Open Access
Issue
A&A
Volume 677, September 2023
Article Number A157
Number of page(s) 17
Section Planets and planetary systems
DOI https://doi.org/10.1051/0004-6361/202346402
Published online 22 September 2023

© The Authors 2023

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication.

1 Introduction

Comets are ice-rich, kilometer-sized bodies composed of volatiles, refractories, and semi-refractory compounds. The key characteristic of these minor bodies of our Solar System is their outgassing activity, especially upon their approach toward the Sun. During the approach, the volatile ices in a cometary nucleus are released as gases into a tenuous coma and dust particles are lifted off of the surface through several mechanisms (Jewitt & Hsieh 2022). Based on their physical properties, it remains unclear if such bodies formed in a protoplanetary disk through hierarchical agglomeration of cometesimals or through pebble accretion (Weissman et al. 2020; Blum et al. 2022). The chemical composition of volatiles does support partial inheritance of ices from the earliest stages of star formation to cometary nuclei (Bockelée-Morvan et al. 2000; Drozdovskaya et al. 2019; Altwegg et al. 2019).

The dominant volatile constituent of comets is water (H2O; Mumma & Charnley 2011). A typical, although not only, method for remote sensing of cometary water is through observations of OH (Schloerb & Gérard 1985), which is a primary product of H2O photodissociation in the coma. OH lines are available in the UV (Feldman et al. 2004; Bodewits et al. 2022), radio (Crovisier et al. 2002a,b), and infrared (IR; Bonev et al. 2006; Bonev & Mumma 2006) regimes. The production rate of H2O is a fundamental parameter of a comet that must be constrained for each comet individually and over the longest-possible period of time in order to get a handle on the individual time-dependent cometary activity. In turn, this facilitates comparisons of chemical compositions from comet to comet relative to the dominant volatile species, H2O.

The nitrogen budget of comets has been a long-standing mystery, because even summing all the volatile N measured in the coma with the refractory N measured in the dust leaves a deficiency in N relative to the solar N/C ratio (Geiss 1988; Rubin et al. 2015; Biver et al. 2022b). The recent discovery of ammonium salts (NH4+X${\rm{NH}}_4^ + {{\rm{X}}^ - }$) on the dust grains of comet 67P/Churyumov-Gerasimenko (hereafter, 67P/C-G) is thought to have uncovered the missing reservoir of nitrogen in comets (Poch et al. 2020; Altwegg et al. 2020). Their presence was already hypothesized on the basis of data for comet 1P/Halley (Wyckoff et al. 1991) and the salt NH4+CN${\rm{NH}}_4^ + {\rm{C}}{{\rm{N}}^ - }$ was suggested to be the precursor for NH3 and HCN release in more recent comets (Mumma et al. 2017, 2018, 2019). Ammonium salts are considered to be semi-volatile with sublimation temperatures in the 160–230 K range (Bossa et al. 2008; Danger et al. 2011). In the laboratory, it was demonstrated that ammonia (NH3) is a key product of thermal desorption of ammonium salts that decompose during this process (Hänni et al. 2019; Kruczkiewicz et al. 2021). Consequently, it is expected that as the comet's heliocentric distance decreases, its ammonium salts desorb and decompose, which in turns leads to an increasing NH3/H2O ratio. This trend was demonstrated on the basis of high-resolution IR spectroscopic observations of 30 comets (Fig. 7b of Dello Russo et al. 2016). To an extent, this is also seen in the NH3/H2O ratio measured with the Rosetta Orbiter Spectrometer for Ion and Neutral Analysis (ROSINA; Balsiger et al. 2007) instrument aboard the ESA Rosetta spacecraft during its 2-yr long monitoring of 67P/C-G, although there is quite a bit of scatter in the ratio that is uncor-related with the latitude and season (Fig. 1a of Altwegg et al. 2020).

Ammonium salts have also been detected on the surface of the dwarf planet Ceres with the Dawn mission (De Sanctis et al. 2015, 2019), but they have not been detected in interstellar ices thus far (Boogert et al. 2015). However, their presence has long been hypothesized (Lewis & Prinn 1980) and expected based on a suite of laboratory experiments involving the UV irradiation or electron bombardment of NH3-containing ices (Grim et al. 1989; Bernstein et al. 1995; Muñoz Caro & Schutte 2003; Gerakines et al. 2004; van Broekhuizen et al. 2004; Bertin et al. 2009; Vinogradoff et al. 2011). It has also been demonstrated in the laboratory that acid-base grain-surface reactions between NH3 and HNCO, NH3 and HCN, and NH3 and HCOOH may form the ammonium cyanate (NH4+OCN${\rm{NH}}_4^ + {\rm{OC}}{{\rm{N}}^ - }$), ammonium formate (NH4+HCOO${\rm{NH}}_4^ + {\rm{HCO}}{{\rm{O}}^ - }$), and ammonium cyanide (NH4+CN${\rm{NH}}_4^ + {\rm{C}}{{\rm{N}}^ - }$) salts, respectively, even at prestellar core temperatures of ~10 K without energetic processing and increase in efficiency with elevated temperatures (Schutte et al. 1999; Raunier et al. 2003, 2004; Gálvez et al. 2010; Mispelaer et al. 2012; Noble et al. 2013; Bergner et al. 2016). At a warmer temperature of 80 K, NH3 and CO2 have also been experimentally shown to react to produce ammonium carbamate (NH4+H2NCOO${\rm{NH}}_4^ + {{\rm{H}}_2}{\rm{NCO}}{{\rm{O}}^ - }$; Bossa et al. 2008).

Detections of the NH3 hyperfine inversion lines, which arise from the oscillation of the nitrogen atom through the plane formed by the three hydrogens, have been challenging to secure on a consistent basis in comets. A possible detection of the JK = 33 − 33 inversion line at 23.870 GHz was made in comet C/1983 H1 (IRAS-Araki-Alcock) with the Effelsberg 100-m telescope (Altenhoff et al. 1983). A sensitive search for the same line with the same instrument in comets 1P/Halley and 21P/Giacobini-Zinner was negative (Bird et al. 1987), which was also the case for the earlier comet C/1973 E1 (Kohoutek), (Churchwell et al. 1976). A confirmed detection of the JK = 11 − 11 and 33 − 33 inversion lines with high signal-to-noise ratios was obtained in C/1996 B2 (Hyakutake) with the 43-m telescope of the National Radio Astronomy Observatory (NRAO) in Green Bank (Palmer et al. 1996). In comet C/1995 O1 (Hale-Bopp), all inversion transitions from JK = 11 − 11 to 55 − 55 were detected with the Effelsberg 100-m telescope (Bird et al. 1997). Some of these were also observed with the NRAO 43-m telescope (Butler et al. 2002) and the 45-m telescope of the Nobeyama Radio Observatory (Hirota et al. 1999). Despite a sensitive search at the Effelsberg 100-m telescope, NH3 was not detected in comets C/2001 A2 (LINEAR), C/2001 Q4 (NEAT), and C/2002 T7 (LINEAR), and only marginally detected in comet 153P/Ikeya-Zhang (Bird et al. 2002; Hatchell et al. 2005). Subsequent rotational lines, which are stronger, lie at submillimeter wavelengths. NH3 was detected in several comets (Table B.1) using the Odin and Herschel satellites through its JK = 10 − 00 transition at 572 GHz that is only accessible from space (Biver et al. 2007, 2012). This transition was also monitored in comet 67P/C-G with the Microwave Instrument for the Rosetta Orbiter (MIRO) instrument aboard Rosetta (Gulkis et al. 2007; Biver et al. 2019). At shorter IR wavelengths, NH3 has been observed in many comets (Table B.1) through its vibrational lines near 3 µm, although its unblended lines are sparse and weak in comparison to other species typically observed in the near-IR (Dello Russo et al. 2016, 2022; Bonev et al. 2021 and the many references therein).

In this paper, we present radio observations of comet C/2020 F3 (NEOWISE), hereafter F3 for simplicity, taken at the Nançay Radio Telescope (NRT) and the Green Bank Telescope (GBT) that targeted OH and NH3 lines (Table 1). The long-period comet F3 originating from the Oort Cloud was a bright naked-eye object in the sky that underwent perihelion on 3 July 2020 and perigee on 22–23 July 2020. A more detailed description of the comet is given in Sect. 2.1. The OH emission from the comet was monitored nearly continuously for 2 months at the NRT pre- and post-perihelion. GBT observations targeted contemporaneously OH and NH3 lines post-perihelion on 24 July 2020. Additional OH data were collected at the GBT on 11 August 2020. The GBT observational strategy and data reduction are presented in Sect. 2.2. The NRT observations are described in Sect. 2.3. The methods behind the determination of the OH production rate are described in Sect. 3.1.1. The analysis of OH line profiles in the context of outflow velocities is in Sect. 3.1.2. GBT and NRT OH observations taken on the same date, a mere 2 h apart, are used to constrain the OH quenching radius in Sect. 3.1.3 based directly on observational data, which is critical for an accurate determination of the OH production rate. The GBT OH observations of 11 August 2020 are analyzed in the context of the Greenstein effect in Sect. 3.1.4. The OH and H2O production rates are finally computed in Sect. 3.1.5. GBT NH3 observations are analyzed in Sect. 3.2 and discussed in the context of the NH3/H2O ratio as a function of heliocentric distance in Sect. 4. The conclusions of the work can be found in Sect. 5.

2 Observations of comet C/2020 F3 (NEOWISE)

2.1 Comet C/2020 F3 (NEOWISE)

Comet F3 is considered to be the brightest comet in the northern hemisphere since comet C/1995 O1 (Hale-Bopp) in 1997 (Fig. 1). It was a bright naked-eye target in midsummer of 2020 and termed by some as the “Great Comet of 2020”. Comet F3 showed a huge (~105 km), curving, banded dust tail and a much fainter, straight, blue ion tail (~106 km). It was viewed from the International Space Station (ISS)1 and photographed by countless members of the public2. F3 is a long-period comet originating from the Oort Cloud with an orbital period of ~4500 yr and a near-parabolic (eccentricity of 0.999) trajectory. It was discovered on March 27th, 2020 by the Near Earth Object Wide-field Infrared Survey Explorer (NEOWISE) mission of the Wide-field Infrared Survey Explorer spacecraft (Mainzer et al. 2011)3. Perihelion took place on July 3rd, 2020 at 0.295 au from the Sun and perigee occurred in the night from July 22nd to 23rd, 2020 at 0.692 au from the Earth. Its nucleus was estimated to be approximately 5 km in diameter (Bauer et al. 2020) with a rotation period of 7.8 ± 0.2 h (Manzini et al. 2021). Sudden short-lived outbursts have not been reported for comet F3 and none were observed during the extended (from May to September 2020) monitoring campaign of the comet by the Solar Wind ANisotropies (SWAN) camera on the Solar and Heliospheric Observatory (SOHO) spacecraft (Combi et al. 2021). However, strong jet activity is seen in the HST images taken on 8 August 2020 (Program ID: 16418, P.I.: Qicheng Zhang)4. There were also no signs of disintegration pre-perihelion based on NASA/ESA’s Solar and Heliospheric Observatory (SOHO) LASCO C3 coronagraph observations (Knight & Battams 2020).

