ALMA view of the circumnuclear disk of the Galactic Center: tidally disrupted molecular clouds falling to the Galactic Center

Abstract

1 Introduction

The Galactic Center is the nuclear region of the nearest spiral galaxy, Milky Way Galaxy. Sagittarius A* (Sgr A*) is a counterpart of the Galactic Center black hole (GCBH) in the regime from radio to X-ray; it is located very near the dynamical center of the galaxy (Reid et al. 2003) and has a mass of ∼4 × 10⁶ M_⊙ (e.g., Ghez et al. 2008; Gillessen et al. 2009; Boehle et al. 2016). The “circumnuclear disk” (CND) has been identified conventionally as a torus-like molecular gas around Sgr A* that is easily observed by high-J molecular emission lines and dust continuum (e.g., Güsten et al. 1987; Jackson et al. 1993; Marr et al. 1993; Christopher et al. 2005; Montero-Castaño et al. 2009; Martín et al. 2012). The overall kinematics of the CND had been interpreted as rotation around Sgr A* with a velocity of ∼100 km s⁻¹. However, the recently observed kinematics of the CND cannot be fully explained only by the rotation (e.g., Jackson et al. 1993; Shukla et al. 2004; Christopher et al. 2005; Montero-Castaño et al. 2009; Martín et al. 2012). How the molecular gas is supplied to the CND remains a controversial topic (e.g., Montero-Castaño et al. 2009; Takekawa et al. 2017; Hsieh et al. 2017). Another prominent structure in the vicinity of Sgr A* is the “Galactic Center mini-spiral” (GCMS), which is a bundle of ionized gas streams orbiting around Sgr A* (e.g., Lacy et al. 1980, 1991; Ekers et al. 1983; Lo & Claussen 1983; Serabyn & Lacy 1985; Serabyn et al. 1988; Roberts et al. 1996; Scoville et al. 2003; Zhao et al. 2009, 2010; Tsuboi et al. 2017) and is likely feeding the GCBH. There remains considerable controversy about the origin and lifetime of the GCMS. The relation between these structures is still an open question, although the GCMS is located along the inner ridge of the CND.

The CND itself has been observed intensively by the existing millimeter interferometers and IR telescopes, as mentioned above, while the outer region surrounding the CND has not been sufficiently observed because of their insufficient sensitivities and small field of view. In addition, the molecular gas in the outer region would be detected better in low-J molecular emission lines because of their lower excitation (e.g., Wright et al. 2001), although the CND itself is prominent in high-J molecular emission lines. The Atacama Large Millimeter/Submillimeter Array (ALMA) can detect both the CND itself and the molecular gas in the outer surrounding region, because ALMA has an unprecedentedly high sensitivity compared to previous telescopes. The more detailed gas kinematics of the wider area including the CND must provide key information to address the issues above mentioned. Therefore, we have observed the molecular gas in the wider region using ALMA.

The distance to the Galactic Center is assumed to be 8 kpc in this paper (e.g., Ghez et al. 2008; Gillessen et al. 2009; Schödel et al. 2009; Boehle et al. 2016). Then, corresponds to about 0.04 pc at the distance. We use Galactic coordinates. The directions used in this paper, for example “north” or “south,” are with reference to Galactic coordinates. We also use the traditional nomenclature of the substructures of the GCMS because it has been established, even though it has been named with reference to equatorial coordinates.

2 Observation and data reduction

We have produced a 136-pointing mosaic of the 12 m array and a 68-pointing mosaic of the 7 m array (ACA) which cover a 330΄ × 330΄ area including the CND and the “Galactic Center 50 km s⁻¹ molecular cloud” (50MC), a most conspicuous star-forming region in the vicinity of Sgr A*, in the CS J = 2–1 (97.980953 GHz), SiO v = 0 J = 2–1 (86.846995 GHz), H¹³CO⁺J = 1–0 (86.754288 GHz), C³⁴S J = 2–1 (96.41950 GHz), and CH₃OH |$J_{K_a, K_c}=2_{1,1}-1_{1,0}A_{--} (97.582808\:$|GHz) emission lines in ALMA Cy.1 (2012.1.00080.S, PI: Tsuboi, M.). The CS J = 2–1 and H¹³CO⁺J = 1–0 emission lines are dense molecular gas tracers with n(H₂) ≳ 10⁴ cm⁻³ and n(H₂) ≳ 10⁵ cm⁻³, respectively. The C³⁴S J = 2–1 emission line also traces the dense gas with n(H₂) ≳ 10⁴ cm⁻³, but is expected to be optically thin in this region, in contrast to the optically thick CS J = 2–1 emission line. The SiO J = 2–1 emission line is a tracer of strong C-shock with ΔV ≳ 30 km s⁻¹ in molecular clouds (e.g., Gusdorf et al. 2008; Jiménez-Serra et al. 2008), while the CH₃OH emission line is known as a tracer of weak shocks (ΔV ∼ 10 km s⁻¹; e.g., Hartquist et al. 1995). The frequency range in our observation also includes the H42α recombination line (85.6884 GHz), which is a tracer of ionized gas (e.g., Tsuboi et al. 2017).

The resultant maps have angular resolutions of (⁠|${2.^{\prime \prime }3}$|–|${2.^{\prime \prime }5}$|⁠) × (⁠|${1.^{\prime \prime }6}$|–|${1.^{\prime \prime }8}$|⁠), |$\mathit {PA} \sim -30^\circ$| using “natural weighting” sampling on the spatial frequency (u–v) plane, which corresponds to (0.09–0.10) pc × (0.06–0.07) pc at the Galactic Center distance. The synthesized beams of this observation are approximately four times smaller than those of previous molecular line observations (e.g., Montero-Castaño et al. 2009; Martín et al. 2012). The original velocity resolution is 1.7 km s⁻¹ (488 kHz). J0006−0623, J1517−2422, J717−3342, J1733−1304, J1743−3058, J1744−3116, and J2148+0657 were used as phase calibrators. The flux density scale was determined using Titan, Neptune, and Mars. The calibration and imaging of the data were performed by Common Astronomy Software Applications (CASA: McMullin et al. 2007). We made the spectral line images with ΔV = 5 km s⁻¹ of the emission lines mentioned above. The rms noise levels of the resultant maps are ∼0.002 Jy beam⁻¹ × 5 km s⁻¹ or ∼0.35 K km s⁻¹. Because the observation has a large time span of one year and seven months, the flux uncertainty is as large as 15%. We will present the full data of the observation and the analysis of 50MC in a future paper (Uehara et al. 2018).

