A belt-like distribution of gaseous hydrogen cyanide on Neptune's equatorial stratosphere detected by ALMA
Takahiro Iino, Hideo Sagawa, Takashi Tsukagoshi, Satonori Nozawa
DDraft version September 30, 2020
Typeset using L A TEX manuscript style in AASTeX62
A belt-like distribution of gaseous hydrogen cyanide on Neptune’s equatorial stratosphere detectedby ALMA
Takahiro Iino, Hideo Sagawa, Takashi Tsukagoshi, and Satonori Nozawa Information Technology Center, The University of Tokyo, 2-11-16, Yayoi, Bunkyo, Tokyo 113-8658, Japan Faculty of Science, Kyoto Sangyo University, Motoyama, Kamigamo, Kita-ku, Kyoto 603-8555, Japan National Astronomical Observatory of Japan, 2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan Institute for Space-Earth Environmental Research, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Aichi464-8601, Japan (Received; Revised; Accepted)
ABSTRACTWe present a spatially resolved map of integrated- intensity and abundance of Nep-tune’s stratospheric hydrogen cyanide (HCN). The analyzed data were obtained fromthe archived 2016 observation of the Atacama Large Millimeter/submillimeter Array.A 0. (cid:48)(cid:48) × (cid:48)(cid:48)
39 synthesized beam, which is equivalent to a latitudinal resolution of ∼ ◦ at the disk center, was fine enough to resolve Neptune’s 2. (cid:48)(cid:48)
24 diameter disk. Af-ter correcting the effect of different optical path lengths, a spatial distribution of HCNemissions is derived over Neptune’s disk, and it clearly shows a band-like HCN enhance-ment at the equator. Radiative transfer analysis indicates that the HCN volume mixingratio measured at the equator was 1.92 ppb above the 10 − bar pressure level, which is40% higher than that measured at the southern middle and high latitudes. The spatialdistribution of HCN can be interpreted as either the effect of the transportation of [email protected] a r X i v : . [ a s t r o - ph . E P ] S e p Iino et al. N from the troposphere by meridional atmospheric circulation, or an external supplysuch as cometary collisions (or both of these reasons). From the meridional circulationpoint of view, the observed HCN enhancement on both the equator and the pole canbe explained by the production and accumulation of HCN at the downward branchesof the previously suggested two-cell meridional circulation models. However, the HCN-depleted latitude of 60 ◦ S does not match with the location of the upward branch of thetwo-cell circulation models.
Keywords: planets and satellites: atmospheres — submillimeter: planetary systems INTRODUCTIONThe presence of hydrogen cyanide (HCN) characterizes Neptune’s stratospheric composition. Thefirst detection of HCN was made by single-dish millimeter/submillimeter telescopes observing therotational transitions of J =3–2 and 4–3 (Marten et al. 1993; Rosenqvist et al. 1992). The observedline widths of HCN were significantly narrow as ∼
20 MHz, indicating that HCN is present in the upperstratosphere. The volume mixing ratio (VMR) determined from J =3–2 and 4–3 observations were3.0 × − and 1.0 × − , respectively. The HCN was restricted to be present at altitudes above theHCN condensation level, 3.5 mbar atmospheric pressure region. Subsequent sensitive disk-averagedobservations also identified vertical distributions, and suggested that HCN was present in the regionabove 0.9–3 mbar, where the condensation cannot occur (Marten et al. 2005; Rezac et al. 2014).To explain the production of HCN in Neptune’s atmosphere, Lellouch (1994) developed a photo-chemical model of N-bearing species. The production of HCN from dissociated N-atoms was explainedby two reactions: N + CH −−→ H CN + H, H CN + H −−→
HCN + H . Two scenarios were proposedfor the origin of the N-atoms: (1) infalling of ionized N-atoms transported from Neptune’s largestmoon, Triton, and (2) the upward transportation of N from the warm troposphere and subsequentdissociation to N-atoms by Galactic Cosmic Ray. The latter scenario suggests the importance of N transportation into the stratosphere by global circulation. The circulation has been inferred by theobservations of continuum emissions and tropospheric gases because the atmospheric transportationmay cause perturbations in the brightness temperature by adiabatic heating and cooling, and molecu-lar opacity variation (de Pater et al. 2014; Fletcher et al. 2014; Tollefson et al. 2019). Previous studiessuggested that the observational signatures of the dry south polar troposphere, cold mid-latitudinalstratosphere and warm equatorial stratosphere are caused by the downward transportation of dryair, adiabatic cooling induced by the upward transportation, and adiabatic heating by the downwardtransportation, respectively. The suggested upward branch in the meridional circulation could trans-port N from the troposphere into the stratosphere in the mid-latitude, and possibly produce HCNby photo-chemical reactions in the stratosphere.In turn, some observations supported an external origin scenario that cometary impact and influxof the Interplanetary Dust Particle