Comet F3 was observed by several facilities in the optical regime. Observations with the High Accuracy Radial velocity Planet Searcher for the Northern hemisphere (HARPS-N) echelle spectrograph on the 360-cm Telescopio Nazionale Galileo (TNG) with a high resolving power (R) of 115 000 revealed detections of C2, C3, CN, CH, NH2, Na, and [O I] (taken on 26 July and 5 August 2020; Cambianica et al. 2021). Lines of C2, NH2, Na, and [O I] were also detected with low resolution, R = 7400-9300, observations made with the Échelle spectrograph FLECHAS at the 90-cm telescope of the University Observatory Jena taken on 21, 23, 29, 30, and 31 July 2020 (Bischoff & Mugrauer 2021; Na was detected only on the first two dates). Lines of C2, C3, CN, CH, NH2, Na, and [O I] were also detected with the Tull Coudé Spectrograph at R = 60 000 on the 2.7-m Harlan J. Smith Telescope at McDonald Observatory on 24 July and 11 August 2020 (Cochran et al. 2020). Lines of C2, C3, CN, NH2, Na, [O I], K, and H2O+ were also detected with ultra-high resolution spectroscopy at R = 140 000 with the EXtreme PREcision Spectrograph (EXPRES) on the 4.3-m Lowell Discovery Telescope on 15 and 16 July 2020 (Ye et al. 2020). Further chemical characterization of comet F3 was executed in the IR. In the 1.1–5.3 µm range, data were taken with the long-slit near-IR high-resolution (R = 35 000–70 000) immersion echelle spectrograph iSHELL at the NASA/IRTF facility. The 0.95–5.5 µm range was observed with NIRSPEC 2.0 (R = 25 000–37 500) at the Keck Observatory. These near-IR observations of Faggi et al. (2021) led to the detection of 9 primary volatiles (H2O, HCN, NH3, CO, C2H2, C2H6, CH4, CH3OH, and H2CO) and 3 product species (CN, NH2, OH*) on several dates in July and August 2020. In the radio, comet F3 was observed with the Institut de Radio Astronomie Millimétrique (IRAM) 30-m and the NOrthern Extended Millimeter Array (NOEMA) telescopes on several dates in July and August with secured detections of HCN, HNC, CH3OH, CS, H2CO, CH3CN, H2S, and CO (Biver et al. 2022a).

Table 1

Targeted lines.

thumbnail Fig. 1

Comet C/2020 F3 (NEOWISE) as seen from the Blue Ridge Mountains outside of White Hall, Virginia on 20 July 2020. The image is an aligned and stacked average of five, 5-min exposures taken using a Nikon D750 DLSR camera with a 50 mm f /1.8 lens held on an equatorial tracking mount (B. A. McGuire). The image is approximately 29.3° × 22.3° in total. The 1° scale is indicated in the lower right corner, which corresponds to 3 × 104 km at the geocentric distance of the comet on this date. A finder chart for this image is available at https://nova.astrometry.net/user_images/7997658# (Lang et al. 2010).

2.2 Green Bank Telescope observations

Comet F3 was observed with the Robert C. Byrd 100-m GBT in Green Bank, West Virginia under the project code GBT20A-587. The Director s Discretionary Time (DDT) proposal was awarded a total amount of 12 h, which was to be executed in two 6 h sessions with a separation of at least 3 days to probe dependencies with heliocentric distance. It was planned for each session to begin with 1 h of OH observations, followed immediately by 5 h of NH3 observations (including 5 and 40 min of overheads, respectively). The targeted lines are tabulated in Table 1. Due to an unfortunate overlap of the comet’s perihelion on 3 July 2020 with scheduled maintenance, both sessions were observed post-perihelion.

2.2.1 Day 1 and 2 OH observations

OH observations were carried out with the L-band receiver coupled with the VErsatile GBT Astronomical Spectrometer (VEGAS) backend (Roshi et al. 2011). For the first session, on day 1 (24.71 July 2020), mode 8 was used with a 100 MHz bandwidth, 65 536 channels, and a spectral resolution of 1.5 kHz (0.3 km s−1 at 1666.0 MHz). These were single beam, dual polarization observations carried out with in-band frequency switching with a throw of 25 MHz (centered on the rest frequency of 1666.0 MHz). For the second session, on day 2 (11.88 August 2020), the signal was routed to four banks of VEGAS in four different modes: mode 8 (same as day 1), mode 9 (100 MHz bandwidth, 131 072 channels, 0.8 kHz or 0.14 km s−1 spectral resolution), mode 11 (23.44 MHz bandwidth, 65 536 channels, 0.4 kHz or 0.07 km s−1 spectral resolution), and mode 12 (23.44 MHz bandwidth, 131 072 channels, 0.2 kHz or 0.04 km s−1 spectral resolution). At the start of observing on both days, pointing and focus were performed using the Digital Continuum Receiver (DCR) backend toward the calibrator source 1011+4628 across a 80 MHz bandwidth. The half-power beam width (HPBW) of the GBT at 1.67 GHz is (7.55 ± 0.083)′ (according to equation 2 of the Green Bank Observatory (GBO) Proposer’s Guide for the GBT of 3 January 2023).

Day 1 OH observations were executed starting 24.71 July 2020. The total integration time on-source was 46 min (23 scans, 2 min each). Day 2 observations were taken starting 11.88 August 2020. Due to the separation between the two sessions being 17 days (much longer than the initially envisioned separation of not much more than 3 days) and the non-detection of NH3 at the more favorable smaller heliocentric distance of Day 1 (Sect. 2.2.2), the remaining allocated time was dedicated solely to OH observations. However, this time, the observations were performed toward three different locations: on-comet, an offset toward the Sun, and an offset away from the Sun. The integration time was 1 h 26 min (43 scans, 2 min each) on-comet, 1 h 38 min (49 scans, 2 min each) toward-offset, and 1 h 58 min (59 scans, 2 min each) away-offset. Total on-source time was 5 h 2 min. It was envisioned to probe one-beam offsets from the on-comet position along the Sun-comet trajectory to accurately quantify the quenching radius of OH (analogously to the methods of Colom et al. 1999 for comet C/1995 O1 (Hale-Bopp)). However, due to an error in offset coordinates in the observing script, the probed positions were ~8 beams away in RA to the northeast (and < 1/5 of a beam away in Dec to the southwest). Specifically, the toward-offset was RA, Dec (J2000) of −7m30s, + 1′20″ and the away-offset was RA, Dec (J2000) of +7m30s, −1′20″.

The OH 1665 and 1667 MHz spectra obtained on Day 1 (24.71 July 2020) at the GBT are shown in Fig. 2. The OH 1667 MHz spectrum obtained on Day 2 (11.88 August 2020) at the GBT is shown in Fig. 3. L-band Day 1 data have been checked for the presence of the 18OH line at 1639.5 MHz (the strongest of three 18OH lines in range), but it was not detected. The observed frequency range also covered one line of 17OH, but based on neither the isotopic ratio of 18O/17O = 3.6 in the local interstellar medium (ISM; Wilson 1999) nor the measured 18O/17O = 5.3 ± 0.4 in water of comet 67P/C-G (Schroeder et al. 2019; Müller et al. 2022), it is not expected to be detected if 18OH is not. The main beam efficiencies (ηmB) given in Table 2 were used to convert the line areas into the main beam brightness temperature scale through TmB=TA*/ηmB${{{T_{{\rm{mB}}}} = T_{\rm{A}}^*} \mathord{\left/ {\vphantom {{{T_{{\rm{mB}}}} = T_{\rm{A}}^*} {{\eta _{{\rm{mB}}}}}}} \right. \kern-\nulldelimiterspace} {{\eta _{{\rm{mB}}}}}}$. Table 3 gives line areas in TA*$T_{\rm{A}}^*$ scale for the OH lines observed at the GBT. The ratio of the line areas of the 1667 and 1665 MHz OH lines observed on 24 July 2020 is 1.93 ± 0.07, which is consistent, within 2σ with the statistical ratio of 1 .8.

Table 2

GBT and NRT half-power beam widths (HPBWs) and efficiencies.

2.2.2 Day 1 NH3 observations

NH3 observations were carried out on Day 1 (starting 24.81 July 2020). The 7-pixel K-band Focal Plane Array (KFPA) receiver (Masters et al. 2011) was employed with the VEGAS backend. The total integration time on-source was 3 h 34 min (107 scans, 2 min each). Mode 24 was used with a 23.44 MHz bandwidth, 65 536 channels, and a spectral resolution of 0.4 kHz (0.005 km s−1 at 23.8 GHz). All seven beams of the KFPA were used and routed to six banks of VEGAS (centered on rest frequencies of 23 694.50, 23 722.63, 23 870.13, 24 139.42, 24 532.99, 23 963.9 MHz). These were dual polarization observations carried out with in-band frequency switching with a throw of 5 MHz. At the start of observing, pointing and focus were performed using a single spectral window (still with the VEGAS backend) toward the calibrator source 1033+4116 across a 1500 MHz bandwidth (centered on the rest frequency of 24 010 MHz). The HPBW of the GBT at 23.8 GHz is (0.53 ± 0.005)’ (Table 2). The data from all seven beams were inspected for the presence of NH3 lines individually; however, solely the centermost on-comet beam has been used for the derivation of the upper limits. Table 3 gives the upper limits in TA*$T_{\rm{A}}^*$ scale on the NH3 lines observed at the GBT.

2.2.3 GBT data reduction

Initial data processing and calibration were performed using GBTIDL5. Day 1 OH and NH3 observations were visually inspected on a scan-by-scan basis. All scans were considered for the final data products with folding performed using the getfs routine. The final spectra are noise-weighted averages of all the scans and both polarizations for the case of L-band observations. For the KFPA data, only the right polarization has been considered given a known and reported issue with the left polarization at the time these observations were taken. Day 2 OH observations suffered from much more severe Radio Frequency Interference (RFI). Furthermore, a mishap occurred for modes 11 and 12, because the adopted 25 MHz throw exceeded the bandwidth, resulting in there being no overlap between the sig and ref phases. All 151 scans were visually inspected in the sig and ref phases (for modes 8 and 9; and only in the sig phase for modes 11 and 12) separately in both polarizations in small regions near the two OH lines. Scans with severe RFI or odd baselines (even if in just one polarization of either of the phases) were discarded. All scans in the ref phase of modes 8 and 9 exhibit an RFI close to (1667.36 − 25.00) = 1642.36 MHz, which after folding results in an absorption artifact close to the 1667. 36 MHz OH line. Consequently, all data in the ref phase from all 4 modes were discarded in the analysis of the 1667. 36 MHz OH line. As a result, calibration of the sig phase had to be done manually on the basis of a line-and RFI-free region, and the application of the noise diode in K factor. With this methodology, any remaining absorption feature must be a real signal rather than an artifact of folding. In the analysis of the 1665. 40 MHz OH line, both sig and ref phases were considered from modes 8 and 9 with folding using the getfs routine, but only the sig phase from modes 11 and 12 was considered with the manual calibration procedure. Finally, for every mode a noise-weighted average was produced. These spectra were then resampled to the coarsest spectral resolution of mode 8 (1 . 5 kHz). The resampling has been performed with the Flux Conserving Resampler from the specutils Python package, which conserves the flux during the resampling process (Carnall 2017)6. The final spectrum per OH line per position is a noise-weighted average of the four modes.

