DETERMINATION OF AN UPPER LIMIT FOR THE WATER OUTGASSING RATE OF MAIN-BELT COMET P/2012 T1 (PANSTARRS)^*

Recent observational and theoretical developments have suggested that the objects between Mars and Jupiter in the Main Asteroid Belt hold considerably more water ice than traditionally thought—as distinct from the widespread observation of hydrated minerals in asteroids (e.g., Rivkin et al. 2002; Vilas & Gaffey 1989). They would hold this water ice despite being in orbits that are presumably stable over the lifetime of the solar system and thus having surfaces that have received significant solar energy input. For example, the snow line at the time of planet formation could have been close to Mars' orbit (Sasselov & Lecar 2000; Lecar et al. 2006); subsurface ice in asteroids can survive for the age of the solar system if buried under a dusty surface (Schorghofer 2008); surface ice has possibly been detected on the large main-belt asteroid (24) Themis (Rivkin & Emery 2010; Campins et al. 2010).

Further evidence for significant water in the Main Belt comes from the so-called Main-Belt Comets (MBCs; Hsieh & Jewitt 2006; Bertini 2011; Jewitt 2012). These objects show cometary activity, and that cometary activity appears to be sustained rather than impulsive. The MBCs have orbits squarely within the Main Belt, and their orbits are found to be stable; as such, the objects are considered likely to be native to their current residence regions in the Main Belt. Thus, the implication is that volatile material incorporated into these bodies when they formed has survived to the present day. This also means that MBCs have the potential to place observational constraints on the temperature and compositional structure of the early solar system and our protoplanetary disk, giving insight into the solar system's formation.

In a recent study (Waszczak et al. 2013) of the Palomar Transient Factory survey database, upper limits on the possible population size of active MBCs were derived at a 2σ level of 22 active MBCs (per million main-belt asteroids down to 1 km diameter). We note that of the 10 known Main Belt objects observed to have shown extended emission from dust, we consider 7 of them to be MBCs (i.e., excluding P/2010 A2, P/2012 F5, and (596) Scheila), apparently having sustained activity due to sublimation of ices as the more traditional comets do. P/2012 T1, the object considered here, would be the seventh MBC.

The MBC P/2012 T1 (PANSTARRS) was discovered on October 6 by the 1.8 m Pan-STARRS1 (PS1) survey telescope on Haleakala with follow-up images confirming the object to be cometary in nature with a 10''–15''-long tail (Wainscoat et al. 2012). Observation campaigns performed by various ground-based observatories (Keck 10 m, the University of Hawaii 2.2 m, 6.5 m Baade, 6.5 m Clay Magellan, 1.8 m Perkins, etc.; Hsieh et al. 2013) found the object's intrinsic brightness roughly doubling from the time of its discovery until mid-November, after which it was then seen to decrease by ≈60% from late December to early February. Similar long-lived photometric behavior has been observed for several other MBCs, suggesting that the activity of P/2012 T1, deemed to belong to the Lixiaohua asteroid family, could likewise be driven by sublimation (Hsieh et al. 2012a, 2012b, 2013).

While sublimation of water ice in main-belt objects is strongly implied by MBC activity, gas has never been directly observed via spectroscopy, though attempts have been made (Jewitt et al. 2009; Hsieh et al. 2012b, 2012c, 2013; de Val-Borro et al. 2012a). Given the difficulty of detecting weak, distant, and transient gas emission, these non-detections do not rule out the presence of gas, however. A bright and actively sublimating MBC offers the unique opportunity to confirm the presence of sublimating volatile material in a main-belt object, and thus confirm the plausibility of ice in all of the other MBCs. This would strongly validate the potential that MBCs are believed to have for tracing the volatile content of the inner solar system. The Heterodyne Instrument for the Far Infrared (HIFI; de Graauw et al. 2010) on board the ESA Herschel Space Observatory (Pilbratt et al. 2010) proves to be the most sensitive instrument for directly observing water in a distant comet (e.g., Bockelée-Morvan et al. 2010). The HIFI instrument provides very high-resolution spectroscopy that can resolve the line shape and enable the determination of accurate production rates (e.g., Hartogh et al. 2010).

