EDP Sciences
Free Access
Issue
A&A
Volume 532, August 2011
Article Number A87
Number of page(s) 11
Section Planets and planetary systems
DOI https://doi.org/10.1051/0004-6361/201116892
Published online 27 July 2011

© ESO, 2011

1. Introduction

The most important scientific rationale for studying comets is to retrieve information on their origin, on their relationship with interstellar, and on interplanetary material and implications for the formation of the Solar System. Deep Impact was the first mission to look beneath the surface of a comet (i.e. 9P/Tempel 1) (A’Hearn et al. 2005). The mission has provided us with unique information about primordial material in comets, global nucleus properties (density, material strength, ageing of cometary surface), and basic properties such as the crater formation on celestial bodies (A’Hearn 2007; Thomas et al. 2007).

Following the Deep Impact encounter with 9P/Tempel 1, the spacecraft was left on an orbit that gave the possibility of redirecting it to 103P/Hartley 2. The comet nucleus was planned to be flown by on November 4, 2010. Comet 103P/Hartley 2 was discovered in 1984 (Hartley 1984). It is an ecliptic comet that moved in the vicinity of the 1:1 mean motion resonance with Jupiter until 1875. After that date, a series of encounters with Jupiter reduced its perihelion from 2.9 AU to  ~1 AU and its aphelion from  ~8 AU to  ~6 AU (Carusi et al. 1985). At the time of the target selection by the mission (October 2007), the activity pattern of the comet was relatively unknown. From ISOPHOT data, Colangeli et al. (1999) obtained a H2O and CO2 production rates of (3.1 ± 0.2) × 1028 s-1 and (2.5 ± 0.3) × 1027 s-1, respectively, on January 1, 1998. This leads to a CO2/H2O abundance ratio of  ~8%, showing that 103P/Hartley 2 is a relatively CO2 abundant comet with respect to several other comets. Also, from data obtained with the instrument SWS on board ISO, Crovisier et al. (2000) report measurements on water rotational temperature, averaged in field of view of  ~20″ and the ortho-to-para ratio of  ~2.7, which corresponds to a spin temperature of about 35 K. The effective radius of the comet was first estimated as 0.71 ± 0.13 km (Groussin et al. 2004), giving rise to an active fraction of  ~1 at perihelion and 0.30 ± 0.11 at 1.11 AU post-perihelion. Later on, Lisse et al. (2009) could determine a comet nucleus radius of 0.57 ± 0.08   km, whereas Snodgrass et al. (2010) also reports a size estimate of  ≤ 1 km from data taken in 2008. Likewise, Lisse et al. (2009) claims that under circumstances of small size and high activity, the comet might suffer strong non-gravitational forces such that the spin state might undergo substantial changes in the instantaneous spin rate, but also the direction of the spin pole might change by as much as  ~30° in apex in a single perihelion passage.

An extensive monitoring of the comet activity long before, during, and after the fly-by has taken place all around the world (Meech et al. 2011). The aims were to characterize the comet activity pattern as the comet approached perihelion, to perform ground-based simultaneous observations with the mission fly-by and to put in context a large-scale gas and dust coma with the detailed nucleus and near-nucleus environment observed by the instruments on board the EPOXI mission.

In this article, we exploit the optical data acquired at the time of the comet perihelion, from the October 27 to the 29, 2010, to analyse the gas and refractory component of the coma. The observations and data reduction are explained in Sect. 2; the data analysis, our results and their comparison with others currently available are given in Sect. 3; and the summary appears in Sect. 4.

2. Observations and data reduction

The comet was observed from the Observatorio El Roque de los Muchachos on October 27 and 29, 2010 with the 4.2-m William Herschel Telescope (WHT) of the Isaac Newton Group (ING). We used the instrument ACAM (aux-port camera), which is mounted permanently at a folded-Cassegrain focus of the telescope. The CCD is a deep depletion device with additional fringe suppression and an AR coating peaking at 900 nm. The pixel size is 15 μm, which gives 0.25″/pixel.

ACAM can be used for imaging with a circular field of view with 8.3′ diameter and for low-resolution spectroscopy over the spectral range  ~350−940 nm with a resolution of  ~0.33 nm/pixel. The performance of the instrument is extraordinary, however scattered light from the Moon represents a severe problem when it is close to the zenith or when the observations are acquired within  ~25° of it. Unfortunately, during the comet perihelion, the Moon was at 13°, 1° and 10° from the comet on October 27, 28, and 29, 2010 respectively. The scattered moonlight typically manifests itself as a broad bright feature covering most of one half of the imaging area, and it has intensity a factor 2−3 times higher than on the uncontaminated imaging area. Therefore, it was impossible to acquire any useful data of 103P/Hartley on October 28, whereas during the other two nights the target had to be placed near the edges of the CCD to avoid the scattered moonlight contaminated area placed to the west (i.e. in the tail direction) as much as possible on October 27 and to the east (i.e. in the Sun direction) on October 29. Also, on October 29, there was a technical problem with one of the filter wheels such that no broadband images could be acquired.