supply volatiles to the stratosphere. Such a process is well knownfor Jupiter, where the collisions of comet Shoemaker-Levy/9 produced a large amount of volatilessuch as carbon monoxide (CO), HCN, carbon monosulfide (CS), and H O, as long-lived species inthe stratosphere (Lellouch et al. 1997; Moreno et al. 2001, 2003; Cavali´e et al. 2013; Iino et al. 2016).Among these species, on Neptune, CS and H O are particularly important probes of the externalvolatiles supply because such species cannot pass through the cold tropopause in the gas phase (theyare easily condensed in low-temperature environment). Some attempts have been made to detect CSon Neptune (Moreno 1998; Iino et al. 2014), and a recent Atacama Large Millimeter/submillimeterArray (ALMA) observation reported the first detection of CS ( J = 7–6) on Neptune with a 2.4–21.0 × − mixing ratio above the 0.5–0.03 mbar pressure level(Moreno et al. 2017). The detection ofCS on Neptune reinforces the evidence for a previous cometary impact that should have suppliedHCN, along with CS (and CO) at the same time as occurred on Jupiter.HCN is also a subject of research for the Saturn’s largest moon, Titan. HCN has been observedby in-situ, space- and ground-based observations. Those observations revealed a remarkable seasonalchange in Titan’s HCN distribution, in which the winter hemisphere has a larger abundance than thatof the summer hemisphere (Coustenis et al. 1989, 2005, 2010, 2016; Thelen et al. 2019). In particular,the Cassini spacecraft illustrated HCN enhancement at the winter pole before the summer solstice.HCN abundance measured at 75 ◦ S showed ∼ Iino et al. (Coustenis et al. 2016). The inhomogeneous distribution is attributed to effects of global circulationand photo-chemistry. Vinatier et al. (2015) successfully obtained the seasonal evolution of HCNvertical distribution by analyzing the number of Cassini observation data using the radiative transfermethod. They concluded that, in 2011, two years after the vernal equinox, a single north-to-south-pole cell appeared in the meridional circulation, and transported HCN-enriched air to the southpole.The above-described case of Titan is indicative that the spatially resolved observation of volatiles,in particular HCN, could give us a new clue on the atmospheric dynamics and chemistry of Neptune.The ALMA achieves high spatial resolution observation of solar system objects in millimeter andsubmillimeter wavelength, whereas previous observations using single-dish telescopes were able toobtain only the disk-averaged spectra of HCN. In this paper, we first report the spatial distributionof HCN in Neptune’s stratosphere obtained with ALMA. IMAGE SYNTHESIS OF ALMA ARCHIVED DATAWe analyzed an archived ALMA data of project ID 2015.1.01471.S (PI: R. Moreno) including theHCN ( J =4–3) rotational transition at 354.505 GHz, the same project that was used in Moreno et al.(2017). The observations were performed originally to search for isotopologues of major species suchas CO and HCN, and minor chemical species such as CS and CH CCH on Neptune.The observation was performed on 30 April 2016 (UTC) using 41 12–m antennae. At the observedtime, the apparent angular diameter of Neptune was 2. (cid:48)(cid:48)
24. Both the sub-observer and sub-solarlatitude were 26 ◦ S. The center frequency of the spectral window used for the analysis was 355.19004GHz. The total bandwidth of the spectral window and the channel spacing were 1875 and 0.977 MHz(the effective spectral resolution was ∼ ×
320 pixels with 0. (cid:48)(cid:48)
025 pixel spacing, ”natural”weighting, 0.1 mJy threshold, 0.1 gain and csclean mode. A circular CLEAN region that had adiameter slightly larger than Neptune was employed with the channel (width = 1) clean mode. Theachieved synthesized beam size was 0. (cid:48)(cid:48) × (cid:48)(cid:48)
39, which was fine enough to resolve the Neptune’s diskspatially. ANALYSIS OF SPATIAL DISTRIBUTION OF HCN EMISSIONTo illustrate the spatial distribution of HCN on Neptune’s disk, an integrated-intensity map thatintegrates the ±
30 MHz frequency range, which covers the entire HCN emission line, was produced.The map is shown in Figure 1(a) and exhibits a clear ring-like structure with a ∼ (cid:48)(cid:48)
95 radius, whichhas been also reported in an unpublished work using the Sub Millimeter Array (Moullet & Gurwell2011). Note that the ring structure is attributed to the increase of the line-of-sight path lengthin the Neptune atmosphere as the emission angle increases. Figure 1(b) shows the HCN intensitymeasured at the same emission angle along the 1. (cid:48)(cid:48)
05– and 0. (cid:48)(cid:48)
75– radius circle in Figure 1(a), whichcan exhibit latitudinal intensity variation without variation of the emission angle. Vertical andhorizontal error bars are r.m.s. noise level measured outside the disk and the latitude range includedin the synthesized beam, respectively. Eastern and western hemispheric intensities are averaged.For both selected circles, an interesting feature is that the greatest intensity peak is locating at theequator. The lowest intensity values of the both circles locate at 60 ◦ S. The peak intensity measuredat the equator is 25–30% higher than the lowest value. In addition, for the 1. (cid:48)(cid:48)