For the Day 1 OH data, baseline subtraction was performed (separately for each line) on a scan-by-scan basis with a first-order polynomial in a region centered on the rest-frequency of the specific OH line, which corresponds to ±54 km s−1 (±300 kHz or ±200 channels), while excluding the central ±2 km s−1 (±11 kHz or ±8 channels) range that includes the line. For the NH3 data, baseline subtraction was performed (separately for each line) on a scan-by-scan basis with a first-order polynomial in a region centered on the rest-frequency of the specific NH3 line, which corresponds to ±8 km s−1 (±633 kHz or ±1766 channels), while excluding the central ±2 km s−1 (±158 kHz or ±442 channels) range that includes the line. For the Day 2 OH data, baseline subtraction was performed on a scan-by-scan basis with a first-order polynomial in a 0.54–0.76 (depending on the mode) and 0.2 MHz wide region near the rest-frequencies of the 1665.40 and 1667.36 MHz OH lines, respectively. The baseline subtraction was performed solely on the sig phase for the cases when it was the only phase used for the final product (all modes for the 1667.36 MHz line, and modes 11 and 12 for the 1665.40 MHz line). The observed spectra were corrected by the comet’s velocity on a scan-by-scan basis based on computations (with a step size of 1 min) from NASA/JPL Horizons On-Line Ephemeris System (EOP files eop.210205.p210429 and eop.210211.p210505)7.

The system temperature (Tsys) on Day 1 in the L-band was in the 16.76-19.41 K range with a mean of 18.01 K. Tsys on Day 2 in the L-band was in the 11.14–25.79, 8.77-27.91, and 17.36– 28.14 K ranges with means of 20.37, 20.21, and 21.65 K at the on-comet, toward-offset, and away-offset positions, respectively.

Tsys on Day 1 for the KFPA was in the 9.55 – 88.29 K range with a mean of 76.43 K (in the central beam of the KFPA), which was on the higher end of expectations. The anticipated RMS based on the GBT Sensitivity Calculator was estimated to be in the 2.7–5.4 mK km s−1 range for Tsys ∈ [50 – 100] K in the proposal for the KFPA for a line width of 3 km s−1. However, the attained RMS of the observations came out to be 46–48 mK channel−1 (in the central beam of the KFPA), which corresponds to 5.6–5.9 mK km s−1 for a line width of 3 km s−1 (using RMS(mK km s−1) = n$\sqrt n $ × RMS (mK channel−1) × dυ, where dυ is the channel width and n is the number of channels spanning the line). Other KFPA beams had comparable Tsys and RMS values, except for the sixth beam, which was a factor of 1 .7 higher. The somewhat higher attained RMS in comparison to the anticipated RMS from the GBT Sensitivity Calculator stems from the actual on-source time being 46 min shorter than planned and the Tsys values being in the higher range of expectations. All spectra are in terms of the antenna temperature corrected for antenna and atmospheric losses, TA*$T_{\rm{A}}^*$ (Ulich & Haas 1976). In the K-band, the calibration uncertainty is generally ~30% for the GBT (McGuire et al. 2020; Sita et al. 2022).

thumbnail Fig. 2

OH 1665 (left) and 1667 MHz (right) lines observed with the GBT (top) and the NRT (bottom) on 24 July 2020 in solid black lines. The red dashed curves show the fitted trapeziums to the spectra based on the methodology of Bockelée-Morvan et al. (1990). For GBT (NRT), the half lower bases of the trapeziums are 2.32 ± 0.07 (2.30 ± 0.14) and 2.43 ± 0.03 (2.23 ± 0.11) km s−1 for the 1665 and 1667 MHz lines, respectively.

Table 3

GBT observations: line intensities and spectral characteristics.

thumbnail Fig. 3

OH 1667 MHz line observed with the GBT on 11.88 August 2020.

2.3 Nançay Radio Telescope observations

Comet F3 was scheduled at the NRT as a Target of Opportunity (ToO) for observations beginning 1 June 2020 and continuing until 27 July 2020. It was observed almost every day during this period, except for 4–17 July 2020 (Table 4). The instrumental characteristics, observing protocol, and data reduction procedure are the same as those used in preceding cometary observations with the NRT. These are described in Crovisier et al. (2002a,b). The NRT is a Meridian telescope, which can observe a given source for ~1 h. Its RA×Dec beam size is 3.5′ × 19′ and its sensitivity is 0.85 K Jy−1 at the 18-cm wavelength. The spectrometer, which can accommodate 8 banks with each having a 195 kHz bandwidth and 1024 channels, was aimed at the four OH transitions at 1665 and 1667 MHz (main lines) and 1612 and 1721 MHz (satellite lines) in, both, left- and right-hand circular polarizations with a 0.4 kHz (0.07 km s−1 at 1666.0 MHz) spectral resolution after Hanning smoothing.

Figure 4 shows the NRT spectra averaged over a few days that were used for kinematic studies based on the line shapes. The depicted spectra are averages of both polarizations and weighted averages of the 1665 and 1667 MHz main lines, converted to the 1667 MHz intensity scale by assuming the statistical ratio of 1.8. The satellite 1612 MHz line is clearly detected in the NRT data averaged over the 18–27 July observing window. The spectrum containing the 1720 MHz satellite line is noisier; and the detection is only marginal. The four OH 18-cm lines observed in comet F3 with the NRT averaged from 18 to 27 July 2020 are shown in Fig. A.1. The expected statistical relative intensities of the 1667 : 1665 : 1612 : 1720 MHz lines are 9:5:1:1, which is in good agreement with the observations. The satellite OH lines were tentatively detected at the 1σ level on 31.8 July 2020 with the ~3′ circular beam (at 1667 MHz) of the Arecibo Telescope (Smith et al. 2021).

The NRT spectra of 24.61 July 2020 were recorded nearly simultaneously with the Day 1 GBT observation and are analyzed individually in Sect. 3.1.3 (sixth line of Table 4). The 1665 and 1667 MHz spectra observed on this date are shown in Fig. 2. The measured 1667/1665 MHz line integrated intensity ratio is 1.82 ± 0.13, which is consistent within uncertainties with the statistical ratio of 1 .8. The other daily spectra are not shown individually. However, in Table 4, daily measurements for July are provided, alongside averages over a few days in June and July. For the dates in June, it was not possible to compute the OH production rate on a daily basis due to the relative weakness of the OH lines (due to a larger geocentric distance). Of particular importance is the entry for 2 July 2020 with its high maser inversion (i < −0.31), which allowed the OH production rate near perihelion to be estimated (Sect. 3.1.5).

Table 4

OH observations at NRT: line intensities and spectral characteristics.

3 Results

3.1 OH data analysis

3.1.1 Model for OH production rate determination

In comets, the excitation of OH through UV pumping and subsequent fluorescence leads to population inversion (or anti-inversion) in the sublevels of the Λ-doublet of the 2Π3/2 J = 3/2 ground-state. The population inversion (i) depends strongly on the comet heliocentric velocity and has been modeled by Despois et al. (1981) and Schleicher & A’Hearn (1988). Depending on the sign of i, the four hyperfine components of the Λ-doublet appear in either emission or absorption. The velocity-integrated flux density (i.e., line area) can be expressed as (e.g., Despois et al. 1981; Schloerb & Gérard 1985):

Svdυ=AulkcT bg4πΔ22Fu+18[ i+i+12hvkTbg ]ΓOH.$\int {{S_{\,v}}{\rm{d}}\upsilon } = {{{A_{{\rm{ul}}}}kc{T_{\,{\rm{bg}}}}} \over {4\pi {{\rm{\Delta }}^2}}}{{2{F_u} + 1} \over 8}\left[ {i + {{i + 1} \over 2}{{hv} \over {k{T_{{\rm{bg}}}}}}} \right]\,{{\rm{\Gamma }}_{{\rm{OH}}}}.$(1)

In this equation, Aul is the Einstein spontaneous coefficient of the line at frequency v, Fu is the statistical weight of the upper level of the transition (Fu = 1 and 2 for the 1667 and 1665 MHz lines, respectively), i is the inversion of the ground state Λ-doublet levels, Tbg is the background temperature, and Δ is the Earth-comet distance. ΓOH is the number of molecules within the beam, which is proportional to the OH production rate (QOH). The second term inside the brackets corresponds to spontaneous emission and is only significant for small values of i. For the 100-m GBT, the area of the 1667 MHz line, expressed in K km s−1 on the main beam brightness temperature scale, is given by:

T mBdυ=9.157×1035TbgΔ2[ i+i+12hvkTbg ]ΓOH,$\int {{T_{\,{\rm{mB}}}}{\rm{d}}\upsilon } = 9.157 \times {10^{ - 35}}\,{{{T_{{\rm{bg}}}}} \over {{{\rm{\Delta }}^2}}}\left[ {i + {{i + 1} \over 2}{{hv} \over {k{T_{{\rm{bg}}}}}}} \right]\,{{\rm{\Gamma }}_{{\rm{OH}}}},$(2)

where the geocentric distance is in astronomical units.

In the inner part of the coma, collisions thermalize the populations of the Λ-doublet and quench the signal. Based on 1P/Halley observations, Gérard (1990) showed that the quenching radius (in km) can be approximated by:

rq=4.7×104rhQOH/1029,${r_{\rm{q}}} = 4.7 \times {10^4}{r_{\rm{h}}}\sqrt {{{{Q_{{\rm{OH}}}}} \mathord{\left/ {\vphantom {{{Q_{{\rm{OH}}}}} {{{10}^{29}}}}} \right. \kern-\nulldelimiterspace} {{{10}^{29}}}}} ,$(3)

where QOH is the OH production rate in mol s−1. For an active comet such as F3, the quenched region can be expected to be comparable to the projected field of view, which is 2.29 × 105 km in diameter for the GBT and (1.07 × 105) × (5.80 × 105) km for the NRT on 24 July 2020. Hence, collisional quenching is an important factor to consider when deriving OH production rates from the GBT and NRT observations of comet F3. The nearly simultaneous observations (a mere 2 h apart) of the comet by these two instruments provide a new opportunity to measure the quenching radius. The very few prior measurements are based on the comparison of 18-cm- and UV-derived OH production rates (Gérard 1990) and on the analysis of 18-cm observations taken at various offset positions from the comet nucleus (Colom et al. 1999; Schloerb et al. 1997; Gérard et al. 1998).

In order to take into account collisional quenching, it is assumed that the maser inversion is zero for cometocentric distances rrq, following Schloerb (1988) and Gérard (1990). More realistic descriptions with a progressive quenching throughout the coma have been investigated by Colom et al. (1999) and Gérard et al. (1998). Equations (1) and (2) are then no longer valid, but can be replaced by similar equations: ΓOH is replaced by the number of unquenched OH radicals within the beam and only the spontaneous emission term (with i = 0) is considered for the OH radicals within the quenched region. Despois et al. (1981) have shown that in the collisional region the maser inversion is very small, that is on the order of 10−4 for a kinetic temperature of 300 K; thus, it can be safely neglected for the purpose of our analysis of comet F3. The calculation of the number of quenched and unquenched OH radicals within the beam is done by volume integration within a Gaussian beam using a description of the distribution of the OH radicals in the coma (see, e.g., Despois et al. 1981).

For the spatial distribution of the OH radicals, the Haser-equivalent model is employed (Combi & Delsemme 1980). The water and OH radical lifetimes at 1 au from the Sun are set to 8.5 × 104 (for the quiet Sun from Crovisier 1989, considering that the Sun was quiet in summer 2020) and 1.1 × 105 s (van Dishoeck & Dalgarno 1984, but poorly constrained as discussed in Schloerb & Gérard 1985), respectively (see also Table 3 in Crovisier et al. 2002a)8. The OH ejection velocity is set to υe = 0.95 km s−1 (Bockelée-Morvan et al. 1990; Crovisier et al. 2002a). The OH parent velocity υp is derived from the OH line profile, as described in Sect. 3.1.2.

thumbnail Fig. 4

Weighted averages of the OH 1665 and 16 657 MHz lines observed at the NRT, normalized to the intensity of the 1667 MHz (Sect. 2.3). The spectra are integrated over specified periods of time, which are indicated above the plots in yymmdd format. The observations are fitted with trapeziums based on the methodology of Bockelée-Morvan et al. (1990).