In this Letter, we present the Herschel DDT (Director Discretionary Time) observation of the 1₁₀–1₀₁ fundamental rotational transition of H₂O at 557 GHz in P/2012 T1. This observation was requested to be performed on 2013 January 16, at the opening of the Herschel visibility window for the target, when the MBC was deemed to be still active; this was confirmed via our DDT observations using the VLT FORS2 instrument (Appenzeller et al. 1998) performed on January 15. The Herschel observation was intended to test the prediction that the observed cometary activity of MBCs is driven by sublimation of water ices and to constrain the production process.

On the basis of its obvious and ongoing activity following its perihelion passage in early 2012 September, the MBC P/2012 T1 was observed by Herschel on UT 2013 January 16.31 with a total on-target integration time of 4.8 hr, when it was at an approximate heliocentric distance of 2.504 AU, a distance of 2.064 AU from the satellite (Herschel ObsID 1342259756), and at a phase angle of 2255. The object was tracked using an up-to-date ephemeris provided by the JPL Horizons system.

The line emission from the fundamental ortho-H₂O 1₁₀–1₀₁ line at 556.936 GHz was searched for in the upper sideband of the HIFI band 1a mixer. The observation was performed in the frequency-switching observing mode with a frequency throw of 94.5 MHz, using both the wide-band spectrometer (WBS) and the high-resolution spectrometer (HRS). In this observing mode there is no need to observe a reference position on the sky and the on-target integration time is maximized. However, the statistical noise may be underestimated for observations in frequency-switched mode owing to uncertainties in baseline removal (Bockelée-Morvan et al. 2012).

In addition to the Herschel observations, three 300 s V-band VLT FORS2 images were taken around 01:00 UT on 2013 January 15, with the telescope tracking the comet's motion. At this time the comet was at 2.502 AU from the Sun, 2.052 AU from Earth, and at a phase angle of 2226.

In the case of the HIFI data set, the standard HIFI processing pipeline v9.2 (belonging to the Herschel interactive processing environment software package; Ott 2010) was used to reduce the data to calibrated level 2 data products.

The spectral resolution of the WBS is 1 MHz (0.54 km s⁻¹ at the frequency of the observed line), while the HRS was used in its high-resolution mode with a resolution of 120 kHz (0.065 km s⁻¹). The main beam brightness temperature scale was computed using a beam efficiency of 0.75 and a forward efficiency of 0.96. The folded spectrum was obtained by averaging the original spectrum with a shifted and inverted copy. Horizontal and vertical polarizations were averaged, weighted by the root mean square amplitude, to increase the signal-to-noise ratio. The pointing offset of horizontal and vertical polarization spectra is 66 in band 1a, which, at the observed frequency, is approximately 20% of the half-power beam width.

The frequency switching observing mode introduces quite a strong baseline ripple that can be removed by performing a least-squares fit of a linear combination of sine waves. We applied the Lomb–Scargle periodogram method (initially proposed by Lomb 1976, and additionally developed by Scargle 1982) to the HRS and WBS spectra and fitted the baseline ripple using the strongest peaks in the frequency spectrum. The reduction methods applied for baseline removal and denoising of the Herschel/HIFI data are described in greater detail in de Val-Borro et al. (2012b). We show the averaged spectra of the two orthogonal polarizations in the HRS and WBS spectra in Figure 1.

Zoom In Zoom Out Reset image size

Table 1. H₂O Production Rate (Q_H2O) and Column Density (N_col) Derived for Different Gas Kinetic Temperatures

Temperature	$Q_{{\rm H_{2}O}}$	Ncol
(K)	(molecules s−1)	(cm−2)
10	<8.0 × 1025	<1.68 × 1011
15	<7.6 × 1025	<1.60 × 1011
20	<7.3 × 1025	<1.54 × 1011
Mean	<7.63 × 1025	<1.61 × 1011

Notes. Assuming a line area upper limit of 3.6 mK km s⁻¹ T_mBdv at 3σ, with the following excitation parameters r_h = 2.504 AU, Δ = 2.064 AU, v_exp = 0.5 km s⁻¹, x_ne = 0.2, pointing offset = 35 (if the comet was in between the two band 1A beams), one can obtain the following H₂O production rates and associated column densities for P/2012 T1, at different gas temperatures. Note that if a v_exp = 0.4 km s⁻¹ is used, this would lower the $Q_{{\rm H_{2}O}}$ by 15%.