Images of comet 103P/Hartley 2 were obtained with Sloan-Gunn R, I, G and Bessel U filters, and narrowband filters centred on strong comet gas emission bands (CN, C3, C2 and NH2) and on clear continuum regions in blue (BC, 443.0/3.5 nm) and red (RC, 684.0/7.7 nm) wavelength ranges. We took consecutive series of five and three images with the broad- and narrowband filters, respectively. Appropriate bias, dark, and flat field frames were also taken each night. There were sky photometric conditions with a typical mean seeing  ~ 0.7 − 1.0″ FWHM of the PSF during the observing campaign, excluding a sudden and brief change in the weather conditions on October 29 around 04:00 UT when the humidity varied almost by a factor of 1.5, the wind changed direction, and the seeing degraded to 2.8″ FWHM. Photometric stars were observed at an airmass similar to the comet observations for absolute flux calibration of the broadband filters. In the case of the CN, C3, C2, BC, and RC narrowband cometary filters, we observed the spectrophotometric standard star GD71 (at similar airmass as the comet) for converting the counts on the detector into erg   cm-2   s-1. To obtain gas (CN, C3 and C2) images of the comet 103P, the continuum contribution in the gas filter images has been removed by making use of the continuum scaling factor (Korsun & Jockers 2002) as in Lin et al. (2007). The continuum emission contributed  ~32% of the total signal in the CN filter passband,  ~28% in the C3 passband and  ~80% in the C2 passband.

Regarding the long-slit spectroscopy, the slit of the spectrograph was oriented along the E-W direction on October 29 giving dust and gas radial profiles in approximately the Sun-antisolar (PA = 276°) direction. However, due to an error during the instrument configuration on October 27, the slit was positioned along the direction 70°–250° counted from north towards east. This provides us with radial profiles that deviate by 23° from the antisolar direction (PA = 273°) (see Fig. 1). For absolute calibration, observations of the HZ 7 spectrophotometric standard star were acquired. For the comet observations, the slit width was 2″, whereas the usable selected length is 5.5′, which provides us with radial profiles along selected directions up to cometocentric projected distances of  ~15  000 km in each direction. The spectrophotometric standard star was observed with a width of 5″ and the same slit length. All comet observations were done with telescope tracking at the comet’s proper motion. Table 1 lists the details of the observations acquired during October 27 and 29, 2010. The spectrum on October 29 was acquired at the time when weather dramatically varied for about one hour during that night.

Since the gas emission of 103P/Hartley 2 covers the entire length of the slit, the sky subtraction has been estimated from the edges of the frame in continuum regions. More details on the images and spectra reduction and calibration can be found in Lara et al. (2001, 2011), Bertini et al. (2009) and Lin et al. (2007, 2009).

thumbnail Fig. 1

Images of 103P/Hartley 2 on October 27 and 29, 2010 on which we have superimposed an imaginary slit to show the directions for which dust and gas radial profiles are obtained during the spectroscopic observations.

Open with DEXTER

Table 1

Log of observations.

The broad- and the BC and RC narrowband images have been used to study (i) the dust coma morphology by applying the radial renormalization (A’Hearn et al. 1986) and the Larson-Sekanina method (Larson & Sekanina 1984); (ii) the azimuthally averaged profile of surface brightnesses; (iii) Afρ, and (iv) the dust colour maps as a function of projected cometocentric distance ρ and of time during the perihelion. The CN, C3, and C2 comet images have provided us with gas coma morphology, whereas the NH2 images were only considered for an initial inspection of the coma since no observations of the calibration star GD71 were acquired with this filter, and thus no flux calibration has been done for the NH2 images.

On the other hand, the spectra have provided us with the production rates for CN, C3, C2 and NH2 in the context of the Haser model (Haser 1957), with dust brightness profiles as a function ρ along the directions imposed by the slit (see Fig. 1), with dust colour variations in those same directions and with Afρ in different continuum spectral ranges.

thumbnail Fig. 2

Images of 103P/Hartley 2 on October 27, 2010, acquired with G, R and I Sloan-Gunn filters (from left to right), together with their corresponding radially renormalized images to enhance the structures in the coma. The FOV is 125″  ×  125″, meaning 11 600  ×  11 600 km at the comet’s distance.

Open with DEXTER

3. Data analysis and results

3.1. Dust

The analysis of the dust coma of comet 103P/Hartley 2 was carried out from the images and the spectra. This analysis comprises: coma morphology, dust production rate through the Afρ parameter, and dust colour from the 2D images and from the long-slit spectra.