3.1.2 Line profiles and H2O outflow velocity

The observed line shapes have been analyzed in the framework of the trapezium modeling as proposed by Bockelée-Morvan et al. (1990) and subsequently applied to the kinematic studies of the coma of many comets in Tseng et al. (2007). OH is a daughter species of H2O photodissociation and is assumed here to be emitted isotropically in the rest frame of H2O. Indeed, some anisotropy may occur since water photodissociation is caused by the unidirectional solar UV radiation (e.g., Crovisier 1990). However, the currently available theoretical work and laboratory data do not permit a complete evaluation of this potential anisotropy. Taking into account collisional quenching, the line width of OH does not provide the velocity of its parent molecule. The maximum radial velocity of OH along the line of sight is υp+υe, assuming that the OH parent and OH ejection velocity distributions are monokinetic. A trapezium centered on the cometary nucleus is expected when the beam is very large with respect to the OH coma. As shown by Bockelée-Morvan et al. (1990), the half lower base of the fitted trapezium to an OH line is expected to be equal to υp+υe. Figures 2 and 4 show the trapezium method applied to the GBT and NRT OH spectra. The derived υp+υe values are given in Tables 3 and 4.

The OH parent velocity derived from the GBT 1667 MHz line observed on 24 July 2020 is υp = 1.48 ± 0.04 km s−1. The trapezium method applied to the Nançay spectrum of 24 July 2020 yields υp = 1.30 ± 0.14 km s−1, which is consistent within 1σ with the GBT-derived value. A slightly lower υp value measured at NRT is not unexpected, as it could be explained by gas acceleration in the coma. With its 3.5′ × 19′ beam, the NRT field of view is probing OH radicals closer to the surface than the 7.55′ GBT beam (at 1667 MHz). As expected, the OH parent expansion velocity υp is observed to increase from ~1.0 to ~1.8 km s−1 when the heliocentric distance rh decreases from ~0.72 to ~0.52 au (Table 4). This trend was observed in other comets (Tseng et al. 2007).

3.1.3 Combined analysis of the GBT and NRT OH observations of 24 July 2020

Due to the significant differences in beam sizes and shapes between the GBT and NRT (of a factor of about two considering the small dimension of the NRT’s elliptical beam), the fractions of quenched OH radicals that the two telescopes sample differ. The quenching radius rq can be determined by searching for the QOH value that is consistent with data from both facilities. Figure 5 shows QOH values derived as a function of rq from the 24 July 2020 GBT and NRT data using the maser inversion of i = 0.20 based on Despois et al. (1981). For the GBT and NRT calculations, the derived υp = 1.48 km s−1 and υp = 1.30 km s−1 values are utilized, respectively (Sect. 3.1.2). The obtained quenching radius is (5.95 ± 0.14) × 104 km and the consistent OH production rate is (4.62 ± 0.38) × 1029 mol s−1. Using the maser inversion i = 0.28 based on Schleicher & A’Hearn (1988), the derived rq value is the same ((5.96 ± 0.14) × 104 km), but the OH production rate is (3.41 ± 0.25) × 1029 mol s−1. The derived quenching radius is consistent with the law provided by Eq. (3), which predicts values of 5.6 × 104 and 6.6 × 104 km for the two obtained QOH values. Averaging the production rates inferred with the Despois et al. (1981) and Schleicher & A’Hearn (1988) maser inversion (i) models yields an OH production rate of (4.02 ± 0.32) × 1029 mol s−1, which is consistent with each model individually within errors. The corresponding averaged water production rate is (4.42 ± 0.35) × 1029 mol s−1. The two considered maser inversion models were developed independently and compared in Schloerb & Gérard (1985). The two models use different sources for the solar spectrum. The model of Schleicher & A’Hearn (1988) includes IR pumping (a minor effect). Both models give remarkably similar results, except when the OH inversion is low.

thumbnail Fig. 5

Derived OH production rates from the GBT (red curve) and NRT (blue curves) observations of the OH 1667 MHz line in comet C/2020 F3 (NEOWISE) on 24 July 2020. OH production rates are plotted as a function of the quenching radius. The dashed lines correspond to the ±1σ uncertainty in the line area measured at NRT. Analogous dashed lines for the GBT are indistinguishable from the solid curve due to the high signal-to-noise ratio of the data. The maser inversion is i = 0.20 based on Despois et al. (1981), and the background temperature is Tbg = 3.1 K. The OH spatial distribution is described by the Haser-equivalent model (Combi & Delsemme 1980).

3.1.4 GBT spectrum of 11 August 2020

The GBT on-nucleus data obtained on 11 August 2020 show a marginal absorption line near the frequency of the OH 1667 MHz line at the nucleus-centered beam position (Fig. 3). Unlike the 24 July 2020 spectrum, the line is narrow (~1 km s−1 ), strongly blueshifted (Δυ = −0.53 km s−1 from a Gaussian fit, Table 2), and does not show any signal at positive Doppler velocities. This line shape can be explained by the Greenstein effect (Greenstein 1958). As a result of the motion of the OH radicals in the coma, their heliocentric velocity is shifted with respect to the nucleus heliocentric velocity, adding another component to the maser inversion (Despois et al. 1981). In most instances, the Greenstein effect only affects weakly the shape of the OH lines. However, the effect is striking at heliocentric velocities where the maser inversion changes its sign over a small velocity range, since some radicals have a positive maser inversion, and others have a negative inversion.

The heliocentric velocity of comet F3 at the time when the GBT observations of 11 August 2020 were carried out was υh = +34.8 km s−1. For the Despois et al. (1981) model, the maser inversion is i = −0.125 at this υh, but varies from −0.240 to +0.006 in the coma assuming that the maximum OH expansion velocity (υpe) is 2 km s−1. For the Schleicher & A’Hearn (1988) model, the inversion is in the [−0.16, +0.09] range with i = −0.020 at this value of υh. Taking into account the Sun-comet-Earth angle of 58.8° (phase angle) and spontaneous emission, it is expected that the line will be in absorption at negative Doppler velocities and will show at positive velocities either weak positive or negative emission depending on the Schleicher et al. or the Despois et al. inversion values, respectively. The low signal-to-noise ratio prevents any conclusion about which model is in better accordance with the observed spectrum.

Equations (1) and (2) hold in the presence of the Greenstein effect, so they have been used for an attempt to determine the OH production rate on 11 August 2020. For this, the quenching law given by Eq. (3), Tbg = 3.1 K, and υp = 1.2 km s−1 have been employed. This υp value is calculated from the value of 1 . 48 km s−1 determined on 24 July (Table 3) and an assumed a rh0.5$r_{\rm{h}}^{ - 0.5}$ dependence. The Schleicher et al. excitation model cannot explain the observed line intensity (Table 3) with this quenching law. A value of QOH in the 0.5-3 × 1029 mol s−1 range is obtained using the Despois et al. inversion value. If the quenching radius is left as a free parameter, the derived QOH strongly depends on rq and i values. It becomes necessary to conclude that it is not possible to derive a reliable QOH value from the 11 August 2020 data. The 1665 MHz spectrum and the 1665, 1667 MHz spectra obtained at offset positions do not show even any marginal hints of lines and were not analyzed.

3.1.5 Evolution of the OH production rate

Table 4 presents the OH production rates determined from the NRT data using the Despois et al. (1981) and the Schleicher & A’Hearn (1988) inversion models. The average spectrum of 2–3 July 2020 observations shows an absorption line consistent with the inversion models. The maser inversion was very low on 3 July 2020. Consequently, as for the GBT data of 11 August 2020 (Sect. 3.1.4), this observation could not be used for determining a production rate. On the other hand, an OH production rate of ~2 × 1030 mol s−1 could be determined for 2 July (i.e., one day before perihelion; Table 4). Despite the maser inversion being large on this date (either −0.31 or −0.37 depending on the inversion model), the signal was weak due to a large fraction of quenched OH radicals in the beam.

The OH 18-cm lines were observed in emission with the Arecibo 305-m dish on 31.8 July 2020 (rh = 0.82 au), from which an unexpectedly low OH production rate of (3.6 ± 0.6) × 1028 mol s−1 was derived (Smith et al. 2021). However, collisional quenching was not taken into account in the analysis presented in that paper. In this work, the Arecibo data were reanalyzed with collision quenching being taken into account. The inferred values are QOH = (1.5 ± 0.3) × 1029 mol s−1 with the inversion model of Despois et al. (1981) and (2.95 ± 0.05) × 1029 mol s−1 with that of Schleicher & A’Hearn (1988). These values are in much closer agreement with the production rate estimated from optical OH line observations for the same observation date of 8.5 × 1028 mol s−1 quoted in the Smith et al. (2021) paper (based on D. Schleicher 2021, personal communication).

The derived water production rates (QH2O=1.1×QOH${Q_{{{\rm{H}}_2}{\rm{O}}}} = 1.1 \times {Q_{{\rm{OH}}}}$ ; Crovisier 1989) are plotted in Fig. 6 together with other water production rate determinations from near-IR water lines observed with long-slit spectroscopy with the iSHELL at NASA/IRTF (Faggi et al. 2021) and from Ly-α observations using SOHO/SWAN (Combi et al. 2021). There is a good agreement between the OH 18-cm and Ly-α QH2O${Q_{{{\rm{H}}_2}{\rm{O}}}}$ data post-perihelion, but a factor of two discrepancy is observed preperihelion. Understanding this is potentially important, but requires dedicated modeling efforts that are beyond the scope of this paper. Near-IR determinations, which were all obtained post-perihelion (see red symbols in Fig. 6), are overall consistent with the OH 18-cm for contemporaneous dates. The large discrepancy between near-IR and Ly-α measurements near-perihelion (rh < 0.4 au) is discussed in Faggi et al. (2021) and may be related to an extended production of water, for example, from icy grains. Generally, robust comparisons across the UV, IR, and radio domains are challenging due to the significantly different spatial scales being probed (FOV of 30°, 0.3–4″, and 32″−19′, respectively).

thumbnail Fig. 6

Water production rates of comet C/2020 F3 (NEOWISE). Black dots are values from this work and include the revised value from the Arecibo data: filled and open dots refer to values obtained with the inversion models of Schleicher & A’Hearn (1988) and Despois et al. (1981), respectively. Red dots: values from near-IR observations from Faggi et al. (2021). The dashed blue lines are a fit to the pre- and post-perihelion heliocentric values derived from Ly-a data (Combi et al. 2021).

3.2 NH3 data analysis

The analysis of the NH3 lines was performed using the excitation model described in Biver et al. (2012). Since the NH3 photodissociation lifetime is short (~2700 s at rh = 0.7 au)9 and the water production rate of comet F3 is high, the excitation of NH3 is dominated by collisional processes and its rotational levels are in thermal equilibrium. A gas kinetic temperature of 90 K is assumed, based on constraints from close-in-date CH3OH observations of this comet at the IRAM 30-m and NOEMA telescopes (Biver et al. 2022a). An expansion velocity of 1 km s−1 is utilized, as derived from the line profiles observed at IRAM 30-m and NOEMA, which is appropriate as the field of view (934″) of these observations matches the size of the expected NH3 coma.

