TEMPORAL AND SPATIAL ASPECTS OF GAS RELEASE DURING THE 2010 APPARITION OF COMET 103P/HARTLEY 2^*

Prior to its apparition in 2010, the ecliptic comet 103P/Hartley 2 (hereafter Hartley 2) was distinguished among short period comets of Kuiper Belt origin by three factors: first, its nucleus exhibited unusually high fractional activity (approaching 100% of the sunlit area; Groussin et al. 2004; Lisse et al. 2009), second, it released an unusually high (as then perceived) abundance of CO₂ (∼10%–20% relative to H₂O; Colangeli et al. 1999; Crovisier et al. 1999), and third, it displayed apparently 'normal' (or 'typical') composition among a dynamical population known for having many members depleted in (the unknown) 'long-chain' hydrocarbons whose destruction could lead to the C₂ and C₃ radicals observed in the coma (A'Hearn et al. 1995).

It is now recognized that the first two factors are connected, through CO₂-controlled nucleus activity that releases icy clumps into the coma; their subsequent dissolution leads to an overestimate of the fractional activity of the nucleus (A'Hearn et al. 2011). The classification as ''typical'' is based on the relative production rates of reactive species (C₂, CN, C₃), which cannot be stored stably in the cometary nucleus and whose parentage is disputed or unknown. While product species are interesting, the identity and abundance of the primary volatiles and rocky grains released from the nucleus are the key clues for testing the origin and evolution of cometary material. Volatiles released directly from the nucleus (whether as gases or ices) are here termed "primary" species, while those produced in the coma are termed ''product'' species.

Production rates of primary volatiles can change dramatically as orbital motion and nucleus rotation modulate the received insolation (and thus gas production) on long and short timescales, respectively. Moreover, the apparent mixing ratios in the coma may change if nucleus regions with distinct chemical compositions are exposed to sunlight sequentially, e.g., by nucleus rotation. And spatial differences in the release of specific volatiles can occur if individual active vents contain material of diverse volatile character. Thus, measurements acquired through snapshots in time can be placed in a global context for the nucleus only if such behavior is characterized. Unfortunately, taking such snapshots of the composition from one or two observations has long biased studies of primary volatiles in comets. To provide that contextual perspective for Hartley 2, our investigation emphasized both the long-term and short-term behavior of gas release from the cometary nucleus.

We acquired high-resolution near-infrared spectra of comet Hartley 2 using CRIRES at the ESO Very Large Telescope (VLT) in late July, and with NIRSPEC at the W. M. Keck Observatory from late July through mid-December 2010. The heliocentric (geocentric) distance (AU) ranged from 1.62 (0.78) on July 26 to 1.06 (0.13) on October 28.25 (perihelion), and then increased to 1.26 (0.37) on December 17. In late July, the comet was fainter than expected by about a factor of 10 so we achieved upper limits only. Although the comet remained less active than expected throughout this apparition (Combi et al. 2011; Meech et al. 2011), we successfully characterized volatile production with NIRSPEC on September 18 (Mumma et al. 2010), October 19 and 22, November 16, and December 16 and 17.

NIRSPEC spectra were acquired by nodding the telescope along the slit in an ABBA sequence, with the two beams separated by 12 arcsec (half the slit length). At each grating setting, spectra of infrared standard stars and calibration frames (flats and darks) were acquired for absolute flux calibration of the cometary data. We followed our standard methodology for data reduction and analysis of the individual echelle orders (Bonev 2005; DiSanti et al. 2001; Mumma et al. 2001). The signals from the two beams were combined, and spectra were extracted over nine rows (1.78 arcsec) centered on the nucleus (taken to be the row containing the peak continuum). We isolated cometary molecular emissions by subtracting a modeled dust continuum multiplied by the atmospheric transmittance. We synthesized the spectral transmittance by using a multiple layer atmosphere and a radiative transfer model (LBLRTM) that accessed the HITRAN-2008 molecular database augmented with our custom updated line parameters and fluorescence models (e.g., Villanueva et al. 2008; Kawakita & Mumma 2011; Radeva et al. 2011; Villanueva et al. 2011).

