a than in 1a because longer occupation does not boost seed ionization without hot electrons.

3.2. Strong Source (Cases 1b–4b)

[22] The effect of the hot electrons becomes more apparent with the strong source because a denser portion of the plume is intersected by the flow lines. The effect of slowing the plasma is similar to that in the weak-source case. The total pickup however is increased by a factor of 3 when comparing the cases with hot electrons (1b/2b) to cases without (3b/4b). The almost linear response of the total pickup to equation imageioz (compare 1b/2b to 3b/4b) suggests that, unlike in the weak source case, most of the pickup is occurring inside the hot electron cutoff at 3 RE.

[23] K07 and S08 were in rough agreement on the total pickup rate. For E1 and E2, they found ≈0.2–0.6 kg s−1, and for E0 they found ≈3 kg s−1. Because our model relies on a physical chemistry calculation alone, it is remarkable to have obtained the same pickup rates using neutral plume distributions similar to those by S08 (E0: strong source, E1 and E2: weak source).

[24] Charge exchange dominates the chemistry by at least a factor of 6 in all cases. Because of this, the water-group abundance ratios (shown only for the nominal case) are qualitatively unaffected. In particular, H3O+/W+ always increases while O+/W+, OH+/W+, and H2O+/W+ always decrease in Enceladus' wake. The dominance of H3O+ elsewhere has been observed by CAPS during Cassini's orbital insertion period [Tokar et al., 2006; Sittler et al., 2008]. Fleshman et al. [2009] found that a steady-state, water-based Enceladus torus underestimates the H3O+ abundance seen in the CAPS data. Though few new ions are produced at Enceladus, the process that produces H3O+ may have an important effect on the large-scale torus composition. A more complete global model of the torus should include the effect of dense H2O on Saturn's corotating plasma demonstrated in this paper.

4. Discussion and Conclusions

[25] To investigate the impact of hot electrons on the chemistry, we have chosen the simplest flow field possible with minimal perturbation by Enceladus. This flow is roughly consistent with the compact source derived from magnetometer data (K07) but is much less perturbed than the flow reported by Tokar et al. [2006]. Similarly, we have started with a single symmetric plume oriented due-south. In future studies we will explore multiple jets (S08), displaced sources (K07), and a minor spherically-symmetric global component [Burger et al., 2007].

[26] Our findings are summarized as follows: (1) Charge exchange dominates the plume–plasma chemistry, confirming previous work by Burger et al. [2007] and consistent with estimates by Johnson et al. [2006]. (2) Charge exchange leads largely to an H3O+-dominated wake, consistent with INMS [Cravens et al., 2009]. Reactions leading to H3O+ are well known in the comet community [Aikin, 1974; Haberli et al., 1997]. (3) Comparing our pickup rates to those derived from the Cassini magnetometer (K07 and S08), INMS, and UVIS [Burger et al., 2007], we find that a persistent source of hot electrons may exist near Enceladus. If present, beams of hot electrons at Enceladus may be related to the weak UV auroral spots recently observed by W. Pryor (personal communication, 2009).

Acknowledgments

[27] This work was supported under NESSF grant Planet09F–0036 and CDAP grant 06-CASS06-0062.

Ancillary