For comparison, we analyzed the mosaic observation data of the CND with ALMA in the CS J = 7–6 (342.882857 GHz) emission line obtained from the JVO portal of NAOJ (ALMA#2012.1.00543.S). The data of the CS J = 7–6 emission line have an angular resolution of |${4.^{\prime \prime }3}$| × |${2.^{\prime \prime }7}$|⁠, |$\mathit {PA} \sim -35^\circ$| using “natural weighting” sampling on the u–v plane, which corresponds to 0.17 pc × 0.11 pc at the Galactic Center distance. The analysis of the data were also performed by CASA.

3 Results

3.1 Integrated intensity (moment 0) maps

Figure 1 shows the integrated intensity (moment 0) maps of the CND (red dashed ellipse) and its surrounding region in the CS J = 2–1, C³⁴S J = 2–1, SiO v = 0, J = 2–1, H¹³CO⁺J = 1–0, and CH₃OH |$J_{K_a, K-c}=2_{1,1}-1_{1,0}A_{--}$| and H42α emission lines. The integrated velocity range of the panels is V_LSR = −150 to 150 km s⁻¹. Because the previous observations show that the CND has a rotation velocity of 100–110 km s⁻¹ and a random or radial velocity of 10–30 km s⁻¹ (e.g., Güsten et al. 1987; Jackson et al. 1993; Christopher et al. 2005), the integrated velocity range contains almost all of the total velocity extent of the CND. These panels are cut from the large-area mosaic data mentioned in the previous section. The contours in the figure show the continuum emission at 100 GHz, which mainly traces the GCMS, for comparison (Tsuboi et al. 2016). Figure 1 also shows the moment 0 map of the CS J = 7–6 emission line by ALMA and the continuum map at 670 GHz by the James Clerk Maxwell Telescope (JCMT) for comparison. Although a mosaic map of the CS J = 7–6 emission line with a similar angular resolution has been obtained by the Submillimeter Array (SMA) (Montero-Castaño et al. 2009), the sensitivities of our maps are five times or more higher than those of the SMA maps. The CND is identified as a fragmented ring surrounding the GCMS in the CS J = 2–1, C³⁴S, SiO, and H¹³CO⁺ emission lines (cf. Montero-Castaño et al. 2009; Martín et al. 2012), and is less prominent in the CH₃OH emission line. Note that the CH₃OH emission line seems to trace the surrounding gas rather than the CND itself. There are many extended components surrounding the CND in the CS J = 2–1, C³⁴S, SiO, and H¹³CO⁺ lines, whereas the CS J = 7–6 line and 670 GHz continuum emissions mainly trace the CND.

Open in new tabDownload slide

Open in new tabDownload slide

Fig. 1.

Integrated intensity (moment 0) maps of the CND (the red dashed ellipse shows the elliptical outline of the CND with an assumed radius = 58, PA = 67°, and inclination =30°) and the surrounding region in the (a) CS J = 2–1, (b) C³⁴S J = 2–1, (c) SiO v = 0, J = 2–1, (d) H¹³CO⁺J = 1–0, (e) CH₃OH |$J_{K_a, K_c}=2_{1,1}$|–1_{1, 0}A₋₋, and (f) H42α emission lines. The integrated velocity range is V_LSR = −150 to 150 km s⁻¹. The FWHM beam sizes are (⁠|${2.^{\prime \prime }3}$|–|${2.^{\prime \prime }5}$|⁠) × (⁠|${1.^{\prime \prime }6}$|–|${1.^{\prime \prime }8}$|⁠), |$\mathit {PA} \sim -30^\circ$| shown at the lower-left corners of the panels as ovals. The contours in each panel show the continuum emission at 100 GHz of the GCMS for comparison (Tsuboi et al. 2016). They are set at 3.75, 7.5, 15, 30, 60, 120, 240, 480, and 960 mJy beam⁻¹. (g) Integrated intensity (moment 0) map in the CS J = 7–6 emission line. The data has an angular resolution of |${4.^{\prime \prime }6} \times {2.^{\prime \prime }7}$|⁠, |$\mathit {PA} \sim -35^\circ$|⁠. (h) Continuum map at 670 GHz, obtained from the JCMT Science Archive. The data has an angular resolution of 8΄. (Color online)

3.2 Channel maps

Figures 2, 3, 4, 5, and 6 show the channel maps of the same region as figure 1, with central velocities V_{c, LSR} = −142.5 to +147.5 km s⁻¹ in the CS, C³⁴S, SiO, H¹³CO⁺, and CH₃OH emission lines, respectively. The velocity width of each panel is 10 km s⁻¹. Figure 7 shows the channel maps with the same velocity range in the CS J = 7–6 emission line for comparison.

Open in new tabDownload slide

Open in new tabDownload slide

Fig. 2.

Channel maps of the CND (red dashed ellipse) and its surrounding region in the CS J = 2–1 emission line. The central velocity range is from V_{c, LSR} = −142.5 to +147.5 km s⁻¹, and the velocity width of each panel is 10 km s⁻¹. The angular resolution is |${2.^{\prime \prime }3}$| × |${1.^{\prime \prime }6}$| in FWHM, which is shown as the oval at the lower-left corner of each panel. The contours in the figure show the continuum emission of the GCMS at 100 GHz for comparison (Tsuboi et al. 2016). They are set at 3.75, 7.5, 15, 30, 60, 120, 240, 480, and 960 mJy beam⁻¹. (Color online)

Open in new tabDownload slide

Open in new tabDownload slide

Fig. 3.