To determine whether some morphological structures are present in the coma of 103P/Hartley 2, we investigated all of the flux-calibrated images acquired, either in broadband filters as in cometary narrowband filters with bandpass centred on clear continuum regions, that is, 443 (BC) and 684 nm (RC). From a first inspection, the green and red broadband images clearly display the dust+gas coma. The bandpass of the G and R Sloan-Gunn filters do contain strong gas emission in C2, NH2, and H2O + . On the other hand, the comet images acquired with the bandpass of the I Sloan-Gunn filter show much less gas contamination, although fainter cometary gas emissions (the CN red system) also fall in the bandpass. Figure 2 shows the comet images in G, R, and I Sloan-Gunn together with their radially renormalized frames (A’Hearn et al. 1986) in each filter. Aside from a clear dust tail, no evident feature can be detected in these images except in the one obtained with R broadband where a faint jet seems to be detected in Sun direction.

Therefore, a meaningful morphological analysis of the dust coma has been done on the flux-calibrated images acquired with the BC and RC cometary filters. Figure 3 displays the dust coma of 103P on October 27 and 29 imaged with the BC narrowband filter, the RC one, the quotient of the images in BC and RC taken a few hours apart, and the radially renormalized RC images.

thumbnail Fig. 3

Flux-calibrated images of 103P/Hartley 2 in a look-up table to display the isophote anisotropies. The images are acquired with clear continuum filters centred at 443 nm with cuts between 0 and 2 × 10-15 erg   cm-2   s-1, and at 684 nm with cuts between 0 and 8 × 10-15 erg   cm-2   s-1. The third column shows the quotient of the comet images obtained a few hours apart (i.e. 2nd column/1st column) with cuts between 1 and 1.6. The fourth column displays the radially renormalized images of the ones in the second column. The FOV is 62.5″  ×  62.5″, meaning 5800  ×  5800 km at the comet’s distance.

Open with DEXTER

A detailed inspection of the processed images shows that, beside the dust tail, on October 27 at 01:41 UT a faint jet in the sunward direction starts to be visible in the red continuum (see Fig. 4), becoming stronger and more obvious in the images acquired approximately two hours later with both the BC and RC narrowband cometary filters.

On October 27, both for the blue and red continuum, the quotient image (third column in Fig. 3) shows a clear dust structure pointing towards the Sun and slightly curved to the south. This feature is not clearly detected on October 29. Both the dust tail and the curved jet pointing to the Sun are also enhanced by means of the Larson-Sekanina method (Larson & Sekanina 1984). An independent verification can be done by looking the anisotropies in the isophotes (first two columns in Fig. 3). The turn on/off of the curved jet in the eastern direction has been explained as the response of an active area that faces the Sun coinciding with the nucleus spin period (Harmon et al. 2011) during October 24–27, 2010. This finding in our data is confirmed by the EPOXI Team when analysing a sequence of images acquired by the HRI and MRI instruments on board. Assuming that on October 27 between 01:42 and 03:33 this active area faced the Sun and that the nucleus spin period is 18.1 h (Harmon et al. 2011), the dust coma of 103P would have displayed the same structure on October 29 at  ~07:30 UT i.e. during dawn at the Observatorio El Roque de los Muchachos but unfortunately was not observable.

From the set of images and spectra on October 27 and 29, we have characterized the dust coma as much as possible in terms of dust colour and production rates (Afρ), and of surface brightness radial profiles. Figure 5 shows the maps of dust reddening in %/100   nm for October 27 and 29, computed following the relation (Turner & Smith 1999) (1)where RC and BC refer to the comet images acquired with the red and blue narrowband cometary filters (centred at 684.0 nm and at 443.0 nm, respectively) after removing the solar contribution. The average dust reddening in a circular aperture of  ~500 km of radius is  ~24%/100   nm for both dates, a rather common value of cometary dust colour. This behaviour of the dust colour seems to indicate that the optical properties of the dust grains do not significantly vary regardless of where they are measure in the jet or in the surrounding coma. However, it is interesting that the dust in the jet has different optical properties when compared with the “older” dust in the background coma (see bottom right in Fig. 5). Comparison of the dust coma in the RC, when the jet is already clearly visible in the coma, with the dust coma in the BC 2 h before, shows dust reddening variations. We measure normalized dust reflectivity of 45%/100   nm at the optocentre decreasing to 24%/100   nm at the outer edge (~840 km) along the jet axis. The origin of this colour gradient along the jet axis can be explained as the consequence of comparing old processed dust and new refractory material dragged by the outgassing when the active area is exposed to the Sun. This higher reddening around the optocentre characterizes this new material freshly expelled from the comet nucleus. These refractory particles get bluer while travelling outwards, meaning that there are variations in the optical properties caused either by fragmentation, ice sublimation, or both.