05 circle, a weakintensity peak is also found on the south pole. The intensity difference between 60 ◦ S and the southpole is ∼ ∼ (cid:48)(cid:48) corresponds to the ring structure shown in Figure 1(a). The radiallyaveraged profile was produced by the polynomial fitting method. The intensity ratio of measured andaveraged intensity is shown in Figure 2. While both the vertical and horizontal errors are omitteddue to the dense distribution of dots, corresponding errors are same as in Figure 1(b) and the sizeof the synthesized beam size, respectively. Iino et al.
To illustrate the latitudinal intensity distribution over the entire disk, an intensity ratio map ofmeasured intensity versus radially averaged intensity was produced (Figure 2 (a)). The same methodwas applied to derive the global continuum emission distribution on Neptune (Iino & Yamada 2018).On the equator, a belt-like HCN-rich region, which corresponds to the latitudinal peaks found in the1 (cid:48)(cid:48) .05 radius circle in Figure 1 (b), is clearly shown. At the southern mid-latitude of ∼ ◦ S, a lowintensity ratio region is found in the western hemisphere. A latitudinal profile of the intensity ratiois shown in Figure 2 (b). For a better visibility of the figure, every third pixel intensity was plotted.The derived structure is similar to that measured along the same emission angle as shown in Figure1 (b). The red curve represents the latitudinally averaged profile of the ratio. The equatorial peakshows a latitudinally symmetrical structure. In addition, a relatively weak peak is found at the southpole. Latitudinal errors corresponding to the synthesized beam size are represented at the bottomof Figure 2 (b). At the 60 ◦ S region, intensity ratio values can be divided into two groups, whichare likely to correspond to a dark spot located in the western hemisphere and other regions on 60 ◦ Sarc. Considering the self rotation period of Neptune of ∼
16 hours, Neptune rotates ∼ ◦ duringthe observation time. In turn, difference of intensity between bright and dark region on 60 ◦ S arc is ∼ ◦ S arc is marginal. RADIATIVE TRANSFER ANALYSISThe radiative transfer method was employed to estimate the latitudinal difference of HCN abun-dance between 60 ◦ S and the equatorial region by searching the best-fit spectrum. For the calculation,we employed the open-source software Planetary Spectrum Generator (PSG) (Villanueva et al. 2018).Because PSG has an online Application Program Interface that is easy to use, one can evaluate ourresult by reproducing the observed spectra.Pressure levels of modeled vertical atmospheric structure were ranged from 100 to 10 − bar with40 layers. For the temperature profile, we employed a disk-averaged result retrieved from (Fletcheret al. 2010). Gaseous H , He, CH , CO and HCN were considered as the atmospheric constituents.Their vertical abundance profile except for HCN were the initial set of PSG. (Moses 2005; Martenet al. 2005; Lellouch et al. 2010) Spectroscopic parameters are as of HITRAN database. BecausePSG can employ a horizontally symmetrical beam, a 0. (cid:48)(cid:48)
41 diameter beam was employed while thetrue beam shape is slightly elliptical. We attempted to obtain the HCN abundance at two points on60 ◦ S and the equator.As the vertical HCN profile, a constant molecular VMR above a specific pressure level, p , wasemployed in this study. We have tested some p values in the range from 0.5 to 2.0 mbar pressurelevel. As a result, we employed 1.0 mbar pressure level as p because the value could reproduce theobserved spectra to a relatively better extent.The best-fit spectrum for each point was identified within the VMR parameter space by the least-square method. The ∆ χ technique was used to obtain the error for the fitted parameter (Teanbyet al. 2006). The 1– σ significance level was used as the error value, corresponding to ∆ χ = 1.0.The derived VMR and errors for 60 ◦ S and the equatorial region were 1.17 ± +0 . − . ppb,respectively, above the 1.0 mbar pressure region. Thus, the equatorial region is determined to have40% higher abundance than that of 60 ◦ S region.Figure 3(a) and (b) shows a series of the measured and best-fit spectra. Residuals between theobserved and modeled spectra are also shown as red curves with an offset of -5 K. Consideringthe atmospheric structure, opacity of the HCN line core is derived to be ∼ ± DISCUSSIONThis new analysis of the spatially-resolved ALMA spectroscopic observation data for Neptuneindicates that Neptune’s stratospheric HCN has a horizontally non-uniform distribution, in which
Iino et al. the equator has a ∼
30% higher line intensity and ∼
40% higher abundance than that of the 60 ◦ Sregion. Possible scenarios that might explain the observed results are discussed from two point ofviews of the origin of HCN: internal and external sources.5.1.