Typical spectra are presented in Figure 1. From such spectra, we quantified H₂O, CH₄, C₂H₂, C₂H₆, H₂CO, CH₃OH, HCN, NH₃, NH₂, and OH (prompt emission) on multiple dates. A robust rotational temperature is derived from the observed water line intensities (Table 1), and the derived ortho-para abundance ratio (OPR) is 2.85 ± 0.20. The OPR standard deviation (0.20, 1σ) reflects the combined effects of measurement error and modeling uncertainty for intensities of ortho and para lines. The upper bound of the OPR (3.05) is consistent with statistical equilibrium (spin temperature, T_spin ≃ 55 K); the lower bound (2.65) defines T_spin > 32 K. Infrared Space Observatory observations in 1997 returned OPR values of 2.76 ± 0.08 and 2.63 ± 0.18, and T_spin of 34 ± 3 K (Crovisier et al. 1999). For UT November 4, 2011, Dello Russo et al. (2011) reported OPR = 3.4 ± 0.3.

Zoom In Zoom Out Reset image size

The long-term evolution of water production obtained from our data is compared with that derived from optical magnitudes in Figure 2(a), and its short-term behavior is presented in Figure 2(b). We estimated long-term production rates of water from optical magnitudes, following Jorda et al. (2008): log[Q(H₂O)] = log(k) + 30.675 – 0.2453 m_h, where k was derived to be 0.26 and m_h is the visual magnitude¹⁰ (m_h = 8.7 + 20 log r_h(t₀ – 20)). Our mean water production rates are in good agreement with those reported by Combi et al. (2011).

Zoom In Zoom Out Reset image size

The trends in mixing ratio (C₂H₆, HCN, CH₃OH) relative to water are shown in Figure 2(c). Detailed parameters for seven primary volatiles and NH₂ (instrument settings, rotational temperatures, production rates, and mixing ratios) on October 19 and 22 are given in Table 1 (for CH₄, see Table 2, footnote d). A summary of water production rates and mixing ratios for all dates is presented in Table 2. We targeted CO in mid-September and mid-December, but poor weather restricted our observations and CO was not attempted. In early November, CO was measured with Hubble Space Telescope (HST) and found to be within the range 0.15%–0.45% (Weaver et al. 2010, 2011).

The spatial profiles measured along slit are shown for H₂O, C₂H₆, and dust on four dates, and for HCN and CH₃OH on two (Figure 3). The ends of the slit are marked (+, −) (see the inset diagram in each panel), and corresponding pixels are so-marked on the abscissa (panels (a)–(d)). The ''+'' direction corresponds to the projection (on the sky plane) of the true comet–Sun direction, but pixels in the ''+'' direction sample material in both sunward and anti-sunward hemispheres (compare diagrams above each panel). For isotropic outflow, about ¾ of the sampled coma gas would be in the sunward hemisphere on UT September 18, however, no exact conclusion can be reached when outflow is anisotropic. We loosely term the ''+" direction as ''sampling the sunward hemisphere'', and the ''–" direction as ''sampling the anti-solar hemisphere".

Zoom In Zoom Out Reset image size

On UT 18 September, we aligned the slit nearly along the Sun–comet line (within ∼6 deg). The spatial profile of the continuum is symmetric, but H₂O is extended in the anti-solar hemisphere (''–" direction) while C₂H₆ is extended in the sunward hemisphere (''+" direction; Figure 3(a)).

On UT October 19, a strong CN jet was reported with position angle nearly due north from the optocenter (Knight & Schleicher 2010, 2011). We aligned the slit along the jet, and nearly orthogonal to the Sun–comet line (Figure 3(b)). The spatial profiles of continuum, H₂O, and CH₃OH are symmetric but gases are more extended than dust in both directions. C₂H₆ is more extended in the jet direction (''+") than are other gases and dust, and HCN is more extended than C₂H₆ in the anti-jet direction (but less so than H₂O and CH₃OH).

On UT October 22, we aligned the slit nearly along the Sun–comet line (orthogonal to the CN jet; Figure 3(c)). All gases are more extended in the sunward hemisphere (''+'') than is dust and their profiles have similar shape. Water is asymmetric and more extended in the anti-solar hemisphere (''–''), compared with CH₃OH, HCN, C₂H_6, and dust.

On UT December 16, we aligned the slit ∼40 deg to the Sun–comet line. Ethane is more extended in the sunward hemisphere (''+'') than is H₂O or dust; H₂O is more extended than dust and ethane in the anti-solar hemisphere (''–''; Figure 3(d)).