The long-slit spectra cover a wide wavelength range and they contain spatial information in several directions. Therefore, the dust colour and its variation with ρ can be determined from these measurements, too. Several band passes representing continuum emissions are 415.0–417.5 nm, 440.0–437.0 nm, 483.3–489.0 nm, 522.0–530.0 nm, 677.5–692.5 nm, and 680.0–686.0 nm. The comet spectra have been divided by the solar spectrum (Kurucz et al. 1984), degraded to the same spectral resolution as the ACAM data to remove the solar colour when computing the cometary dust colour. Normalized reflectivity, or dust reddening, is computed as in Eq. (1) where we substitute BC and RC for different combinations of the blue and red spectral ranges mentioned above in wavelength regions where differences in sky colour are not dramatically high. Figure 6 shows the continuum reflectivity slope for October 27 and 29, 2010 along the directions dictated by the slit positioning (see Sect. 3 and Fig. 1). We have only considered the innermost coma, i.e.  − 1500 ≤ ρ ≤ 1500 km, to investigate the dust colour reddening from the spectroscopic measurements, because the scattered moonlight which falls in the slit length poses a severe problem at distances ρ ≥ 2000 km from the optocentre. For both dates, we see that dust colour gets considerably bluer in the sunward hemisphere, whereas it can be considered constant in approximately tail direction. The absolute values of this dust reddening depend on the spectral ranges, but the trend explained above is confirmed for all of them (see Fig. 6).

thumbnail Fig. 4

Radially renormalized image of comet 103P/Hartley 2 obtained on October 27, 2010 at  ~ 01:40 UT with the RC cometary filter. The arrow points to the faint jet that starts to be visible at that time. Its intensity increases two hours later as clearly visible in Fig. 3. The FOV is 62.5″  ×  62.5″, meaning 5800  ×  5800 km at the comet’s distance.

Open with DEXTER

Table 2 lists the slope of the surface brightness radial profiles for both October 27 and 29, and Fig. 7 shows the dust surface brightness profiles in the direction at PA 70–250° and in E-W direction (PA 90–270°), respectively. The surface brightness B is computed from the flux-calibrated continuum profiles multiplied by 2πΔ2 and and normalized to a unit area of 1 km2 at the position of the comet. This leads to a surface brightness in units of energy per second per square kilometer per angstrom (erg s-1 km-2 Å-1), which is independent of the geocentric distance, as well as of the size of the aperture and normalized to the solar flux at rH = 1 AU. It therefore allows a direct comparison of the day-to-day variations in the continuum profiles. The results in Table 2 and in Fig. 7 support the idea that the feature in the sunward hemisphere was still detectable in the coma of 103P/Hartley 2 about four hours after its activation at 01:00 UT according to the EPOXI Team (see http://epoxi.umd.edu/3gallery/20101027-jets.shtml) since our data were acquired at  ~ 05:00 UT on October 27, 2010. The similarity between the log B versus log ρ profiles on October 27 and 29 also indicates that the jet was present in the eastern hemisphere on October 29 at 04:01 UT, although this cannot be confirmed through dust images since none was acquired after the spectrum acquisition (see Table 1).

thumbnail Fig. 5

Dust reddening, in % per 100 nm, of the coma of 103P/Hartley 2 on October 27 and 29, 2010. Bottom right frame in the figure displays the dust reddening of the new material in the jet compared with the older dust in the coma on October 27. The look-up table is linear and stretches from 15 to 40%/100   nm.

Open with DEXTER

thumbnail Fig. 6

Dust colour profiles, or normalized reflectivity S′ in units of % per 100 nm as a function of projected cometocentric distance, ρ, on October 27 and 29, 2010 obtained from the spectroscopic measurements. Negative ρ refers to the 70° direction on October 27 and to east direction on October 29. Positive ρ refers to the direction at PA = 250° on October 27 and to the westward direction on October 29. Black crosses: normalized reflectivity computed between the ranges 415.0–417.5 nm and 522.0–530.0 nm. Red asterisks: normalized reflectivity computed between the ranges 483.3–489.0 nm and 680.0–686.0 nm. Green open squares: normalized reflectivity computed between the ranges 440.0–437.0 nm and 677.5–692.5 nm, which are the same regions as the BC and RC filters.

Open with DEXTER

thumbnail Fig. 7

Dust surface brightness profiles B as a function of projected cometocentric distance ρ in double logarithmic representation, obtained from the spectroscopic measurements on October 27 and 29, 2010 in the  ~ tailwards (red crosses) – sunward (black open circles) directions. The selected continuum region is 482–485 nm. Symbols refer to the observational data and a solid line refers to the linear fit.

Open with DEXTER

Table 2

Linear fits to the log B – log ρ profiles obtained from the long-slit spectroscopic measurements, giving the slope m.

The release of dust from the comet is approximately determined through optical measurements of the parameter Afρ (cm) as a function of the projected cometocentric distance ρ (A’Hearn et al. 1984) (2)where Δ (in cm) and rh (in AU) are the comet’s geocentric and heliocentric distances, respectively, Fc is the measured cometary flux in the selected filter (narrowband cometary filters BC and RC) integrated in the radius of aperture ρ, and FS is the total solar flux in the same filter.