Possible sources of HCN intensity gradient
Spatial variations in HCN line intensity can be caused both by spatial variations in the temperatureof the foreground stratosphere and the background troposphere, and differences in HCN abundance.The variation in stratospheric or tropospheric temperature is equivalent to that of the HCN intensitybecause the HCN line core opacity of ∼ Implication for global circulation
The spatial distribution of short-lived trace species is a powerful tool for investigating atmosphericdynamics. The obtained HCN latitudinal gradient shows a morphological similarity with that ob-served on Titan, where the global single-cell circulation possibly produces a latitudinal abundancedifference in which the winter and summer hemispheres exhibit the highest and lowest peaks ofHCN, respectively (Coustenis et al. 1989, 2005, 2010, 2016; Thelen et al. 2019). On Titan, similar tothe terrestrial Brewer–Dobson circulation, a global single cell in the meridional circulation is likelyto transport N-bearing species depleted air parcel from summer to winter hemisphere. Various N-bearing species are being produced during the horizontal transportation, and accumulated on thewinter pole where the subsiding transportation is present. In addition, the result obtained here issimilar to that for Earth’s stratospheric ozone (e.g. Wayne (2000)), which is also caused by a summerto winter single cell meridional circulation.On Neptune, various observation techniques have already been used to propose the presence of aglobal tropospheric and stratospheric circulation. A re-analysis of Voyager/Infrared InterferometerSpectrometer and Radiometer spectra, Fletcher et al. (2014) concluded that a global circulation thatcold air rises in the mid-latitude and subsides both on the equator and the poles. A warm troposphericequator and pole are likely to be produced by the adiabatic heating induced by the subsiding air.Similarly, de Pater et al. (2014) suggested that upward transportation at mid-latitude, ∼ ◦ S, createsa belt-like structure of the tropospheric cloud that is caused by the adiabatic cooling of the risingair parcel. A recent ALMA continuum observation expects that mid-latitude upward transportationis present at relatively lower latitude region, ∼ ◦ S (Tollefson et al. 2019). From a morphologicalpoint of view, our obtained HCN distribution map can be connected to the previous observationsof the global circulation as follows: the high abundances at the equator and south pole are likelyto be due to the accumulation of HCN produced during the horizontal transportation in the samemanner as the mechanism of Titan’s HCN distribution. Considering the troposphere–stratospherecirculation, N transported to the stratosphere dissociates into N-atom by the photolysis, and leadsto HCN production via H CN, as mentioned in the Introduction section. HCN enhancements at boththe equator and south pole are quite consistent with a two-cell circulation model for the southernhemisphere.In turn, the HCN-depleted region observed at 60 ◦ S disagrees with the location of the upward branchof the two-cell circulation model (the circulation model shows the upward transportation at 40 ◦ S).Although the reason for this remains for further study, it should be noted that our study probesthe upper stratosphere while the previously suggested two-cell meridional circulation model has beenmainly based on observations of the upper troposphere. An effective use of HCN maps may bringan additional constraint to the atmospheric circulation at higher altitudes. In addition, because of0
Iino et al. the coarse resolution of this analysis, 40 ◦ S and 60 ◦ S region were not resolved clearly. Thus, a newobservation with a higher spatial resolution enables us to determine the latitude of the HCN-depletedbelt. 5.3.