The NH3 lines at 23–25 GHz present a hyperfine structure with several components that are well-separated (by more than 7 km s−1) from the central frequencies of the lines (as can be seen in the Cologne Database of Molecular Spectroscopy, CDMS, Müller et al. 2001, 2005; Endres et al. 2016). Each of these components, in turn, harbors closely spaced (by less than 0.03 km s−1 ) quadrupole satellite lines. Two to three such satellite components contribute to the signal in the velocity interval from −1.5 to 1.5 km s−1 that is chosen for computing the NH3 line areas in Table 3. The fraction of the intensity in the central hyperfine components is 50/80/89/93/96 % for the JK = 11 − 11 /22 − 22 /33 − 33 /44 − 44/55 − 55 lines (which can be computed based on the CDMS spectroscopic entry). With this taken into consideration, the modeled line strengths show that the JK = 22 − 22 and 33 − 33 lines are, by far, the major contributors to the expected signal in this velocity interval. Based on these two lines and an RMS−2 weighting, the resulting 3σ upper limit on the NH3 production rate is QNH3<1.3×1027${Q_{{\rm{N}}{{\rm{H}}_3}}} &gt; 1.3 \times {10^{27}}$ mol s−1. Using the water production rate of 4.4 × 1029 mol s−1 inferred from OH observations on 24 July 2020 (Sect. 3.1.3, an average of the Despois et al. 1981 and Schleicher & A’Hearn 1988 maser inversion models), the resulting 3σ upper limit for QNH3/QH2O${{{Q_{{\rm{N}}{{\rm{H}}_3}}}} \mathord{\left/ {\vphantom {{{Q_{{\rm{N}}{{\rm{H}}_3}}}} {{Q_{{{\rm{H}}_2}{\rm{O}}}}}}} \right. \kern-\nulldelimiterspace} {{Q_{{{\rm{H}}_2}{\rm{O}}}}}}$ is < 0.29%.

Faggi et al. (2021) detected near-IR lines of NH3 in comet F3 and derived NH3 abundances relative to water of 0.73 ± 0.10% (on 31 July 2020 at rh = 0.81 au, QH2O=(1.6±0.05)×1029 mol s1${Q_{{{\rm{H}}_2}{\rm{O}}}} = \left( {1.6 \pm 0.05} \right) \times {10^{29}}\,{\rm{mol}}\,{{\rm{s}}^{ - 1}}$) and 0.92 ± 0.19% (on 6 August 2020 at rh = 0.94 au, QH2O=(1.5±0.2)×1029 mol s1${Q_{{{\rm{H}}_2}{\rm{O}}}} = \left( {1.5 \pm 0.2} \right) \times {10^{29}}\,{\rm{mol}}\,{{\rm{s}}^{ - 1}}$). The discrepancy with our value is puzzling. The measurements of NH3 in Faggi et al. (2021) were usually performed on one or two very faint spectral lines, which prevented a direct derivation of Trot. Consequently, it was assumed that Trot(NH3) = Trot(H2O). Based on near-IR H2O observations, the rotational temperature was derived to be 130 K on 20 July and 90 K on 31 July. If for the analysis of the GBT NH3 lines, a Trot of 130 K is assumed, then the 3σ upper limit on the NH3 production rate is QNH3<1.4×1027 mol s1${Q_{{\rm{N}}{{\rm{H}}_3}}} &gt; 1.4 \times {10^{27}}\,{\rm{mol}}\,{{\rm{s}}^{ - 1}}$ (based on the JK = 22 − 22 and 33 − 33 lines and a RMS−2 weighting). This is ~11% higher than with a Trot of 90 K; however, still not enough to explain the discrepancy with the near-IR values. Possibly, NH3 displayed abundance variations with time. An increase in abundance relative to H2O with increasing rh was observed for CH3OH, C2H6, and CH4 species in these near-IR observations, while abundances of HCN, C2H2, and H2CO were stable relative to H2O (Faggi et al. 2021). On the other hand, increases relative to H2O with increasing rh were not observed for CH3OH nor H2CO in the IRAM 30-m data (Biver et al. 2022a).

It is not likely that the water production rate has been strongly overestimated in this work, thereby resulting in a low QNH3/QH2O${{{Q_{{\rm{N}}{{\rm{H}}_3}}}} \mathord{\left/ {\vphantom {{{Q_{{\rm{N}}{{\rm{H}}_3}}}} {{Q_{{{\rm{H}}_2}{\rm{O}}}}}}} \right. \kern-\nulldelimiterspace} {{Q_{{{\rm{H}}_2}{\rm{O}}}}}}$ ratio. For the 24 July 2020 date in question (rh = 0.7 au), here a QH2O=(4.42±0.35)×1029${Q_{{{\rm{H}}_2}{\rm{O}}}} = \left( {4.42 \pm 0.35} \right) \times {10^{29}}$ mol s−1 is used (Sect. 3.1.3). Unfortunately, near-IR measurements are not available for this exact date. For 20 July 2020 (rh = 0.56 au), Faggi et al. (2021) obtain (4.2 ± 0.2) × 1029 mol s−1, whereas the QH2O${Q_{{{\rm{H}}_2}{\rm{O}}}}$ value obtained from radio OH observation on this date is (6.3 ± 0.7) × 1029 mol s−1 (weighted average of values deduced using the two maser inversion models, Table 4). Hence, the OH-derived value is only slightly (factor of 1.5) larger than the IR-derived value. If QH2O${Q_{{{\rm{H}}_2}{\rm{O}}}}$ was a factor of 1 .5 lower than what has been obtained in the current analysis on 24 July 2020, the upper limit for QNH3/QH2O${{{Q_{{\rm{N}}{{\rm{H}}_3}}}} \mathord{\left/ {\vphantom {{{Q_{{\rm{N}}{{\rm{H}}_3}}}} {{Q_{{{\rm{H}}_2}{\rm{O}}}}}}} \right. \kern-\nulldelimiterspace} {{Q_{{{\rm{H}}_2}{\rm{O}}}}}}$ would increase to < 0.44%, which is still well below the ratio obtained from near-IR.

The OH that is observed in comet F3 at rh = 0.7 au stems predominantly from H2O that has exited the nucleus between ~4 × 104 s (~12 h, lifetime of water) and ~8 × 104 s (~22 h, time for the OH radicals to reach a distance corresponding to the projected GBT beam radius) earlier. On the other hand, NH3 would be freshly released and would not survive for more than ~2.7 × 103 s (~45 min). There is no evidence for strong daily variability in QH2O${Q_{{{\rm{H}}_2}{\rm{O}}}}$ in the radio observation presented nor when they are compared to the near-IR observation of (Faggi et al. 2021, Fig. 6). Consequently, the discrepancy between near-IR and radio determinations of QNH3/QH2O${{{Q_{{\rm{N}}{{\rm{H}}_3}}}} \mathord{\left/ {\vphantom {{{Q_{{\rm{N}}{{\rm{H}}_3}}}} {{Q_{{{\rm{H}}_2}{\rm{O}}}}}}} \right. \kern-\nulldelimiterspace} {{Q_{{{\rm{H}}_2}{\rm{O}}}}}}$ may be related to the temporal variability of QNH3${Q_{{\rm{N}}{{\rm{H}}_3}}}$.

thumbnail Fig. 7

NH3 abundances relative to water measured in comets from centimeter and submillimeter observations (hollow symbols), and near-IR observations (filled symbols). Different colors are used for Jupiter-family comets, dynamically new comets from the Oort Cloud, and long-period comets originating from the Oort Cloud. Downward-pointing arrowheads are 3σ upper limits. Dark blue upward-pointing triangles refer to the measurements of comet F3 from Faggi et al. (2021). The dark blue downward-pointing arrowhead is the value derived in this work. All data used for this figure are tabulated in Tables B.1 and B.2 with the corresponding references.

4 Discussion

Comet F3 was supposed to be a prime target for the investigation of the evolution of the NH3/H2O ratio as a function of heliocentric distance. It boasted a high water production rate (> 1030 mol s−1 around perihelion), which was firmly quantified during the two month-long monitoring of OH with the NRT. Its heliocentric distance at perihelion (0.295 au) and geocentric distance at perigee (0.692 au) were short. The GBT observational campaign targeted OH and NH3 near-simultaneously, thereby ensuring robust constraints on the NH3/H2O ratio. However, the NH3 hyperfine inversion lines at 23–25 GHz eluded detection like for many other comets in the past (Sect. 1). The NH3 abundance relative to water was quantified in several preceding comets (Fig. 7) based on detections of NH3 at 572 GHz using Odin, Herschel, and Rosetta/MIRO (Biver et al. 2007, 2012, 2019, and unpublished results from N. Biver, personal communication) and in the near-IR (Dello Russo et al. 2016; Lippi et al. 2021, and other references in Table B.1). The upper limit obtained for comet F3 (blue downward-pointing arrowhead in Fig. 7) is in the low range of values measured in comets at 0.7 au from the Sun.

There is a trend for higher NH3 abundances at low (< 1 au) heliocentric distances (Fig. 7), suggesting a possible contribution of NH3 released from the thermal degradation of compounds on grains at small rh. This is supported by the radial profiles of NH3, which are often more extended than expected for a nuclear source of NH3 (e.g., DiSanti et al. 2016; Dello Russo et al. 2022). A similar increase at low heliocentric distances is also seen in the Q(NH3)/Q(C2H6) ratio (Fig. 5 of Mumma et al. 2019). The thermal decomposition of ammonium salts discovered in comet 67P/C-G (Quirico et al. 2016; Altwegg et al. 2020; Poch et al. 2020) has been proposed to explain the excess of NH3 production at small rh (Mumma et al. 2019). This excess of production is not observed for comet F3, based on the here-obtained NH3/H2O upper limit and the NH3 abundance determined in the near-IR (Faggi et al. 2021). A contributing factor could be the amount of water-poor dust launched into the coma of a comet. At 0.7 au from the Sun, the equilibrium temperature of dust grains is already above the ~160–230 K threshold for thermal degradation of some ammonium salts. The increase in the NH3/H2O ratio must therefore be correlated with a higher dust density in the coma. Furthermore, given enough dust in the coma, an initial grain size distribution can be altered by fragmentation as it moves outward through the coma, which may lead to an increase in the small grain population as a function of distance from the nucleus. The small grains can be super-heated above the equilibrium temperature; whereby, salts with the highest binding energies would also thermally disintegrate. This dust must be dry, i.e., water-poor, because otherwise the amount of H2O would increase together with NH3 and potentially mask the increase in the relative abundance of NH3.

The dust-to-gas ratio in a cometary coma may vary drastically with outbursts. Comet F3 was not observed to undergo outbursts during its 2022 perihelion passage, but rather to only display strong jet activity (Combi et al. 2021, Sect. 1). Additional observations of NH3 at low rh are obviously needed. Although, it is not excluded that enhanced NH3/H2O ratios may be shortlived as ammonium salts may not survive for long on dust in the coma following an injection of dust from an outburst. Typically, outbursts are more frequent at smaller heliocentric distances, in agreement with the NH3/H2O increasing trend. Continuous monitoring pre- and post-outburst would allow the lifetime of ammonium salts in the coma to be estimated.