Table 3 contains the Afρ values measured on October 27 and 29, 2010 from the BC and RC comet images (see Table 1), as well as from the spectroscopic measurements in three different continuum regions. In the case of the images, we considered a circular aperture of radius ρ ~ 4700 km. The parameter Afρ in the blue and red continuum is 59 and 102 cm, respectively, on October 27, and 56 and 111 cm on October 29. The dust colour reddening computed from these two values of Afρ is  ~ 22%   nm, in good agreement with results from spectroscopic data. Regarding the spectroscopic measurements, Afρ can also be computed in several continuum regions (see Table 2). For this, we have determined the equivalent radius of the aperture provided by the long-slit (width = 2″, useful length = 330″, pixel size = 0.25″) according to the expression given in Roettger (1991), yielding 4.329″ (~400 km). The results are also listed in Table 3 on both dates. While there is an overall agreement between the Afρ computed from the BC and RC images and the long-slit spectrum on October 27, this not the case for October 29 when the long-slit spectrum generally provides much lower Afρ values than two days before and also noticeably different in the BC range. Since the comet activity variations reported by Biver et al. (2010, priv. comm.) do not reach a factor of  ~2, we have checked the weather conditions at the time of the spectrum acquisition. Unfortunately, the humidity increased to  ~50%, the wind dramatically changed direction and the seeing increased to 2.8″. Therefore, the reason for such a discrepancy most likely lies in the unstable weather conditions during data acquisition and not on intrinsic comet activity variability.

Table 3

Dust (Afρ) and gas production rates of the comet 103P/Hartley 2.

3.2. Gas

Although the signal-to-noise ratio of the data acquired with the gas narrowband cometary filters is not very high, we have investigated the general spatial distribution of gas species in the coma of comet 103P/Hartley 2 in the CN, C2, and C3 images.

In CN and C3, the coma is noticeably asymmetric on both dates, showing a north-south elongation with clear differences between October 27 and October 29, 2010 in the case of CN. The first two frames in Fig. 8 show the CN image after processing with the Larson-Sekanina technique. It is clear that on October 27, there is a more pronounced over abundance (~10%) of CN particles in the southern hemisphere (as projected on the plane of the sky), whereas on October 29, both cones in the northern and southern hemispheres display similar over abundances (~5%) when compared to the isotropic gas coma. These variations in the CN jet appearance are related to the spin period of the comet nucleus as already reported by Samarasinha et al. (2011). On October 29, we detected that the C3 coma is also asymmetric, displaying two cones in the north-south direction, centred on the nucleus, where the column density is enhanced with respect to any other direction. This behaviour is rather similar to the one seen in CN. No C3 data were acquired on October 27, therefore we cannot conclude whether a similar anisotropic C3 coma already existed. On the other hand, the C2 gas coma does not display any evident asymmetry.

thumbnail Fig. 8

Enhanced gas images of the comet obtained on October 27 and 29, 2010. The Larson-Sekanina method has been applied to fully calibrated images after subtracting the underlying continuum as explained in Sect. 2. From left to right: CN on October 27 at 02:45 UT (mid-time during exposure), CN on October 29 at 05:15 UT (mid-time during exposure), C3 on October 29 at 03:02 UT (mid-time during exposure). The FOV is 11   800 × 11   800 km. Aside from the reported CN jet by Samarasinha et al. (2011) and Meech et al. (2011), the C3 coma distribution also displays similar jet structures in the north-south direction.

Open with DEXTER

The acquired spectra of the comet are used to investigate the CN, C3, C2, and NH2 column density profiles in the directions imposed by the slit positioning, that is, in the 70°–250° direction from north towards east on October 27 and in E-W direction on October 29, 2010. Figure 9 shows the flux-calibrated spectrum of comet 103P/Hartley 2 acquired on October 27, 2010 where the strongest cometary emissions in the optical range (CN, C2, C3, and NH2) are indicated. The slit was positioned as indicated in Fig. 1.

thumbnail Fig. 9

Flux-calibrated spectrum of comet 103P/Hartley 2 acquired on October 27, 2010 where the strongest cometary emissions in the optical range (CN, C2, C3, and NH2) are marked. The slit was positioned as indicated in Fig. 1. The spatial scale covers from the optocentre up to  ~12   000 km in both directions. The spatial resolution is  ~23 km/pix. The spectral wavelength range runs from 366.0 to 780.3 nm with resolution 0.33 nm/pix. The colour code stretches from 0 to 5 × 10-16 erg   cm-2   s-1   Å-1.

Open with DEXTER

The spectral regions and the subtraction of the underlying continuum in the gas emission bands is done as described by Lara et al. (2001). The conversion of the emission band fluxes into column densities makes use of constant g-factors for C2, C3, and NH2 (A’Hearn et al. 1995), whereas the g-factor of the CN molecule is calculated for the heliocentric distance and velocity of 103P/Hartley 2 for both dates from the set of values given by Schleicher (1983). Although it has been reported that the gas production rate of 103P is not steady, showing variations in days and minutes Biver et al. (2010, priv. comm.), for giving any indication of the gas production rates, we computed them by means of the Haser modelling (Haser 1957) with parent velocity vp scaled with rh ( km   s-1), customary values for the daughter velocity vd = 1 km   s-1 and scale lengths given in A’Hearn et al. (1995). For the corresponding set of parameters in the Haser modelling, we produced theoretical column density profiles for each species by varying the production rate until the best match between observations and theoretical predictions was achieved.

thumbnail Fig. 10

Observed gas column density profiles N in units of cm-2 versus the projected cometocentric distance ρ in log-log representation in 70° and in east direction for October 27 and 29, 2010, respectively. The best achievable fit obtained by means of the Haser modelling (Haser 1957) with parameters given in the text and production rates listed in Table 3 is plotted with solid lines.