Download table as:  ASCIITypeset image

In the case of the three VLT FORS2 V band images, the first image was affected by a cosmic ray near to the comet, but the sum of the remaining frames is shown in Figure 2. The comet was clearly active at that time, approximately one day before the Herschel observations. The seeing, measured using the FWHM of the star trails in this image, was around 08 and conditions were photometric. Photometry was performed with an aperture of radius 306 to avoid contamination from the nearby trailed star. The magnitude within this aperture is m_V = 21.47 ± 0.01, corresponding to a reduced magnitude of V(1,1,0) = 17.47, assuming a phase function of 0.02 mag deg⁻¹, appropriate for cometary dust (Meech & Jewitt 1987). The corresponding Afρ value (A'Hearn et al. 1989) is 2.7 cm (not including any phase angle correction), with the ρ = 306 aperture radius corresponding to 4560 km at the distance of the comet.

Zoom In Zoom Out Reset image size

The slope of the radial profile is −1.6, slightly steeper than expected for a steady-state coma under the influence of radiation pressure (Jewitt & Meech 1987), which means that Afρ is not independent of the choice of ρ in this case. We can say that it is very low, implying a weakly active comet. This low value is comparable with what has been provided by Hsieh et al. (2013) on this target for the January timeframe. The photometry also allows us to place an upper limit on the size of the nucleus: We obtain r < 1.3 km, assuming typical comet nucleus values for the geometric albedo (4%) and phase function (0.04 mag deg⁻¹).

Upon analyzing our data, we found no detection of H₂O, although with the fact that it was shown to be still active at the time of our observation, we expected the dust emission activity to be driven by the sublimation of subsurface material.

We used a molecular excitation model to calculate the population of the rotational levels of water as a function of nucleocentric distance as well as to derive the production rates. The model in use is based upon the publicly available accelerated Monte Carlo radiative transfer code ratran (Hogerheijde & van der Tak 2000). To derive the production rates, the code includes collisional effects as well as infrared fluorescence by solar radiation. We used the one-dimensional spherically symmetric version of the code as described in Bensch & Bergin (2004). This same version has been utilized to analyze both Herschel and ground-based cometary observations (Hartogh et al. 2010, 2011; de Val-Borro et al. 2010, 2012a, 2012b).

Input parameters to this model include the electron density, gas kinetic temperature, expansion velocity, and the radial gas density profile.

Electron density and gas kinetic temperature profile values extracted from (Biver et al. 1997; de Val-Borro et al. 2010) were fed into the model. For the case of the kinetic temperature (used to control the molecular excitation in the collisional region), a range of values from 10 to 20 K was input. Since the electron density in the coma is not well constrained, an electron density scaling factor of $x_{n_{e}}$ = 0.2 with respect to the standard profile derived from observations of comet 1P/Halley has been used (e.g., Hartogh et al. 2010).

The expansion velocity (assumed constant in the coma), and the radial gas density profile for water was obtained based upon a standard spherically symmetric Haser distribution. For lower-activity comets, and in this case P/2012 T1, an expansion velocity of 0.5 km s⁻¹1 has been used (Tseng et al. 2007; Biver et al. 2007). While Tseng et al. (2007) are limited by measurements taken for Q_OH < 10²⁸ at r_h distances >2 AU, the Biver et al. (2007) paper gives the FWHM = 0.95 km s⁻¹ for the comet C/2003 K4 (LINEAR) water line at 2.2 AU from the Sun, which can be interpreted as v_exp = 0.5 km s⁻¹, taking into account the fact that self-absorption makes the line a little narrower. Finally, from the temperature expected at the subsolar point where ice sublimates, we derived a thermal velocity of 0.35 km s⁻¹.

For the line area upper limit, based upon removal of sine baselines, with smoothing of approximately 25% and computed on a [−0.7 0.7 km s⁻¹] window, we derive a 1σ ∫ T_mBdv value of 1.63 mK km s⁻¹ from the HRS spectrum and 0.88 mK km s⁻¹ from the WBS spectrum.

Taking an upper limit of 3.6 mK km s⁻¹ ∫ T_mBdv at 3σ from the mean of the HRS and WBS values, we can obtain upper limits for the water production rate (Q) at different temperature ranges.

With this line area upper limit, using expansion velocity characteristic of weak comets (0.5 km s⁻¹), a gas kinetic temperature of 20 K, we derive a sensitive 3σ upper limit of Q < 7.3 × 10²⁵ molecules s⁻¹ from the WBS and HRS data (see Table 1). An upper limit of <7.63 × 10²⁵ molecules s⁻¹ is derived from the mean of the WBS and HRS upper limits for a gas expansion velocity of 0.5 km s⁻¹ and gas kinetic temperatures of 10, 15, and 20 K.