Several salts have been identified to be present on the surface of 67P/C-G: NH4+Cl${\rm{NH}}_4^ + {\rm{C}}{{\rm{l}}^ - }$, NH4+CN${\rm{NH}}_4^ + {\rm{C}}{{\rm{N}}^ - }$, NH4+OCN${\rm{NH}}_4^ + {\rm{OC}}{{\rm{N}}^ - }$, NH4+HCOO${\rm{NH}}_4^ + {\rm{HCO}}{{\rm{O}}^ - }$, NH4+CH3COO${\rm{NH}}_4^ + {\rm{C}}{{\rm{H}}_3}{\rm{CO}}{{\rm{O}}^ - }$, NH4+SH${\rm{NH}}_4^ + {\rm{S}}{{\rm{H}}^ - }$, and NH4+F${\rm{NH}}_4^ + {{\rm{F}}^ - }$ (Altwegg et al. 2020, 2022). As ammonium salts degrade at higher temperatures at smaller heliocentric distances, not only the NH3/H2O ratio should display an increasing trend, but also the ratio of the acid-counterparts (HCl, HCN, HOCN, HCOOH, HCOOCH3) relative to H2O. However, to what extent each of these individual salts makes a significant contribution relative to the amount of these species already present in the ice in the nucleus is not clear. Modeling work is required to quantify these effects. For the case of the CN radical, it has been shown that it does appear to have a distributed source (e.g., Opitom et al. 2016) and requires an additional parent molecule to match its signal strength in the ROSINA measurements of the inner coma of 67P/C-G (Hänni et al. 2020, 2021), thus likely being a product of ammonium salt thermal degradation. This has also been observed for NH2, a photodissociation product of NH3 (Opitom et al. 2019). An important step forward would be the comparison of the 14N/15N isotopic ratio in the dust with that in various N-bearing volatiles. It has been shown that the 14N/15N isotopic ratio for NH3, NO, N2 in 67P/C-G is in the 93–160 range, which is consistent with the ~140 measured in HCN, CN, and NH2 in other comets (Altwegg et al. 2019; Bockelée-Morvan et al. 2015; Biver et al. 2022b). This ratio has not been reported for the dust of 67P/C-G.

5 Conclusions

This paper presents the two month-long monitoring campaign of the OH emission from comet C/2020 F3 (NEOWISE) at the Nançay Radio Telescope, which is used to determine the H2O production rate. Furthermore, GBT observations of F3 targeting OH lines on two separate days (24 July and 11 August 2020) and the NH3 line observations contemporaneous with the first date are presented. The main results are as follows.

  1. The OH parent expansion velocity (υp) increases from 1.0 to 1 .8 km s−1 with decreasing heliocentric distance from 0.72 to 0.52 au;

  2. The OH quenching radius (rq) is determined to be (5.96 ± 0.10) × 104 km on 24 July 2020 (rh = 0.7 au) from the analysis of the concurrent NRT and GBT OH observations. This value consistently explains the OH line intensities measured by the GBT and NRT for an OH production rate (QOH) of (4.02 ± 0.32) × 1029 mol s−1. This rq value is consistent with the Gérard (1990) prescription;

  3. The Greenstein effect (Greenstein 1958) is observationally demonstrated in the OH observations of 11 August 2020 taken at the GBT, yielding a narrow, strongly blueshifted line in absorption. The signal-to-noise ratio of the data is not high enough to distinguish between the Despois et al. (1981) and Schleicher & A’Hearn (1988) maser inversion models;

  4. One day before perihelion (2 July 2020), the H2O production rate (QH2O${Q_{{{\rm{H}}_2}{\rm{O}}}}$) was very high (~2 × 1030 mol s−1) in comet F3. Pre-perihelion, QH2O${Q_{{{\rm{H}}_2}{\rm{O}}}}$ increases with decreasing heliocentric distance and agrees within a factor of 2 with Ly-a observations using SOHO/SWAN of Combi et al. (2021). Post-perihelion, QH2O${Q_{{{\rm{H}}_2}{\rm{O}}}}$ decreases with the increasing heliocentric distance and is in excellent agreement with Ly-α observations;

  5. The 3σ upper limit for QNH3/QH2O${{{Q_{{\rm{N}}{{\rm{H}}_3}}}} \mathord{\left/ {\vphantom {{{Q_{{\rm{N}}{{\rm{H}}_3}}}} {{Q_{{{\rm{H}}_2}{\rm{O}}}}}}} \right. \kern-\nulldelimiterspace} {{Q_{{{\rm{H}}_2}{\rm{O}}}}}}$ is < 0.29% for comet F3 at 0.7 au from the Sun (post-perihelion on 24 July 2020), which is in the low range of values obtained for other comets at similar heliocentric distances;

  6. The differences in the NH3/H2O ratios measured for comet F3 with radio and near-IR observations may hint at this ratio being highly variable with time in a cometary coma.

Acknowledgements

This work is supported by the Swiss National Science Foundation (SNSF) Ambizione grant no. 180079, the Center for Space and Habitability (CSH) Fellowship, and the IAU Gruber Foundation Fellowship. MAC, SBC, and SNM’s work was supported by the NASA Planetary Science Division Internal Scientist Funding Program through the Fundamental Laboratory Research (FLaRe) work package. This work benefited from discussions held with the international team #461 “Provenances of our Solar System’s Relics” (team leaders Maria N. Drozdovskaya and Cyrielle Opitom) at the International Space Science Institute, Bern, Switzerland. The Green Bank Observatory is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc. The authors are grateful for the guidance from Tapasi Ghosh and project friend Larry Morgan in reducing L-band and KFPA data, respectively. The Nançay Radio Observatory is operated by the Paris Observatory, associated with the French Centre National de la Recherche Scientifique (CNRS) and with the University of Orléans. The authors thank the referee, Mike Mumma, for constructive comments that have strengthened the manuscript.

Appendix A Averaged NRT spectra of comet F3 for 18-27 July 2020

thumbnail Fig. A.1:

Four OH 18-cm lines observed in comet F3 with the NRT averaged from 18 to 27 July 2020. The expected statistical relative intensities of the 1 667 : 1 665 : 1 612 : 1 720 MHz lines are 9 : 5 : 1 : 1, which is in good agreement with the displayed observations.

Appendix B Ammonia in comets

NH3 abundances relative to water measured in comets from ground-based radio and IR observations are tabulated in Table B.1, while those obtained with in situ measurements are given in Table B.2 with the corresponding references. For Fig. 7, all ground-based values have been used from Table B.1 except for the measurement for comet C/1995 O1 (Hale-Bopp) from Hirota et al. (1999) due to its large, unconstraining error bars. From Table B.2, only the two reported in lines two and three have been used for Fig. 7. The other measurements for comet 67P/Churyumov-Gerasimenko have been tabulated in order to illustrate that a range of values has been measured at this comet over the 2-yr duration of the ESA Rosetta mission with two independent instruments: MIRO and ROSINA. The data point of 1P/Halley has been excluded from Fig. 7, because its error bars are large and do not allow for a meaningful comparison.

Table B.1

Abundances of ammonia in comets from ground-based observations.

Table B.2

Abundances of ammonia in comets from in situ measurements.