Open with DEXTER

thumbnail Fig. 11

Observed gas column density profiles N in units of cm-2 versus the projected cometocentric distance ρ in log-log representation in 250° and in west direction for October 27 and 29, 2010, respectively. The best achievable fit (solid line) obtained by means of the Haser modelling (Haser 1957) with parameters given in the text and production rates listed in Table 3.

Open with DEXTER

The production rates, Qi, led to a best match for the observed column densities, the quotient [Q(C2)/Q(CN)], and log [Afρ(520 − 525   nm)/Q(C2)] are listed in Table 3, whereas Figs. 10 and 11 show the observed CN, C3, C2, and NH2 column density radial profiles in the directions probed by the slit, together with the best achievable Haser fit. The observed and theoretical gas profiles in Figs. 10 and 11 show the complexity of fitting these profiles as the coma is rather asymmetric. The use of any model (either vectorial modelling (Festou 1981a,b) or time-dependent one as considered by (Lara et al. 2001) for comet 1996/Q1 Tabur) will be unable to reproduce the E-W asymmetries. Some improvement in the fits can be achieved when “time-dependent” production rates are considered. For implementing this model, the production rate is doubled for CN (and to a lesser extent for the other species) in a short fraction of a day,  ~0.1, whereas the steady comet activity can be reproduced by using the values listed in Table 3. Since the step-function change in production rate is merely for computational ease, this result only confirms the rapid and periodical increase in the cometary activity.

Also, another problem is the short geocentric distance and the available slit length that has allowed us to measure the gas column densities up to projected cometocentric distances of only 10 000 km. This, together with an approximate sky subtraction, gives rise to observed profiles noticeably deviate from the Haser theoretical profiles beyond log    (ρ) ≈ 3.8. For the species with long scale lengths, CN, C2, and NH2, the theoretical profiles predict column densities which fall with ρ in a much less pronounced way than the observed ones. However, better fits are obtained for C3 at every ρ as the parent and daughter scale lengths are shorter than for the other analysed species.

On October 27, the fit is less accurate for CN than on October 29 as the enhanced emission of the two cones structures falls within the slit given its orientation in the PA 70–250°. This does not happen on October 29 when the slit is placed exactly in the E-W direction, that is, perpendicular to the direction where CN structures are detected. Furthermore, regarding C3, the situation is reversed, and the disagreement between Haser modelling and observations is greater on October 29 than on October 27, 2010.

Table 4

Comparison of gas and dust activity of 103P/Hartley 2 at the 1991 (A’Hearn et al. 1995) and 2010 perihelion passages.

Based on the results in A’Hearn et al. (1995) and in this work (see Table 4), we can conclude that, for a rather similar OH production rates in both passages (2.1 × 1028 s-1 in 1991 and 1.8 × 1028 s-1 in 2010), the production rates Q of CN, C3, C2, and NH2 measured at the 2010 perihelion passage are slightly higher than the ones measured during the 1991 perihelion passage.

Mumma et al. (2011) report spectroscopic observations of HCN on UT 22.5 October. The HCN production rate on this pre-perihelion date is (1.7 ± 0.1) × 1025 s-1. They oriented the slit along the sun-comet line, and the retrieved profiles were rather symmetric around the nucleus. The QCN obtained in this work is consistent with the Mumma et al. (2011) results that assume that HCN is the main parent species of CN and that the expected variations of QHCN around the perihelion are not very large. Our CN spectroscopic measurements also agree with the fact that the CN (assuming HCN as the parent species) asymmetry lies in the north-south direction and not in sun-tail one. It has been reported that the CN jet appearance changes in morphology as a function of time with a periodicity of  ~ 17.6 h (Samarasinha et al. 2011) around the perihelion date, thus the north-south jet of CN on UT 27.21 October would have been placed at the same orientation on UT 22.81 October, about five hours later than the Mumma et al. (2011) observations. Biver et al. (2010, priv. comm.) have indicated that HCN production rate varied in the range (1−2) × 1025 s-1 with a period of 17.6 h. These variations peaked around UT 25.10 and 27.24 October (i.e. three nucleus rotations) and broad minima were found around UT 25.4, 26.2, and 26.9 October. The comet spectrum on October 27 was acquired at about the same time as the reported maximum on UT 27.24 October. Comparing the CN production in Table 3 and the maxima of HCN production rates, we can thus conclude that on UT 27.24 October the 50−100% of the HCN outgassed from the nucleus is producing CN.