With the Herschel capability to directly detect H₂O in MBCs, it is important to stress the key relevance of such results to the knowledge of water production on these bodies.

In addition to the current work, this capability for direct measurements on MBCs has been demonstrated when MBC 176P/LINEAR was observed using Herschel/HIFI in 2011 August, about one month post-perihelion. In that case, no water line emission was detected and a 3σ upper limit was obtained on the production rate of <4 × 10²⁵ molecules s⁻¹ (de Val-Borro et al. 2012b).

Besides a direct detection, other means exist to derive water production rates although they are limited in nature and primarily model-dependent. We will look at two of these briefly for comparative purposes.

One particular way to derive H₂O production is via the search for CN. Searches for CN (and thus deriving a water production rate via a log based ratio linking Q_CN and Q_OH in "typical" comets) have taken place with four of the other MBCs, all of which were unsuccessful. For those MBCs, estimated water production rates of Q < 10²⁶ molecules s⁻¹ were proposed using a Q_CN/Q_OH mixing ratio of 10⁻³. For P/2012 T1, Hsieh et al. (2013) report that CN was also not detected but an estimate of the expected H₂O upper limit of Q < 5 × 10²⁵ molecules s⁻¹ based upon an average ratio for other comets has been made.

One can also derive a value for the production rate via the Afρ measurement using visible light observations to estimate the coma brightness of a typical Jupiter Family Comet. The visible light proxy for dust production is the quantity Afρ, where A is the bond albedo of the dust, f is the dust area filling factor in the field-of-view, and ρ is the projected radius of the field-of-view, typically expressed in units of cm (A'Hearn et al. 1984). For comae with constant isotropic outflows, Afρ is independent of aperture size. With our FORS2 observation, a corresponding Afρ value (A'Hearn et al. 1989) of 2.7 cm (not including any phase angle correction) has been estimated. Assuming a first order approximation, where Afρ in cm is found to be roughly equal to the dust production rate in kg s⁻¹ (A'Hearn et al. 1995), we find 2.7 kg s⁻¹ of H₂O = 9 × 10²⁵ molecules s⁻¹. Separate to this, we note that the photometry from our FORS2 observation also allowed us to place an upper limit on the radius of the nucleus which we estimate to be <1.3 km, assuming typical comet nucleus values for the geometric albedo (4%) and phase function (0.04 mag deg⁻¹).

The goal of the Herschel DDT observation was to detect H₂O in a clearly active MBC, assumed to be produced from sublimating subsurface ices. Besides sublimation of subsurface ices, other mechanisms have been proposed to drive mass loss from small bodies, including rotational instability, impact ejection and thermal fracture (see Jewitt 2012 for a recent review of mass loss mechanisms in MBCs). A study performed on this comet by Hsieh et al. (2013) for these different possibilities concluded that sublimation is indeed considered the key behind the observed activity.

Although the goal conditions of observing an active MBC were met, with our VLT FORS2 observation clearly showing activity and ground-based observations since that date confirming that the activity continued until the end of 2013 February (Moreno et al. 2013), we were, in the end, unsuccessful in achieving that goal as no detection of H₂O was made.

However, based upon our analysis above, and recognizing that: (1) the activity of the MBC observed at the time of the Herschel observation was already less than what was observed in the preceding months (Hsieh et al. 2013); (2) Herschel did not detect water but the measurement was successful in providing a sensitive 3σ upper limit; (3) the upper limits derived from less direct methods (Hsieh et al. 2013), e.g., Afρ measurements, provide values equivalent and comparable to our direct measurement (thus effectively supporting the strategies being used in such cases, where no direct measurements are available), we conclude that if the activity was due to sublimation of subsurface water ice, the water production rate at the time of our observations was lower than <7.63 × 10²⁵ molecules s⁻¹ for MBC P/2012 T1 (PANSTARRS).

We thank the Herschel Project Scientist and Time Allocation Committee for awarding five hours of Director Discretionary Time for this observation.

M.d.V.B. acknowledges support from grants NSF AST-1108686 and NASA NNX12AH91H.

Based in part on observations collected at the European Southern Observatory, Paranal, Chile, under programme 290.C-5028.

C.S. received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement No. 268421.