External source model
A large cometary impact is also a possible cause of the observed HCN distribution. Strong evidencefor such an impact was previously provided by the presence of CS (Moreno et al. 2017) and by theCO-rich upper stratosphere (Lellouch et al. 2005; Hesman et al. 2007; Fletcher et al. 2010). Becauseno S-bearing species can be supplied from the troposphere, CS is considered to have only a cometaryorigin. Thus, it is also possible for HCN to be supplied by a past cometary impact; likely an impactat the equator. To evaluate the cometary impact hypothesis, a new analysis of the latitudinaldistribution of CS is crucial. If CS shows the equatorial enhancement seen in HCN, the impacthypothesis is strongly supported. It is noted that our observation shows HCN enhancement at thesouth pole as well. Such a latitudinal distribution does not fit well with a simple meridional diffusionof HCN from a single collision. 5.4.
Future perspectives
This study presented that a spatially resolved HCN observation has significant potential for provid-ing various information for discerning the atmospheric circulation and/or a past cometary impact.As mentioned in Section 5.3, highly sensitive observation of CS to illustrate its spatial distribution,along with the determination of the 3-D distribution of CO, would strongly constrain the cometaryimpact scenario. Also the spatial distribution of H O, which is not observable by ALMA, is crucialto evaluate the scenario as for the case for Jupiter (Cavali´e et al. 2013). In addition, precise ob-servations to determine the N/ N isotopic ratio in HCN may constrain its source and productionprocess. For example, on Jupiter, a large nitrogen isotopic fractionation in HCN (4.3-16.7 timeshigher than the typical solar system value) was detected (Matthews et al. 2002). This fractionationmay be caused by the thermo-chemical processed induced by the cometary collision. In addition,on Titan, a different isotopic fractionation between N and its daughter species, HCN, is known1(Hidayat et al. 1997; Gurwell 2004; Molter et al. 2016). A theoretical chemical model suggests thatthe nitrogen-bearing species is fractionated in a different extent according to different dissociationprocesses of N (Dobrijevic & Loison 2018). These applications to other planets and satellites leadone to expect that a new determination of the nitrogen isotopic ratio in Neptune HCN will providenew implication on its origin.This paper makes use of the following ALMA data: ADS/JAO.ALMA Software:
Astropy (Robitaille et al. 2013)
Software:
PSG (Villanueva et al. 2018)2
Iino et al.
RA (J2000) D e c ( J ) (a) Integrated intensity map I n t e g r a t e d i n t e n s i t y [ J y / b e a m k m / s ]
80 60 40 20 0 20 40Latitude [degree]2.502.753.003.253.503.754.004.25 I n t e g r a t e d i n t e n s i t y [ J y / b e a m . k m / s ] (b) Latitudinal HCN line intensities I n t e g r a t e d i n t e n s i t y [ J y / b e a m . k m / s ] (c) Radial intensity profile Figure 1. (a) Integrated intensity map of HCN ( J =4–3) for Neptune. The white ellipse at the bottom leftillustrates the shape of the synthesized beam. (b) HCN intensity profiles along the 1 (cid:48)(cid:48) .05 and 0 (cid:48)(cid:48) .75 radiuscircle indicated by white dashed lines in (a). (c) Radial HCN intensity profile (blue dots) and averagedprofile (red curve) . Horizontal and vertical error bars are the same as the synthesized beam size, ∼ (cid:48)(cid:48) .2 and0.2 Jy/beam km/s, respectively. RA (J2000) D e c ( J ) (a) Intensity ratio map I n t e n s i t y r a t i o
80 60 40 20 0 20 40 60Latitude [degree]0.80.91.01.11.2 I n t e n s i t y r a t i o (b) Latitudinal intensity ratio profile Figure 2. (a) Intensity ratio (measured value versus radially averaged profile) map of HCN. The white ellipseat the bottom left illustrates the shape of the synthesized beam. (b) Latitudinal profile of the intensity ratio.Blue dots and red curve are for each pixel and the latitudinally averaged profile, respectively. Horizontalerror bars at the bottom are latitudinal error measured on the central meridian line. B r i g h t n e ss T e m p e r a t u r e [ K ] Latitude: -60.29 (a) B r i g h t n e ss T e m p e r a t u r e [ K ] Latitude: 0.87 (b)
Figure 3.
Observed (blue) and modeled (red) HCN spectra measured and derived at 60 ◦ S(a) and theequator(b) on the central meridian line. Residual is plotted with separation of -5 K in red curve. Iino et al.
REFERENCES { % } { } DPS/dps2011/a { } dps2011program+abstracts/pdf/EPSC-DPS2011-1153-3.pdfhttp://yly-mac.gps.caltech.edu/A { }