Channel maps of the CND (red dashed ellipse) and its surrounding region in the C³⁴S J = 2–1 emission line. The central velocity range is from V_{c, LSR} = −142.5 to +147.5 km s⁻¹, and the velocity width of each panel is 10 km s⁻¹. The angular resolution is |${2.^{\prime \prime }3} \times {1.^{\prime \prime }7}$| in FWHM, which is shown as the oval at the lower-left corner of each panel. The contours in the figure show the continuum emission of the GCMS at 100 GHz for comparison (Tsuboi et al. 2016). They are set at 3.75, 7.5, 15, 30, 60, 120, 240, 480, and 960 mJy beam⁻¹. (Color online)

Open in new tabDownload slide

Open in new tabDownload slide

Fig. 4.

Channel maps of the CND (red dashed ellipse) and its surrounding region in the SiO v = 0, J = 2–1 emission line. The central velocity range is from V_{c, LSR} = −142.5 to +147.5 km s⁻¹, and the velocity width of each panel is 10 km s⁻¹. The angular resolution is |${2.^{\prime \prime }5} \times {1.^{\prime \prime }8}$| in FWHM, which is shown as the oval at the lower-left corner of each panel. The contours in the figure show the continuum emission of the GCMS at 100 GHz for comparison (Tsuboi et al. 2016). They are set at 3.75, 7.5, 15, 30, 60, 120, 240, 480, and 960 mJy beam⁻¹. (Color online)

Open in new tabDownload slide

Open in new tabDownload slide

Fig. 5.

Channel maps of the CND (red dashed ellipse) and its surrounding region in the H¹³CO⁺J = 1–0 emission line. The central velocity range is from V_{c, LSR} = −142.5 to +147.5 km s⁻¹, and the velocity width of each panel is 10 km s⁻¹. The angular resolution is |${2.^{\prime \prime }5} \times {1.^{\prime \prime }8}$| in FWHM, which is shown as the oval at the lower-left corner of each panel. The contours in the figure show the continuum emission of the GCMS at 100 GHz for comparison (Tsuboi et al. 2016). They are set at 3.75, 7.5, 15, 30, 60, 120, 240, 480, and 960 mJy beam⁻¹. (Color online)

Open in new tabDownload slide

Open in new tabDownload slide

Fig. 6.

Channel maps of the CND (red dashed ellipse) and its surrounding region in the CH₃OH emission line. The central velocity range is from V_{c, LSR} = −142.5 to +147.5 km s⁻¹, and the velocity width of each panel is 10 km s⁻¹. The angular resolution is |${2.^{\prime \prime }3}$| × |${1.^{\prime \prime }7}$| in FWHM, which is shown as the oval at the lower-left corner of each panel. The contours in the figure show the continuum emission of the GCMS at 100 GHz for comparison (Tsuboi et al. 2016). They are set at 3.75, 7.5, 15, 30, 60, 120, 240, 480, and 960 mJy beam⁻¹. (Color online)

Open in new tabDownload slide

Open in new tabDownload slide

Fig. 7.

Channel maps of the CND (red dashed ellipse) and its surrounding region in the CS J = 7–6 emission line. The central velocity range is from V_{c, LSR} = −142.5 to +147.5 km s⁻¹, and the velocity width of each panel is 10 km s⁻¹. The angular resolution is |${4.^{\prime \prime }6} \times {2.^{\prime \prime }7}$| in FWHM, which is shown as the oval at the lower-left corner of each panel. The contours in the figure show the continuum emission of the GCMS at 100 GHz for comparison (Tsuboi et al. 2016). They are set at 3.75, 7.5, 15, 30, 60, 120, 240, 480, and 960 mJy beam⁻¹.

Although the components of the CND identified in these channel maps are roughly consistent with those in the previous observations (e.g., Güsten et al. 1987; Jackson et al. 1993; Marr et al. 1993; Shukla et al. 2004; Montero-Castaño et al. 2009; Martín et al. 2012), significant differences are found in the different emission lines (cf. Martín et al. 2012). The kinematics of the CND do not always follow a simple rotation law around Sgr A*, as mentioned in section 1. There is a feature apparently distributed just outside the “western arc” (WA) of the GCMS (see also figure 1f). This feature is also prominent in the integrated intensity maps of several previous papers (e.g., Montero-Castaño et al. 2009; Martín et al. 2012). However, the feature has two distinct velocity components with negative and positive radial velocities. The positive velocity component will be discussed separately, in sub-subsection 3.2.1. The negative velocity component is identified in the panels in the CS emission line of V_{c, LSR} = −142.5 to +7.5 km s⁻¹. The component is also identified in the C³⁴S, SiO, and H¹³CO⁺ emission lines, while the component cannot be identified in the CH₃OH emission line. The peak position of the component is seen to be shifting from west to east along the outer edge of the WA with increasing velocity. A similar velocity shift has been observed for the WA in the H42α emission line (e.g., Tsuboi et al. 2017). This component would therefore be physically associated with the WA and follow a simple rotation law around Sgr A*. The component corresponds to the “northwestern sector” of the CND.

Another component is identified on the opposite side of the GCMS in the velocity range V_{c, LSR} ∼ −82.5 to +7.5 km s⁻¹. This component is seen to correspond to the “southwestern sector” of the CND. The peak position of the component would be also shifting from west to east with increasing velocity, although the component has no clear association with the ionized gas of the GCMS. The “northwestern sector” and “southwestern sector” form the “western half” of the CND, which is also identified clearly in the CS J = 7–6 emission line.