The quotients of the C2 and CN gas production rates slightly increases by 5% from October 27 to 29, which we cannot consider as representative of comet outgassing variations because of the error in the flux calibration and Qi determination (5−7%) and because of the varying weather conditions during the 20 min integration time for the spectrum acquisition. A mean value of C2/CN  ~1.2 is given in A’Hearn et al. (1995), and  ~1.3 is reported by Meech et al. (2011); therefore our determination can be considered as realistic, aside from the unstable sky conditions on October 29. According to this quotient, C2/CN, comet 103P/Hartley 2 can be classified as typical, in terms of abundance of long-chain hydrocarbons, following the taxonomic types identified in A’Hearn et al. (1995).

The log [Q(C2)/Afρ], with Afρ measured in the 520–525 nm spectral range, provides us with an approximate indication of the gas or dust enrichment of comet 103P. The result of this quotient is  ~23. In line with A’Hearn et al. (1995), the gas-to-dust mass ratio can be evaluated by (3)According to this expression and using Q(H2O) = (1 − 2) × 1028 s-1 (Biver et al. 2010, priv. comm.) and Afρ from results in Table 3, we compute a gas-to-dust mass ratio of  ~3−6, a relatively gas-rich comet.

4. Summary

The comet 103P/Hartley 2, target of the EPOXI mission (NASA) was observed for three days around its perihelion, from October 27 to 29, 2010. The images and long-slit spectra that we acquired have provided us with the following picture of the comet at a distance of 1.058 AU from the Sun.

Both the gas and dust coma show features. The CN and C3 coma is not spherically symmetric, but a double jet in the north-south direction is detected for both species on October 27 and 29, 2010. The CN double jet (or cone) changes morphology from 27 to 29 October: on October 27, the over-abundance of CN is more pronounced at the southern hemisphere than at the northern one, whereas on October 29, the two cones show similar CN over-abundance. The C3 coma displays a north-south asymmetry similar to the CN coma on October 29, since it was not being possible to determine its existence on October 27 because no C3 images were acquired. The dust coma shows the tail in the anti-sunward direction, and a short Sunward jet curved to the south, which turns on and off following the comet nucleus periodicity (i.e. about  ~ 18.3 h at the time of the perihelion).

The Afρ parameter, a proxy to the dust production rate, measured in a circular aperture of  ~ 4700 km ranges from  ~60 cm in the blue and  ~110 cm in the red. However, Afρ is rather stable in the innermost coma (equivalent aperture radius of  ~400 km) when it is computed from the spectroscopic measurements within several continuum spectral ranges from 482 to 685 nm. The strong variation in the Afρ parameter between the blue and red spectral ranges when considering large aperture radius indicate a considerably dust reddening of  ~24%/100 nm. This normalized dust reflectivity varies with ρ in the Sunward and tailward directions. Both directions display a blueing of the cometary dust, more noticeably in the Sun’s direction.

The BC and RC surface brightness profiles, computed from azimuthally averaged profiles of the comet images, show a nominal behaviour of log    B ~  −1.0   log    ρ. However, a detailed study in sunward and tailward direction clearly shows that this slope m is flatter (m ~  −0.5   to    −0.6) along the tail and steeper (m ~  −1.1 to  −1.3) towards the Sun.

The gas (CN, C3, C2, and NH2) production rates measured at perihelion, rh = 1.058 AU, are a bit higher than those during the 1991 perihelion passage (A’Hearn et al. 1995). The quotient QC2/QCN ~ 1.3 places 103P/Hartley 2 as a typical comet in terms of long-chain hydrocarbons abundance. The gas-to-dust mass ratio is  ~3−6, allowing us to consider 103P as a gas-rich comet.

Acknowledgments

This research has been supported by the Spanish Ministerio de Ciencia e Innovación under contract AyA 2009-08011. Z.-Y. Lin acknowledges a grant by the Junta de Andalucia under the project JA2007-TIC2744. K. J. Meech acknowledges support from NASA’s Discovery Program through contract NNM07AA99C. L. M. Lara is very grateful to the WHT staff, especially to O. Vaduvescu, S. Rix, and F. Riddick for their support before and during the observations. L. M. Lara also acknowledges J. M. Bacaicoa for his help in editing the images.

References

All Tables

Table 1

Log of observations.

Table 2

Linear fits to the log B – log ρ profiles obtained from the long-slit spectroscopic measurements, giving the slope m.

Table 3

Dust (Afρ) and gas production rates of the comet 103P/Hartley 2.

Table 4

Comparison of gas and dust activity of 103P/Hartley 2 at the 1991 (A’Hearn et al. 1995) and 2010 perihelion passages.

All Figures

thumbnail Fig. 1

Images of 103P/Hartley 2 on October 27 and 29, 2010 on which we have superimposed an imaginary slit to show the directions for which dust and gas radial profiles are obtained during the spectroscopic observations.