The activity of Hartley 2 varied strongly with time. The short-term behavior is especially significant—the global production rates for water (the dominant volatile) changed by approximately a factor of two from October 19 to 22, and again from December 16 to 17 (Figure 2(b), Table 1). The production rates for water, ethane, HCN, and methanol vary in a manner consistent with nucleus rotation, similar to the behavior reported for the optical brightness of gas and dust in the coma and for radar images (Harmon et al. 2010; Harmon et al. 2011), CN jet activity (Jehin et al. 2010), activity in HCN, and other phenomena linked directly to the nucleus (Meech et al. 2011). N. Biver et al. (2010, private communication) reported variations in HCN with amplitude ∼2.1 (Q_max/Q_min), with maxima observed around 25.10 and 27.24 UT and (broader) minima around October ∼25.4, 26.2, and ∼26.9. Compared with our measurements, smaller amplitude is expected for the larger beam sizes used by Biver et al., because the time since release increases linearly with nucleocentric distance (assuming constant outflow velocity).

The mixing ratios of trace gases measured by us (the ratios of production rates) are steady throughout the apparition (Figure 2(c)), demonstrating that Hartley 2 releases material that is rather homogeneous in the bulk average. Spacecraft images of the nucleus revealed many individual vents and jets (A'Hearn et al. 2011). Production rates from individual vents may turn on and off as they rotate into and out of sunlight, but our data show little change—if any—in the mean chemical composition of released material. CO₂ comprised about 20% of the primary volatiles detected by the on board spectrometer, and its sublimation is thought to control the release of ''clumps'' of water ice detected in images (A'Hearn et al. 2011). If so, this apolar ice (CO₂) must be largely segregated from the polar water ice in the active nucleus regions. Moreover, ethane and water show distinct spatial profiles (e.g., Figure 3(b)) but the ethane mixing ratio (C₂H₆/H₂O) changes little as the nucleus rotates, suggesting that they are segregated into two distinct ice phases as well. If ethane and carbon dioxide reside in a common apolar ice phase, carbon dioxide should show a similar steady mixing ratio (relative to co-released water) with changing rotational phase. But if it does not, then two distinct apolar ice moieties are implied along with a third polar ice. Future spacecraft results (of CO₂/H₂O versus rotational phase) may test this hypothesis.

The mean mixing ratio for HCN in Hartley 2 agrees well with those measured in several Oort cloud comets tentatively called "organics-normal" based on primary volatiles in the relatively small number (∼20) of comets measured in the IR. However, the mixing ratios for ethane and methanol are slightly higher (and the mixing ratios of C₂H₂ and CO are much lower) in Hartley 2 than in this hypothesized "organics normal" group (Table 2; Mumma et al. 2003; DiSanti & Mumma 2008). CO₂ (currently estimated at ∼20%) is at the high end of the range found by the Akari space observatory for comets within 2.5 AU of the Sun (H. Kawakita 2011, private communication; Ootsubo et al. 2010). The high abundance ratio of CO₂/H₂O is not expected if hydrogenation of C₂H₂ and CO is responsible for the high ethane/acetylene and CH₃OH/CO ratios in Hartley 2 (Mumma et al. 1996). A distinct additional mechanism is needed for efficient formation of CO₂, perhaps by quantum-mechanical tunneling (of CO and O(³P)) on CO-enriched pre-cometary grains or by prompt recombination of CO with O(¹D) produced by radiation processing of icy grain mantles.

3.1. Interpreting Differences in Spatial Profiles

3.2. Constancy of Mixing Ratios

Keck telescope time was granted by NOAO (through the Telescope System Instrumentation Program funded by NSF), the University of Hawaii, and the California Institute of Technology. VLT time was granted by the European Southern Observatory. We gratefully acknowledge support by the NSF Astronomy and Astrophysics Research Grants Program (PI/co-PI Bonev/Gibb), by the NASA Astrobiology Institute (PI: Meech, PI: Mumma), by NASA's Planetary Astronomy (PI: DiSanti; PI: Mumma), Planetary Atmospheres (PI: DiSanti; PI: Villanueva), and Discovery (Meech) Programs, and by the German–Israel Foundation (D. Prialnik; M. Lippi).