Meanwhile, the “eastern half” of the CND has a complicated structure. Several components with positive radial velocity are seen in the eastern area of the CND. However, we should keep in mind that all the molecular clouds seen toward the CND do not always belong to it, because this is a very overlapping region, as shown in figure 1. The molecular clouds coming from outer regions and intruding into the CND may be molecular gas flows feeding the CND. On the other hand, the molecular clouds passing the CND from one outer region to another outer region are thought not to belong to the CND even if they are seen toward the CND. The distinction between the CND components and clouds overlapping the CND by chance in the same line of sight is necessary to obtain the true image of the CND.

3.2.1 Anomaly A

The positive velocity component apparently distributed along the WA is prominent in the panels of the CS emission line with V_{c, LSR} = +67.5 to +117.5 km s⁻¹. The component is also identified in the C³⁴S and SiO emission lines, while it is faint in the H¹³CO⁺ emission line. The “southwestern half” of the component overlaps positionally with the northwestern sector of the CND, as mentioned above. The velocity of the component is different from that of ionized gas in the WA (e.g., Tsuboi et al. 2017). Thus, this component would not be physically associated with the WA. The component has also been identified conventionally as part of the CND, although the positive radial velocity in the western half of the CND means that the component should rotate in the opposite direction around Sgr A* if it really belongs to the CND. This component is called Anomaly A (AA) hereafter.

AA extends to the northeast in the panels of the CS emission line with V_{c, LSR} = 67.5 and 87.5 km s⁻¹ (see figure 2). This seems to cross another component extending from southeast to northwest (Anomaly B, see below) around |$l\sim {359{^{\circ}_{.}}955}$|⁠, |$b\sim - {0{^{\circ}_{.}}030}$| in the panel of V_{c, LSR} = 87.5 km s⁻¹. In the CH₃OH emission line, the “northeastern half” of the component with V_{c, LSR} = +77.5 to +87.5 km s⁻¹ is especially prominent, although the “southwestern half” is faint. Because the CH₃OH emission line is known as a weak shock a weak shock tracer, as mentioned in section 2, the “northeastern half” seems abundant in mildly shocked molecular gas (ΔV ∼ 10 km s⁻¹).

3.2.2 Anomaly B

The panels with V_{c, LSR} = +87.5 to +117.5 km s⁻¹ in the CS emission line show that an elongated molecular cloud seems to approach Sgr A* along an arc from the north to the outer end of the eastern arm (EA) of the GCMS with increasing velocity (see also figure 1f). The component is most prominent in the CS emission line, and is identified partly in the C³⁴S, SiO, H¹³CO⁺, and CH₃OH emission lines. The component appears to be a bundle of filamentary substructures. Although the component has been identified as a small protrusion from the CND since the early days of CND observation using the low-J emission lines (e.g., Güsten et al. 1987; Zylka et al. 1990; Wright et al. 2001), it has never been revealed as a bundle of filamentary substructures by existing telescopes because it is deeply embedded in the CND. Note that the component does not obey the simple rotation law of the CND. This component is called Anomaly B (AB) hereafter. The south end of AB seems to be continuously connected with the EA. On the other hand, the north end of AB is located beyond the CND. The filamentary substructures remarkable in the CS emission line are faint in the SiO and H¹³CO⁺ emission lines, except for the contact area between AB and the EA, |$l \sim {359{^{\circ}_{.}}958}$|⁠, |$b \sim {-0{^{\circ}_{.}}045}$|⁠. This suggests that strongly shocked molecular gas with ΔV ≳ 30 km s⁻¹ and n(H₂) ≳ 10⁵ cm⁻³ is abundant in the contact area. Because the northern half of AB is prominent in the CH₃OH emission line, the northern half seems abundant in mildly shocked molecular gas (ΔV ∼ 10 km s⁻¹).

Meanwhile, the southern half of AB is not seen in the CH₃OH emission line. The CH₃OH molecule is reported to be destroyed easily through the cosmic-ray photodissociation reaction in the vicinity of Sgr A* [CH₃OH→H₂CO + H₂ (Harada et al. 2015)]. This is probably why the CND itself is faint in the CH₃OH emission line. This suggests that the southern half of AB may be located near Sgr A* and the northern half is really farther from it. Similarly, the CH₃OH maps suggest that the northeastern half of AA is farther from Sgr A* and the southwestern half is located near Sgr A*.

3.2.3 Anomaly C

The panels with V_{c, LSR} = 77.5 km s⁻¹ in the CS, C³⁴S, and CH₃OH emission lines show that a nearly vertical structure is located around |$l\sim {359{^{\circ}_{.}}948}$|⁠, |$b\sim -{0{^{\circ}_{.}}032}$| (see figures 2, 3, and 6). This component is identified faintly in the SiO emission line (see figure 4), but is not identified in the H¹³CO⁺ and CS J = 7–6 emission lines. The south end of this component crosses AA. The south end seems to be continuously connected with the horizontal arm of the GCMS (see figure 1f). This component is called Anomaly C (AC) hereafter.

3.2.4 Anomaly D

The panels with V_{c, LSR} = −72.5 to −32.5 km s⁻¹ in the CS emission line show that a compact molecular cloud is located around l ∼ 359|${^{\circ}_{.}}$|945, b ∼ −0|${^{\circ}_{.}}$|035 (see figure 2). The component is prominent only in the CS emission line. This component is identified partly in the C³⁴S emission line, but is faint in the SiO, H¹³CO⁺, and CH₃OH emission lines (see figures 3, 4, 5, and 6). The negative radial velocity at the eastern half of the CND means that the component should rotate in the opposite direction around Sgr A* if it belongs to the CND. Because this component is faint in the CH₃OH emission line, the cloud is likely located near Sgr A*, as it appears. In addition, the southern edge of the component is detected faintly in the H42α recombination line (see figure 1f). This component is called Anomaly D (AD) hereafter.