Open with DEXTER
In the text
thumbnail Fig. 2

Images of 103P/Hartley 2 on October 27, 2010, acquired with G, R and I Sloan-Gunn filters (from left to right), together with their corresponding radially renormalized images to enhance the structures in the coma. The FOV is 125″  ×  125″, meaning 11 600  ×  11 600 km at the comet’s distance.

Open with DEXTER
In the text
thumbnail Fig. 3

Flux-calibrated images of 103P/Hartley 2 in a look-up table to display the isophote anisotropies. The images are acquired with clear continuum filters centred at 443 nm with cuts between 0 and 2 × 10-15 erg   cm-2   s-1, and at 684 nm with cuts between 0 and 8 × 10-15 erg   cm-2   s-1. The third column shows the quotient of the comet images obtained a few hours apart (i.e. 2nd column/1st column) with cuts between 1 and 1.6. The fourth column displays the radially renormalized images of the ones in the second column. The FOV is 62.5″  ×  62.5″, meaning 5800  ×  5800 km at the comet’s distance.

Open with DEXTER
In the text
thumbnail Fig. 4

Radially renormalized image of comet 103P/Hartley 2 obtained on October 27, 2010 at  ~ 01:40 UT with the RC cometary filter. The arrow points to the faint jet that starts to be visible at that time. Its intensity increases two hours later as clearly visible in Fig. 3. The FOV is 62.5″  ×  62.5″, meaning 5800  ×  5800 km at the comet’s distance.

Open with DEXTER
In the text
thumbnail Fig. 5

Dust reddening, in % per 100 nm, of the coma of 103P/Hartley 2 on October 27 and 29, 2010. Bottom right frame in the figure displays the dust reddening of the new material in the jet compared with the older dust in the coma on October 27. The look-up table is linear and stretches from 15 to 40%/100   nm.

Open with DEXTER
In the text
thumbnail Fig. 6

Dust colour profiles, or normalized reflectivity S′ in units of % per 100 nm as a function of projected cometocentric distance, ρ, on October 27 and 29, 2010 obtained from the spectroscopic measurements. Negative ρ refers to the 70° direction on October 27 and to east direction on October 29. Positive ρ refers to the direction at PA = 250° on October 27 and to the westward direction on October 29. Black crosses: normalized reflectivity computed between the ranges 415.0–417.5 nm and 522.0–530.0 nm. Red asterisks: normalized reflectivity computed between the ranges 483.3–489.0 nm and 680.0–686.0 nm. Green open squares: normalized reflectivity computed between the ranges 440.0–437.0 nm and 677.5–692.5 nm, which are the same regions as the BC and RC filters.

Open with DEXTER
In the text
thumbnail Fig. 7

Dust surface brightness profiles B as a function of projected cometocentric distance ρ in double logarithmic representation, obtained from the spectroscopic measurements on October 27 and 29, 2010 in the  ~ tailwards (red crosses) – sunward (black open circles) directions. The selected continuum region is 482–485 nm. Symbols refer to the observational data and a solid line refers to the linear fit.

Open with DEXTER
In the text
thumbnail Fig. 8

Enhanced gas images of the comet obtained on October 27 and 29, 2010. The Larson-Sekanina method has been applied to fully calibrated images after subtracting the underlying continuum as explained in Sect. 2. From left to right: CN on October 27 at 02:45 UT (mid-time during exposure), CN on October 29 at 05:15 UT (mid-time during exposure), C3 on October 29 at 03:02 UT (mid-time during exposure). The FOV is 11   800 × 11   800 km. Aside from the reported CN jet by Samarasinha et al. (2011) and Meech et al. (2011), the C3 coma distribution also displays similar jet structures in the north-south direction.

Open with DEXTER
In the text
thumbnail Fig. 9

Flux-calibrated spectrum of comet 103P/Hartley 2 acquired on October 27, 2010 where the strongest cometary emissions in the optical range (CN, C2, C3, and NH2) are marked. The slit was positioned as indicated in Fig. 1. The spatial scale covers from the optocentre up to  ~12   000 km in both directions. The spatial resolution is  ~23 km/pix. The spectral wavelength range runs from 366.0 to 780.3 nm with resolution 0.33 nm/pix. The colour code stretches from 0 to 5 × 10-16 erg   cm-2   s-1   Å-1.

Open with DEXTER
In the text
thumbnail Fig. 10

Observed gas column density profiles N in units of cm-2 versus the projected cometocentric distance ρ in log-log representation in 70° and in east direction for October 27 and 29, 2010, respectively. The best achievable fit obtained by means of the Haser modelling (Haser 1957) with parameters given in the text and production rates listed in Table 3 is plotted with solid lines.

Open with DEXTER
In the text
thumbnail Fig. 11

Observed gas column density profiles N in units of cm-2 versus the projected cometocentric distance ρ in log-log representation in 250° and in west direction for October 27 and 29, 2010, respectively. The best achievable fit (solid line) obtained by means of the Haser modelling (Haser 1957) with parameters given in the text and production rates listed in Table 3.

Open with DEXTER
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.