References

  1. Altenhoff, W. J., Batrla, W. K., Huchtmeier, W. K., et al. 1983, A&A, 125, L19 [NASA ADS] [Google Scholar]
  2. Altwegg, K., Balsiger, H., & Fuselier, S. A. 2019, ARA&A, 57, 113 [NASA ADS] [CrossRef] [Google Scholar]
  3. Altwegg, K., Balsiger, H., Hänni, N., et al. 2020, Nat. Astron., 4, 533 [NASA ADS] [CrossRef] [Google Scholar]
  4. Altwegg, K., Combi, M., Fuselier, S. A., et al. 2022, MNRAS, 516, 3900 [NASA ADS] [CrossRef] [Google Scholar]
  5. Balsiger, H., Altwegg, K., Bochsler, P., et al. 2007, Space Sci. Rev., 128, 745 [NASA ADS] [CrossRef] [Google Scholar]
  6. Bauer, J., Gicquel, A., Kramer, E., et al. 2020, in AAS/Division for Planetary Sciences Meeting Abstracts, 52, 316.04 [NASA ADS] [Google Scholar]
  7. Bergner, J. B., Öberg, K. I., Rajappan, M., & Fayolle, E. C. 2016, ApJ, 829, 85 [NASA ADS] [CrossRef] [Google Scholar]
  8. Bernstein, M. P., Sandford, S. A., Allamandola, L. J., Chang, S., & Scharberg, M. A. 1995, ApJ, 454, 327 [NASA ADS] [CrossRef] [Google Scholar]
  9. Bertin, M., Martin, I., Duvernay, F., et al. 2009, Phys. Chem. Chem. Phys. (Incorp. Faraday Trans.), 11, 1838 [CrossRef] [Google Scholar]
  10. Bird, M. K., Huchtmeier, W. K., von Kap-Herr, A., Schmidt, J., & Walmsley, C.M. 1987, in Cometary Radio Astronomy, eds. W.M. Irvine, F.P. Schloerb, & L.E. Tacconi-Garman, 85 [Google Scholar]
  11. Bird, M. K., Huchtmeier, W. K., Gensheimer, P., et al. 1997, A&A, 325, L5 [NASA ADS] [Google Scholar]
  12. Bird, M. K., Hatchell, J., van der Tak, F. F. S., Crovisier, J., & Bockelée-Morvan, D. 2002, in ESA SP, 500, 697 [NASA ADS] [Google Scholar]
  13. Bischoff, R., & Mugrauer, M. 2021, Astron. Nachr., 342, 833 [NASA ADS] [CrossRef] [Google Scholar]
  14. Biver, N., Bockelée-Morvan, D., Crovisier, J., et al. 2007, Planet. Space Sci., 55, 1058 [Google Scholar]
  15. Biver, N., Crovisier, J., Bockelée-Morvan, D., et al. 2012, A&A, 539, A68 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  16. Biver, N., Bockelée-Morvan, D., Hofstadter, M., et al. 2019, A&A, 630, A19 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  17. Biver, N., Boissier, J., Bockelée-Morvan, D., et al. 2022a, A&A, 668, A171 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  18. Biver, N., Dello Russo, N., Opitom, C., & Rubin, M. 2022b, arXiv e-prints, [arXiv:2207.04800] [Google Scholar]
  19. Blum, J., Bischoff, D., & Gundlach, B. 2022, Universe, 8, 381 [CrossRef] [Google Scholar]
  20. Bockelée-Morvan, D., Crovisier, J., & Gérard, E. 1990, A&A, 238, 382 [NASA ADS] [Google Scholar]
  21. Bockelée-Morvan, D., Lis, D. C., Wink, J. E., et al. 2000, A&A, 353, 1101 [Google Scholar]
  22. Bockelée-Morvan, D., Calmonte, U., Charnley, S., et al. 2015, Space Sci. Rev., 197, 47 [Google Scholar]
  23. Bodewits, D., Bonev, B. P., Cordiner, M. A., & Villanueva, G. L. 2022, arXiv e-prints, [arXiv:2209.02616] [Google Scholar]
  24. Bonev, B. P., & Mumma, M. J. 2006, ApJ, 653, 788 [NASA ADS] [CrossRef] [Google Scholar]
  25. Bonev, B. P., Mumma, M. J., DiSanti, M. A., et al. 2006, ApJ, 653, 774 [NASA ADS] [CrossRef] [Google Scholar]
  26. Bonev, B. P., Dello Russo, N., DiSanti, M. A., et al. 2021, PSJ, 2, 45 [NASA ADS] [Google Scholar]
  27. Boogert, A. C. A., Gerakines, P. A., & Whittet, D. C. B. 2015, ARA&A, 53, 541 [Google Scholar]
  28. Bossa, J. B., Theulé, P., Duvernay, F., Borget, F., & Chiavassa, T. 2008, A&A, 492, 719 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  29. Butler, B. J., Wootten, A., Palmer, P., et al. 2002, in Asteroids, Comets, Meteors 2002 Book of Abstracts, Berlin, Germany [Google Scholar]
  30. Cambianica, P., Cremonese, G., Munaretto, G., et al. 2021, A&A, 656, A160 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  31. Carnall, A. C. 2017, arXiv e-prints, [arXiv: 1705.05165] [Google Scholar]
  32. Churchwell, E., Landecker, T., Winnewisser, G., Hills, R., & Rahe, J. 1976, in NASA SP, 393, 281 [NASA ADS] [Google Scholar]
  33. Cochran, A. L., Nelson, T., McKay, A. J., et al. 2020, in AAS/Division for Planetary Sciences Meeting Abstracts, 52, 111.03 [NASA ADS] [Google Scholar]
  34. Colom, P., Gérard, E., Crovisier, J., et al. 1999, Earth Moon and Planets, 78, 37 [Google Scholar]
  35. Combi, M. 2002, Earth Moon and Planets, 89, 73 [Google Scholar]
  36. Combi, M. R., & Delsemme, A. H. 1980, ApJ, 237, 633 [NASA ADS] [CrossRef] [Google Scholar]
  37. Combi, M. R., Reinard, A. A., Bertaux, J. L., Quemerais, E., & Mäkinen, T. 2000, Icarus, 144, 191 [NASA ADS] [CrossRef] [Google Scholar]
  38. Combi, M. R., Mäkinen, T., Bertaux, J. L., Quémerais, E., & Ferron, S. 2021, ApJ, 907, L38 [NASA ADS] [CrossRef] [Google Scholar]
  39. Crovisier, J. 1989, A&A, 213, 459 [NASA ADS] [Google Scholar]
  40. Crovisier, J. 1990, in Asteroids, Comets, Meteors III, eds. C.I. Lagerkvist, H. Rickman, & B.A. Lindblad, 297 [Google Scholar]
  41. Crovisier, J., Colom, P., Gérard, E., Bockelée-Morvan, D., & Bourgois, G. 2002a, A&A, 393, 1053 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  42. Crovisier, J., Colom, P., Gérard, É., et al. 2002b, in ESA SP, 500, 685 [NASA ADS] [Google Scholar]
  43. Danger, G., Borget, F., Chomat, M., et al. 2011, A&A, 535, A47 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  44. Dello Russo, N., Kawakita, H., Vervack, R. J., & Weaver, H. A. 2016, Icarus, 278, 301 [NASA ADS] [CrossRef] [Google Scholar]
  45. Dello Russo, N., Kawakita, H., Bonev, B. P., et al. 2020, Icarus, 335, 113411 [NASA ADS] [CrossRef] [Google Scholar]
  46. Dello Russo, N., Vervack, R. J., Kawakita, H., et al. 2022, PSJ, 3, 6 [NASA ADS] [Google Scholar]
  47. De Sanctis, M. C., Ammannito, E., Raponi, A., et al. 2015, Nature, 528, 241 [NASA ADS] [CrossRef] [Google Scholar]
  48. De Sanctis, M. C., Vinogradoff, V., Raponi, A., et al. 2019, MNRAS, 482, 2407 [NASA ADS] [CrossRef] [Google Scholar]
  49. Despois, D., Gérard, E., Crovisier, J., & Kazes, I. 1981, A&A, 99, 320 [NASA ADS] [Google Scholar]
  50. DiSanti, M. A., Bonev, B. P., Gibb, E. L., et al. 2016, ApJ, 820, 34 [NASA ADS] [CrossRef] [Google Scholar]
  51. DiSanti, M. A., Bonev, B. P., Russo, N. D., et al. 2017, AJ, 154, 246 [NASA ADS] [CrossRef] [Google Scholar]
  52. DiSanti, M. A., Bonev, B. P., Gibb, E. L., et al. 2018, AJ, 156, 258 [NASA ADS] [CrossRef] [Google Scholar]
  53. Drozdovskaya, M. N., van Dishoeck, E. F., Rubin, M., Jørgensen, J. K., & Altwegg, K. 2019, MNRAS, 490, 50 [Google Scholar]
  54. Endres, C. P., Schlemmer, S., Schilke, P., Stutzki, J., & Müller, H. S. P. 2016, J. Mol. Spectrosc., 327, 95 [NASA ADS] [CrossRef] [Google Scholar]
  55. Faggi, S., Villanueva, G. L., Mumma, M. J., & Paganini, L. 2018, AJ, 156, 68 [NASA ADS] [CrossRef] [Google Scholar]
  56. Faggi, S., Lippi, M., Camarca, M., et al. 2021, AJ, 162, 178 [NASA ADS] [CrossRef] [Google Scholar]
  57. Feldman, P. D., Cochran, A. L., & Combi, M. R. 2004, in Comets II, eds. M.C. Festou, H.U. Keller, & H.A. Weaver, 425 [Google Scholar]
  58. Gálvez, O., Maté, B., Herrero, V. J., & Escribano, R. 2010, ApJ, 724, 539 [NASA ADS] [CrossRef] [Google Scholar]
  59. Gasc, S., Altwegg, K., Balsiger, H., et al. 2017, MNRAS, 469, S108 [Google Scholar]
  60. Geiss, J. 1988, Rev. Mod. Astron., 1, 1 [NASA ADS] [CrossRef] [Google Scholar]
  61. Gerakines, P. A., Moore, M. H., & Hudson, R. L. 2004, Icarus, 170, 202 [Google Scholar]
  62. Gérard, E. 1990, A&A, 230, 489 [NASA ADS] [Google Scholar]
  63. Gérard, E., Crovisier, J., Colom, P., et al. 1998, Planet. Space Sci., 46, 569 [CrossRef] [Google Scholar]
  64. Giorgini, J. D., Yeomans, D. K., Chamberlin, A. B., et al. 1996, in AAS/Division for Planetary Sciences Meeting Abstracts, 28, 25.04 [NASA ADS] [Google Scholar]
  65. Giorgini, J. D., Chodas, P. W., & Yeomans, D. K. 2001, in AAS/Division for Planetary Sciences Meeting Abstracts, 33, 58.13 [NASA ADS] [Google Scholar]
  66. Greenstein, J. L. 1958, ApJ, 128, 106 [Google Scholar]
  67. Grim, R. J. A., Greenberg, J. M., de Groot, M. S., et al. 1989, A&As, 78, 161 [NASA ADS] [Google Scholar]
  68. Gulkis, S., Frerking, M., Crovisier, J., et al. 2007, Space Sci. Rev., 128, 561 [Google Scholar]
  69. Hänni, N., Gasc, S., Etter, A., et al. 2019, J. Phys. Chem. A, 123, 5805 [CrossRef] [Google Scholar]
  70. Hänni, N., Altwegg, K., Pestoni, B., et al. 2020, MNRAS, 498, 2239 [Google Scholar]
  71. Hänni, N., Altwegg, K., Balsiger, H., et al. 2021, A&A, 647, A22 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  72. Harris, W. M., Scherb, F., Mierkiewicz, E., Oliversen, R., & Morgenthaler, J. 2002, ApJ, 578, 996 [NASA ADS] [CrossRef] [Google Scholar]
  73. Hatchell, J., Bird, M. K., van der Tak, F. F. S., & Sherwood, W. A. 2005, A&A, 439, 777 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  74. Heays, A. N., Bosman, A. D., & van Dishoeck, E. F. 2017, A&A, 602, A105 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  75. Hirota, T., Yamamoto, S., Kawaguchi, K., Sakamoto, A., & Ukita, N. 1999, ApJ, 520, 895 [NASA ADS] [CrossRef] [Google Scholar]
  76. Huebner, W. F., & Mukherjee, J. 2015, Planet. Space Sci., 106, 11 [Google Scholar]
  77. Huebner, W. F., Keady, J. J., & Lyon, S. P. 1992, Ap&SS, 195, 1 [NASA ADS] [CrossRef] [Google Scholar]
  78. Jewitt, D., & Hsieh, H. H. 2022, arXiv e-prints, [arXiv:2283.81397] [Google Scholar]
  79. Khan, Y., Gibb, E. L., Bonev, B. P., et al. 2021, PSJ, 2, 20 [NASA ADS] [Google Scholar]
  80. Knight, M., & Battams, K. 2020, ATel, 13853, 1 [NASA ADS] [Google Scholar]
  81. Kruczkiewicz, F., Vitorino, J., Congiu, E., Theulé, P., & Dulieu, F. 2021, A&A, 652, A29 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  82. Lang, D., Hogg, D. W., Mierle, K., Blanton, M., & Roweis, S. 2010, AJ, 139, 1782 [Google Scholar]
  83. Läuter, M., Kramer, T., Rubin, M., & Altwegg, K. 2022, ACS Earth Space Chem/, 6, 1189 [CrossRef] [Google Scholar]
  84. Le Roy, L., Altwegg, K., Balsiger, H., et al. 2015, A&A, 583, A1 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  85. Lewis, J. S., & Prinn, R. G. 1980, ApJ, 238, 357 [NASA ADS] [CrossRef] [Google Scholar]
  86. Lippi, M., Villanueva, G. L., Mumma, M. J., & Faggi, S. 2021, AJ, 162, 74 [NASA ADS] [CrossRef] [Google Scholar]
  87. Manzini, F., Oldani, V., Ochner, P., et al. 2021, MNRAS, 506, 6195 [CrossRef] [Google Scholar]
  88. Masters, J., Garwood, B., Langston, G., & Shelton, A. 2011, ASP Conf. Ser., 442, 127 [NASA ADS] [Google Scholar]
  89. McGuire, B. A., Burkhardt, A. M., Loomis, R. A., et al. 2020, ApJ, 900, L10 [Google Scholar]
  90. Meier, R., Eberhardt, P., Krankowsky, D., & Hodges, R. R. 1994, A&A, 287, 268 [NASA ADS] [Google Scholar]
  91. Mispelaer, F., Theule, P., Duvernay, F., Roubin, P., & Chiavassa, T. 2012, A&A, 540, A40 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  92. Müller, D. R., Altwegg, K., Berthelier, J. J., et al. 2022, A&A, 662, A69 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  93. Müller, H. S. P., Thorwirth, S., Roth, D. A., & Winnewisser, G. 2001, A&A, 370, L49 [Google Scholar]
  94. Müller, H. S. P., Schlöder, F., Stutzki, J., & Winnewisser, G. 2005, J. Mol. Struct., 742, 215 [Google Scholar]
  95. Mumma, M. J., & Charnley, S. B. 2011, ARA&A, 49, 471 [NASA ADS] [CrossRef] [Google Scholar]
  96. Mumma, M. J., Disanti, M. A., dello Russo, N., et al. 1996, Science, 272, 1310 [NASA ADS] [CrossRef] [Google Scholar]
  97. Mumma, M. J., Charnley, S. B., Cordiner, M., Paganini, L., & Villanueva, G. L. 2017, in AAS/Division for Planetary Sciences Meeting Abstracts, 49, 414.19 [NASA ADS] [Google Scholar]
  98. Mumma, M. J., Charnley, S., Cordiner, M., et al. 2018, in AAS/Division for Planetary Sciences Meeting Abstracts, 50, 209.02 [NASA ADS] [Google Scholar]
  99. Mumma, M., Charnley, S., Cordiner, M., et al. 2019, in EPSC-DPS Joint Meeting 2019, 1916 [Google Scholar]
  100. Muñoz Caro, G. M., & Schutte, W. A. 2003, A&A, 412, 121 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  101. Noble, J. A., Theule, P., Borget, F., et al. 2013, MNRAS, 428, 3262 [NASA ADS] [CrossRef] [Google Scholar]
  102. Opitom, C., Guilbert-Lepoutre, A., Jehin, E., et al. 2016, A&A, 589, A8 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  103. Opitom, C., Yang, B., Selman, F., & Reyes, C. 2019, A&A, 628, A128 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  104. Palmer, P., Wootten, A., Butler, B., et al. 1996, AAS Meet. Abstr., 188, 62.12 [NASA ADS] [Google Scholar]
  105. Poch, O., Istiqomah, I., Quirico, E., et al. 2020, Science, 367, aaw7462 [NASA ADS] [CrossRef] [Google Scholar]
  106. Quirico, E., Moroz, L. V., Schmitt, B., et al. 2016, Icarus, 272, 32 [CrossRef] [Google Scholar]
  107. Raunier, S., Chiavassa, T., Marinelli, F., Allouche, A., & Aycard, J. P. 2003, Chem. Phys. Lett., 368, 594 [NASA ADS] [CrossRef] [Google Scholar]
  108. Raunier, S., Chiavassa, T., Marinelli, F., & Aycard, J.-P. 2004, Chem. Phys., 302, 259 [Google Scholar]
  109. Roshi, D. A., Bloss, M., Brandt, P., et al. 2011, in 2011 XXXth URSI General Assembly and Scientific Symposium, 1 [Google Scholar]
  110. Roth, N. X., Gibb, E. L., Bonev, B. P., et al. 2018, AJ, 156, 251 [NASA ADS] [CrossRef] [Google Scholar]
  111. Rubin, M., Tenishev, V. M., Combi, M. R., et al. 2011, Icarus, 213, 655 [NASA ADS] [CrossRef] [Google Scholar]
  112. Rubin, M., Altwegg, K., Balsiger, H., et al. 2015, Science, 348, 232 [NASA ADS] [CrossRef] [Google Scholar]
  113. Rubin, M., Altwegg, K., Balsiger, H., et al. 2019, MNRAS, 489, 594 [Google Scholar]
  114. Schleicher, D. G., & A’Hearn, M. F. 1988, ApJ, 331, 1058 [Google Scholar]
  115. Schloerb, F. P. 1988, ApJ, 332, 524 [NASA ADS] [CrossRef] [Google Scholar]
  116. Schloerb, F. P., & Gérard, E. 1985, AJ, 90, 1117 [NASA ADS] [CrossRef] [Google Scholar]
  117. Schloerb, F. P., De Vries, C. H., Lovell, A. J., et al. 1997, Earth Moon and Planets, 78, 45 [NASA ADS] [CrossRef] [Google Scholar]
  118. Schroeder, I. R. H. G., Altwegg, K., Balsiger, H., et al. 2019, MNRAS, 489, 4734 [NASA ADS] [CrossRef] [Google Scholar]
  119. Schutte, W. A., Boogert, A. C. A., Tielens, A. G. G. M., et al. 1999, A&A, 343, 966 [NASA ADS] [Google Scholar]
  120. Sita, M. L., Changala, P. B., Xue, C., et al. 2022, ApJ, 938, L12 [NASA ADS] [CrossRef] [Google Scholar]
  121. Smith, A. J., Anish Roshi, D., Manoharan, P., et al. 2021, PSJ, 2, 123 [NASA ADS] [Google Scholar]
  122. Tseng, W. L., Bockelée-Morvan, D., Crovisier, J., Colom, P., & Ip, W. H. 2007, A&A, 467, 729 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  123. Ulich, B. L., & Haas, R. W. 1976, ApJS, 30, 247 [Google Scholar]
  124. van Broekhuizen, F. A., Keane, J. V., & Schutte, W. A. 2004, A&A, 415, 425 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  125. van Dishoeck, E. F., & Dalgarno, A. 1984, ApJ, 277, 576 [NASA ADS] [CrossRef] [Google Scholar]
  126. Vinogradoff, V., Duvernay, F., Danger, G., Theulé, P., & Chiavassa, T. 2011, A&A, 530, A128 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  127. Weissman, P., Morbidelli, A., Davidsson, B., & Blum, J. 2020, Space Sci. Rev., 216, 6 [CrossRef] [Google Scholar]
  128. Wilson, T. L. 1999, Rep. Progr. Phys., 62, 143 [CrossRef] [Google Scholar]
  129. Wootten, A., Butler, B., Bockelee-Morvan, D., et al. 1996a, in Asteroids, Comets, Meteors 1996, 8–12 July 1996, Versailles, France [Google Scholar]
  130. Wootten, A., Butler, B., Bockelee-Morvan, D., et al. 1996b, IAU Circ., 6362, 1 [NASA ADS] [Google Scholar]
  131. Wyckoff, S., Tegler, S. C., & Engel, L. 1991, ApJ, 367, 641 [NASA ADS] [CrossRef] [Google Scholar]
  132. Ye, Q., Zhang, Q., Brewer, J., Knight, M., & Kelley, M. 2020, in AAS/Division for Planetary Sciences Meeting Abstracts, 52, 111.02 [NASA ADS] [Google Scholar]