3.2.5 −80 km s⁻¹ molecular cloud

The panels with V_{c, LSR} = −102.5 to −42.5 km s⁻¹ in the CS emission line show that an isolated molecular cloud is located around |$l\sim {359{^{\circ}_{.}}963}$|⁠, |$b\sim -{0{^{\circ}_{.}}065}$|⁠. This component is also identified in the channel maps of the C³⁴S, SiO, H¹³CO⁺, and CH₃OH emission lines. This component has a half-shell-like appearance in the panels with V_{c, LSR} = −82.5 to −62.5 km s⁻¹ of these emission lines. As mentioned above, the prominent component in the CH₃OH emission line suggests that the cloud is farther from Sgr A*. The component would be independent from the CND. Because the intensity ratio of SiO/H¹³CO⁺ is 3–4 in the component, shocked molecular gas is likely abundant in it. This component has no counterpart in the H42α recombination line (see figure 1f), indicating no ionized gas in it. This component is called the −80 km s⁻¹ molecular cloud (−80MC) hereafter.

3.2.6 −70 km s⁻¹ molecular cloud

The panels with V_{c, LSR} = −82.5 to −52.5 km s⁻¹ in the CS emission line show that a molecular cloud is located around |$l\sim {359{^{\circ}_{.}}969}$|⁠, |$b\sim -{0{^{\circ}_{.}}030}$|⁠. Although this component is also identified clearly in the channel maps of the C³⁴S emission line, it is faint in the SiO, H¹³CO⁺, and CH₃OH emission lines. This component has no counterpart in the H42α recombination line (see figure 1f), indicating no ionized gas in it.

This component is called the −70 km s⁻¹ molecular cloud (−70MC) hereafter.

4 Discussion

4.1 Kinematics of the circumnuclear disk

As mentioned in section 1, many previous works have shown that the CND is a torus-like collection of molecular gas with an inner boundary radius of r_in ∼ 1.5–1.6 pc with a rotation velocity of V_rot ∼ 100–120 km s⁻¹ around Sgr A* (e.g., Güsten et al. 1987; Jackson et al. 1993; Christopher et al. 2005; Montero-Castaño et al. 2009; Martín et al. 2012). In addition, the velocity of the radial motion in the CND has also been estimated to be |$\bar{V}_\mathrm{rad} \sim 20$|–50 km s⁻¹ in these works. However, the observed kinematics of the CND have not been fully explained by only the rotation with radial motion, and the outer boundary has not been clarified.

The inner radius of the CND is estimated to be R_in ∼ 1.5 pc from the appearances in the moment 0 maps (see figures 8a, 8c, 8e, and 8g) and the channel maps in the CS J = 2–1, SiO v = 0, J = 2–1, H¹³CO⁺J = 1–0, and CS J = 7–6 emission lines (see figures 2, 4, 5, and 7). There are many outer components that do not obey the rotation law with radial motion in the position–velocity diagrams (see figure 8) besides the anomalies mentioned in the previous section. The outer components are distributed mainly in the velocity range of V_LSR ∼ −30 to +70 km s⁻¹ (see figure 8). These components would not belong to the CND (see figures 2, 4, 5, and 7) because they are identified to be large features outside the CND, for example “northeast arm,” “linear filament,” and so on; they have been recognized as parts of the CND in previous observations (e.g., Montero-Castaño et al. 2009; Martín et al. 2012). In this case, the outer radius is estimated to be as small as R_out ∼ 2 pc. The inclination angle of the CND is estimated to be i ∼ 30 ± 5° (i = 0° for an edge-on ring) from the aspect ratios of the appearances of the CND in the moment 0 maps.

Open in new tabDownload slide

Open in new tabDownload slide

Fig. 8.

(a) Moment 0 map in the CS J = 2–1 emission line as the finding chart of the CND (white dashed ellipse). The contours in the finding charts show the continuum emission of the GCMS at 100 GHz for comparison (Tsuboi et al. 2016). They are set at 3.75, 7.5, 15, 30, 60, 120, 240, 480, and 960 mJy beam⁻¹. (b) Position–velocity (P–V) diagram along the major axis of the CND in the CS J = 2–1 emission line. The sampling areas of the diagrams are shown as the red rectangles in the finding chart. The pseudo-color and contours show the components in the northern and southern areas, respectively. The contour levels are set at 10.3, 20.6, 30.9, 41.2, 51.5, 61.8, and 82.4 mJy beam⁻¹. The white dashed parallelogram indicates the CND conventionally identified in the P–V diagram (e.g., Güsten et al. 1987; Jackson et al. 1993; Christopher et al. 2005; Montero-Castaño et al. 2009; Martín et al. 2012). “A” and “B” indicate the anomalies A and B of the CND shown in figures 2, 3, 4, 5, and 6. (c) Moment 0 map in the SiO v = 0, J = 2–1 emission line as the finding chart of the CND (white dashed ellipse). (d) P–V diagram along the major axis of the CND in the SiO v = 0, J = 2–1 emission line. The contour levels are set at 2, 4, 6, 8, 10, 12, and 16 mJy beam⁻¹. (e) Moment 0 map in the H¹³CO⁺J = 1–0 emission line as the finding chart of the CND (white dashed ellipse). The contours in the finding charts show the continuum emission of the GCMS at 100 GHz for comparison (Tsuboi et al. 2016). They are set at 3.75, 7.5, 15, 30, 60, 120, 240, 480, and 960 mJy beam⁻¹. (f) P–V diagram along the major axis of the CND in the H¹³CO⁺J = 1–0 emission line. The contour levels are set at 1.1, 2.2, 3.3, 4.4, 5.5, 6.6, and 8.8 mJy beam⁻¹. (g) Moment 0 map in the CS J = 7–6 emission line as the finding chart of the CND (white dashed ellipse). (h) P–V diagram along the major axis of the CND in the CS J = 7–6 emission line. The contour levels are set at 0.2, 0.4, 0.6, 0.8, 1.0, 1.2, and 1.6 Jy beam⁻¹. (Color online)

The kinematics of the CND are shown in the position–velocity diagrams along the major axis in the CS J = 2–1, SiO v = 0, J = 2–1, H¹³CO⁺J = 1–0, and CS J = 7–6 emission lines (see figures 8b, 8d, 8f, and 8h). Because the CND is not clear in the CH₃OH emission line, as mentioned in subsection 3.1, the position–velocity diagram is not shown here. The CND is identified as a series of components which are roughly along an inclined line in the diagrams (broken-line parallelograms). These have been observed in previous observations (e.g., Montero-Castaño et al. 2009; Martín et al. 2012). From the position–velocity diagrams (see figures 8b, 8d, 8f, and 8h), the rotation velocity and radial motion of the CND are estimated to be |$V_{\mathrm{rot}}\sim \frac{100}{\cos i} =115\pm 10\:$|km s⁻¹ and |$V_\mathrm{rad}\sim \frac{20}{\cos i}=23\pm 5\:$|km s⁻¹, respectively. The derived values are consistent with those in the previous observations (e.g., Güsten et al. 1987; Marr et al. 1993; Montero-Castaño et al. 2009; Martín et al. 2012). Figure 9 shows the Keplerian orbit with the derived inner radius, outer radius, inclination angle, and radial motion. An enclosed mass of 4.3 × 10⁶ M_⊙ reproduces the observed position–velocity diagrams.

Fig. 9.

Upper panel: Trajectories of the Keplerian motions with an enclosed mass of 4.3 × 10⁶ M_⊙, a radius of 1.5–2.0 pc, an eccentricity of 0, and radial motion of 23 km s⁻¹. The inclination angle is i = 30°. Lower panel: Position–velocity diagram of the trajectories along their major axis.

Open in new tabDownload slide

4.2 Physical properties of the circumnuclear disk

It has been reported that the moderately dense molecular gas in the CND with |$\log {N_\mathrm{H_2}}\sim 4.5$| has kinetic temperatures of T_K ∼ 200 K, and the slightly denser gas with |$\log {N_ \mathrm{H_2}}\sim 5.5$| has kinetic temperatures of T_K ∼ 300–500 K (Requena-Torres et al. 2012). It has also been reported that the CND would be dominated by the former component. The kinetic temperature of the molecular gas including CS molecules is assumed to be T_K = 200 K here.

We calculate the brightness and excitation temperatures of the CS J = 2–1 and CS J = 7–6 emission lines using the RADEX LVG program (Van der Tak et al. 2007) to estimate the physical properties of the CND. Figure 10a shows the relations of the CS J = 2–1 brightness temperature, T_B(CS J = 2–1), and the brightness temperature ratio, T_B(CS J = 7–6)/T_B(CS J = 2–1), on the plane of H₂ number density versus CS fractional abundance per velocity gradient, are calculated at T_K = 200 K. Figure 10b shows the excitation temperatures of the CS J = 2–1 and CS J = 7–6 emission lines, T_ex(CS J = 2–1) and T_ex(CS J = 7–6), on the same plane. The observed intensity peaks of the CS J = 2–1 brightness temperature in the CND and Anomaly A are plotted in the figures as filled and open circles, respectively. They are clearly separated in the figure. This indicates that the physical properties of the CND and AA are different from each other.

Fig. 10.

(a) Relation of T_B(CS J = 2–1) (black contours) vs. T_B(CS J = 7–6)/T_B(CS J = 2–1) (red contours) at T_K = 200 K, calculated by the RADEX LVG program (Van der Tak et al. 2007). The observed intensity peaks of the CS J = 2–1 brightness temperature in the CND and Anomaly A are plotted as filled and open circles, respectively. (b) Excitation temperatures of CS J = 2–1 (black contours) and CS J = 7–6 (red contours) emission lines at T_K = 200 K. (Color online)

Open in new tabDownload slide

In figure 10b, the data points of the CND are distributed above the boundary where the excitation temperature of the CS J = 2–1 emission line is equal to the kinetic temperature. The calculated excitation temperature often has a negative sign, which nominally indicates that the emission is a maser one. However, this is likely an artifact made by a simplified velocity structure. Because the CS spectra in the CND do not show the very high and very narrow profiles typical in maser emissions, it is reasonable that the J = 2 level of the CS molecules is thermalized. Then we consider that the excitation temperature is equal to the kinetic temperature of T_K = 200 K assumed above. On the other hand, the excitation temperature of the CS J = 7–6 emission line in the CND is in the range of T_ex(CS J = 7–6) = 18–25 K. This indicates that the J = 7 level of the CS molecules is sub-thermally excited. In addition, the excitation temperatures of the CS J = 2–1 and CS J = 7–6 emission lines in AA are in the ranges of T_ex(CS J = 2–1) = 56–200 K and T_ex(CS J = 7–6) = 16–20 K, respectively.

In figure 10a, the mean H₂ number density and CS fractional abundance per velocity gradient of the CND are estimated to be |$n_{\mathrm{H_2}}\sim 2.2\times 10^{5}\:$|cm⁻³ and |$\log \left[X(\mathrm{CS})\left(\frac{dV}{dr}\right)^{-1}\right]\sim -10.8$|⁠, respectively. Assuming that the CS fractional abundance and the velocity width of the CS emission line are X(CS) = 10⁻⁸ and ΔV = 50 km s⁻¹, the path length of the molecular gas is estimated to be Δl ∼ 0.1 pc by |$X(\mathrm{CS})\left(\frac{dV}{dr}\right)^{-1} \sim X(\mathrm{CS})\left(\frac{\Delta l}{\Delta V}\right)$|⁠. The molecular column density based on the LVG model is derived to be |$N{\mathrm{_{LVG}(H_2) }}\sim n_{\mathrm{H_2}}\times \Delta l =7\times 10^{22}\:$|cm⁻².