1

Observed on 5 July 2020 and viewed at https://science.nasa.gov/comet-neowise-iss on 27 June 2023.

2

For example, even winning first place in the 2021 IAU OAE Astrophotography Contest in the Comets category: “Neowise’s metamorphosis” by Tomàš Slovinský and Petr Horálek (Slovakia) spanning observations from 9 July to 2 August 2020 and viewed at https://www.iau.org/public/images/detail/ann21047d/ on 27 June 2023.

3

Minor Planet Electronic Circular (MPEC) 2020-G05: https://minorplanetcenter.net/mpec/K20/K20G05.html

4

Observed on 8 August 2020 and seen at https://hubblesite.org/contents/media/images/2020/45/4731-Image?news=true on 27 June 2023.

5

GBTIDL is an interactive package for reduction and analysis of spectral line data taken with the GBT. See http://gbtidl.nrao.edu/

6

https://specutils.readthedocs.io/en/stable/api/ specutils.manipulation.FluxConservingResampler.html# specutils.manipulation.FluxConservingResampler

7

Giorgini, JD and JPL Solar System Dynamics Group, NASA/JPL Horizons On-Line Ephemeris System, https://ssd.jpl.nasa.gov/horizons/, data retrieved 8 and 12 February 2021 (Giorgini et al. 1996, 2001), solution JPL#23.

8

Using the updated value from Heays et al. (2017), OH radical lifetime at 1 au from the Sun is 1.6 × 105 s.

9

Computed from 1/kpd×rh2${1 \mathord{\left/ {\vphantom {1 {{k_{{\rm{pd}}}} \times r_{\rm{h}}^2}}} \right. \kern-\nulldelimiterspace} {{k_{{\rm{pd}}}} \times r_{\rm{h}}^2}}$, where kpd is the photodissociation rate of NH3 in a Solar radiation field at 1 au and rh is the comet’s heliocentric distance in au. Here, kpd = 1.8 × 10−4 s−1 is assumed, which is the average rate for a quiet and an active Sun from Huebner et al. (1992) and Huebner & Mukherjee (2015) for the NH2(X2B1 ) + H channel, which is the dominant photodissociation channel in the Solar radiation field (Heays et al. 2017). This rate is marginally higher than the 1.5 × 10−4 s−1 value obtained in Heays et al. (2017) for the Solar radiation field at 1 au.

All Tables

Table 1

Targeted lines.

Table 2

GBT and NRT half-power beam widths (HPBWs) and efficiencies.

Table 3

GBT observations: line intensities and spectral characteristics.

Table 4

OH observations at NRT: line intensities and spectral characteristics.

Table B.1

Abundances of ammonia in comets from ground-based observations.

Table B.2

Abundances of ammonia in comets from in situ measurements.

All Figures

thumbnail Fig. 1

Comet C/2020 F3 (NEOWISE) as seen from the Blue Ridge Mountains outside of White Hall, Virginia on 20 July 2020. The image is an aligned and stacked average of five, 5-min exposures taken using a Nikon D750 DLSR camera with a 50 mm f /1.8 lens held on an equatorial tracking mount (B. A. McGuire). The image is approximately 29.3° × 22.3° in total. The 1° scale is indicated in the lower right corner, which corresponds to 3 × 104 km at the geocentric distance of the comet on this date. A finder chart for this image is available at https://nova.astrometry.net/user_images/7997658# (Lang et al. 2010).

In the text
thumbnail Fig. 2

OH 1665 (left) and 1667 MHz (right) lines observed with the GBT (top) and the NRT (bottom) on 24 July 2020 in solid black lines. The red dashed curves show the fitted trapeziums to the spectra based on the methodology of Bockelée-Morvan et al. (1990). For GBT (NRT), the half lower bases of the trapeziums are 2.32 ± 0.07 (2.30 ± 0.14) and 2.43 ± 0.03 (2.23 ± 0.11) km s−1 for the 1665 and 1667 MHz lines, respectively.

In the text
thumbnail Fig. 3

OH 1667 MHz line observed with the GBT on 11.88 August 2020.

In the text
thumbnail Fig. 4

Weighted averages of the OH 1665 and 16 657 MHz lines observed at the NRT, normalized to the intensity of the 1667 MHz (Sect. 2.3). The spectra are integrated over specified periods of time, which are indicated above the plots in yymmdd format. The observations are fitted with trapeziums based on the methodology of Bockelée-Morvan et al. (1990).

In the text
thumbnail Fig. 5

Derived OH production rates from the GBT (red curve) and NRT (blue curves) observations of the OH 1667 MHz line in comet C/2020 F3 (NEOWISE) on 24 July 2020. OH production rates are plotted as a function of the quenching radius. The dashed lines correspond to the ±1σ uncertainty in the line area measured at NRT. Analogous dashed lines for the GBT are indistinguishable from the solid curve due to the high signal-to-noise ratio of the data. The maser inversion is i = 0.20 based on Despois et al. (1981), and the background temperature is Tbg = 3.1 K. The OH spatial distribution is described by the Haser-equivalent model (Combi & Delsemme 1980).

In the text
thumbnail Fig. 6

Water production rates of comet C/2020 F3 (NEOWISE). Black dots are values from this work and include the revised value from the Arecibo data: filled and open dots refer to values obtained with the inversion models of Schleicher & A’Hearn (1988) and Despois et al. (1981), respectively. Red dots: values from near-IR observations from Faggi et al. (2021). The dashed blue lines are a fit to the pre- and post-perihelion heliocentric values derived from Ly-a data (Combi et al. 2021).

In the text
thumbnail Fig. 7

NH3 abundances relative to water measured in comets from centimeter and submillimeter observations (hollow symbols), and near-IR observations (filled symbols). Different colors are used for Jupiter-family comets, dynamically new comets from the Oort Cloud, and long-period comets originating from the Oort Cloud. Downward-pointing arrowheads are 3σ upper limits. Dark blue upward-pointing triangles refer to the measurements of comet F3 from Faggi et al. (2021). The dark blue downward-pointing arrowhead is the value derived in this work. All data used for this figure are tabulated in Tables B.1 and B.2 with the corresponding references.

In the text
thumbnail Fig. A.1:

Four OH 18-cm lines observed in comet F3 with the NRT averaged from 18 to 27 July 2020. The expected statistical relative intensities of the 1 667 : 1 665 : 1 612 : 1 720 MHz lines are 9 : 5 : 1 : 1, which is in good agreement with the displayed observations.

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.