4.3 Molecular gas mass of the circumnuclear disk

The molecular gas mass of the CND would be key information in understanding its origin. In the previous observations, the molecular gas mass of the CND was derived based on dust continuum emission maps or integrated intensity maps of molecular emission lines, because the CND is identified easily as a ring-like feature surrounding Sgr A* in these maps. However, velocity-resolved analysis has been hard to perform owing to the lack of sensitivity. The molecular gas mass of the CND even in the recent dust observations is scattered widely from M = 0.6 × 10³ M_⊙ to 5 × 10⁴ M_⊙: 0.6 × 10³ M_⊙ by Stratospheric Observatory for Infrared Astronomy (SOFIA) (Lau et al. 2013), 1–2 × 10⁴ M_⊙ by SMA (Liu et al. 2013), 2 × 10⁴ M_⊙ by JCMT (e.g., Oka et al. 2011), and 5 × 10⁴ M_⊙ by the Herschel Space Observatory (Etxaluze et al. 2011). On the other hand, the molecular gas mass of the CND derived from molecular emission line observations has been not well determined, although a larger mass of M = 1–3 × 10⁵ M_⊙ has been reported (Smith & Wardle 2013). Because these works have different integration areas for the CND, the various CND masses do not always mean ambiguity in the derivation method. It is still necessary to examine the derivation method of the molecular gas mass and to clarify what is the boundary of the CND. This has been done in subsection 4.1.

4.3.1 CS J = 2–1 and C³⁴S J = 2–1 emission lines

The total corrected integrated line intensity in the CND of the CS J = 2–1 emission line is ΣI_corr.(CS) = 2.1 × 10³ Jy km s⁻¹, whereas that of the C³⁴S J = 2–1 emission line is ΣI(C³⁴S) = 8.8 × 10¹ Jy km s⁻¹. The total velocity range is from −150 to 150 km s⁻¹ in V_LSR for both emission lines. The mean line intensity ratio of the CND is calculated to be |$\frac{\Sigma I\mathrm{_{corr.}(CS)}}{\Sigma I(\mathrm{C^{34}S})}\sim 16.7$|⁠. Then, the typical optical thickness of the CS emission line of the CND is estimated to be |$\bar{\tau }\mathrm{(CND)}\sim 0.63$|⁠, whereas that of AA is as large as |$\bar{\tau }\mathrm{(AA)}\sim 1.5$|⁠. The LTE molecular gas mass of the CND is derived to be M_LTE(CND) = 3.1 × 10⁴(T_ex/200)M_⊙. The LTE mass is consistent with those derived in the previous observations mentioned above. In addition, the LTE molecular gas masses of AA, AB, and AC are derived to be M_LTE(AA) = 1.3 × 10⁴(T_ex/200) M_⊙, M_LTE(AB) = 7.0 × 10³(T_ex/200) M_⊙, and M_LTE(AC) = 1.0 × 10³(T_ex/200) M_⊙, respectively. However, the excitation temperature for these components may be lower than T_ex = 200 K (see figure 10b).

We also derive the LTE molecular gas mass of the CND based on the data of the CS J = 7–6 emission line by the same procedure, for comparison. The integrated line intensity in the CND of the CS J = 7–6 emission line is ΣI(CS J = 7–6) = 4.8 × 10⁵ Jy km s⁻¹. Here, the Einstein A coefficient of the CS J = 7–6 emission line is assumed to be A₇₆ = 0.8 × 10⁻³s⁻¹. The excitation temperature of the CS J = 7–6 emission line is calculated to be T_ex = 18–25 K using the RADEX LVG program (Van der Tak et al. 2007), as mentioned in the previous subsection (see figure 10b). Then, the LTE molecular gas mass derived from the CS J = 7–6 emission line becomes M_LTE(CND) = 1.5 × 10⁴(T_ex/22) M_⊙.

4.3.2 H¹³CO⁺J = 1–0 emission line

We derive the LTE molecular gas mass of the CND based on the observation in the H¹³CO⁺ emission line by the same procedure (see figure 5). We consider that the J = 1 level of the H¹³CO⁺ molecules is also thermalized, and the excitation temperature is equal to the kinetic temperature of T_K = 200 K. The Einstein A coefficient of the H¹³CO⁺ emission line is assumed to be A₁₀ = 3.9 × 10⁻⁵ s⁻¹.

The integrated line intensity in the CND of the H¹³CO⁺ emission line is ΣI(H¹³CO⁺) = 9.2 × 10¹ Jy km s⁻¹. Then, the LTE molecular gas mass of the CND from the H¹³CO⁺ emission line observation is derived to be M_LTE = 1.5 × 10⁴(T_ex/200) M_⊙, assuming a fractional abundance of the H¹³CO⁺ molecule of |$X\mathrm{(H^{13}CO^{+})}=\frac{N\mathrm{(H^{13}CO^{+})}}{N\mathrm{(H_2)}}=4\times 10^{-10}$| (for the CND; Amo-Baladrón et al. 2011). Although the fractional abundance for the CND is nine times larger than that for the Orion A molecular cloud, X(H¹³CO⁺) = 5 × 10⁻¹¹ (Ikeda et al. 2007), the consistency in the derived gas masses from the CS and H¹³CO⁺ observations suggests that the estimations of the excitation temperature and the fractional abundance are good for the CND.

The LTE mass of the CND derived from the CS J = 7–6 and H¹³CO⁺J = 1–0 emission lines, M_LTE ∼ 1.5 × 10⁴(T_ex/200) M_⊙, is half of that derived from the CS J = 2–1 emission line, M_LTE ∼ 3.1 × 10⁴(T_ex/200) M_⊙. These relations are consistent with the effective critical densities of the CS J = 7–6 and H¹³CO⁺J = 1–0 emission lines, n_eff ∼ 10⁵ cm⁻³, being higher than that of the CS J = 2–1 emission line, n_eff ∼ 10³ cm⁻³. The derived masses of tidally disrupted molecular clouds, which will be discussed in the next subsection, are also summarized in table 1.

4.4 Tidally disrupted molecular clouds falling to the Galactic Center

AA, AB, and AC could be clouds falling from the outer regions onto the CND or the GCMS because they apparently connect with the inner structures. The channel maps with V_{c, LSR} = 77.5 to 117.5 km s⁻¹ in the CS emission line show that AA has a continuous appearance from the northeast to the vicinity of WA (see figure 2). Although the northeastern part of AA is prominent in the CH₃OH emission line, the inner part apparently adjacent to WA is faint in the emission line (see the channel maps with V_{c, LSR} = 77.5 to 97.5 km s⁻¹ in the CH₃OH emission line in figure 6). On the other hand, although the northeastern part is faint in the SiO emission line, the inner part is prominent in the emission line (see figure 4). This part is also identified by SOFIA as a compact component in extremely high-J CO emission lines (up to J = 16–15) (Requena-Torres et al. 2012). Because the CH₃OH emission line is activated by shock with a milder shock velocity, ΔV ∼ 10 km s⁻¹, than that of the SiO emission line, the northeastern part is abundant in mildly shocked molecular gas. The SiO enhancement shows that the inner part is abundant in strongly shocked molecular gas. It is likely that the CH₃OH emission line of the inner part is also enhanced by the shock, but the observed weak line suggests the depletion of the CH₃OH molecules in the inner part.

The CH₃OH molecule is reported to be destroyed easily through the cosmic-ray photodissociation reaction in the vicinity of Sgr A*: CH₃OH + c.r.→H₂CO + H₂ (Harada et al. 2015). The disappearance of AA in the CH₃OH emission line suggests that the inner part of AA is really located near Sgr A* rather than being chance coincident. We argue that AA is falling from the outer region to the CND, being disrupted by the tidal shear of Sgr A* and affected by cosmic rays. AA is probably a route for feeding molecular gas to the CND. If this is the case, it is possible that the falling cloud does not obey the rotation law of the CND as observed because the orbit depends on the initial conditions including impact parameter, and the conditions would be various. The molecular cloud would currently be falling from the outer region to the CND because the dynamical relaxation by collisions with molecular gas in the CND has not yet occurred. In addition, the photodissociation by cosmic rays is probably why it is hard to depict the CND itself in the CH₃OH emission line (see figure 1e).

Figure 11 shows the positional relation between the molecular gas of AB in the CS emission line and the ionized gas of the eastern arm (EA) in the H42α recombination line. The ionized gas motion in EA has been explained as a part of a Keplarian elliptical orbit which is determined by the appearance and kinematics (e.g., Zhao et al. 2009, 2010; Tsuboi et al. 2017). AB and EA are identified as a continuous and united body, although they are made of different gas components.

Fig. 11.

Right panel: Integrated intensity map of Anomaly B at V_LSR = 97.5 km s⁻¹ in the CS J = 2–1 emission line (upper part) and the eastern arm at V_LSR = 87.5 km s⁻¹ in the H42α recombination line (lower part). The central velocities are shown at the upper-left and lower-right corners of the upper and lower parts, respectively. The velocity widths of the two parts are 10 km s⁻¹. The red dashed ellipse shows the elliptical outline of the CND. The white contours show the continuum emission at 100 GHz of the GCNS at 100 GHz for comparison (Tsuboi et al. 2016). Upper-left panel: P–V diagram along Anomaly B in the CS J = 2–1 emission line. The sampling area is shown by the upper red rectangle in the right panel. Lower-left panel: P–V diagram along the eastern arm in the H42α recombination line. The sampling area is shown by the lower red rectangle in the right panel. (Color online)

Open in new tabDownload slide

Because a compact component is seen in the SiO emission line around the boundary between AB and EA (see figure 4), the molecular gas may have been ionized by strong UV emission from the central cluster and compressed by the expanding ionized gas. The area is also identified by SOFIA as a compact component in extremely high-J CO emission lines (up to J = 16–15) (Requena-Torres et al. 2012). On the other hand, the other area of AB is not clear in the SiO emission line. As mentioned in sub-subsection 3.2.2, AB has filamentary substructures. The interactions between the filaments may produce mildly shocked molecular gas (Δ ∼ 10 km s⁻¹). Therefore, almost all parts of AB are not abundant in strongly shocked molecular gas except for the boundary area. Meanwhile, the northern half of AB can also be identified in the channel maps with V_{c, LSR} = 77.5 to 97.5 km s⁻¹ in the CH₃OH emission line. The northern half of AB is abundant in mildly shocked molecular gas. Therefore, the shock velocity in the structure is probably in the range of 10 to 30 km s⁻¹. The southern half of AB cannot be identified in the channel maps of V_{c, LSR} = 107.5 and 117.5 km s⁻¹ except for the small patch located around |$l \sim {359{^{\circ}_{.}}960}$|⁠, |$b \sim {-0{^{\circ}_{.}}034}$| (see figure 6). Consequently, we argue that AB is falling from the outer region to the vicinity of Sgr A*, being disrupted by the tidal shear of Sgr A* and affected by cosmic rays. When AB further approaches Sgr A*, it would be immediately ionized by extremely strong Lyman continuum emission (∼4 × 10⁵⁰ s⁻¹) from the central cluster (e.g., Scoville et al. 2003). The final destination of the falling clouds depends on the initial conditions including the impact parameter, as mentioned above. Because AB is ionized and connects with EA, the impact parameter of AB is considered to be smaller than that of AA.

Figure 12 shows the positional relation between the molecular gas of AC in the CS emission line and the ionized gas of the horizontal arm (HA) in the H42α recombination line. HA is identified as a nearly vertical substructure of the ionized gas in the Galactic coordinates. AC and HA are apparently identified as a continuous and nearly vertical filament in the channel map. However, AC is not continuous with HA in the position–velocity diagram. This would also be falling molecular gas from the outer region to the vicinity of Sgr A*, but is not ionized. The impact parameter of AC seems larger than that of AB.

Fig. 12.

Left panel: Integrated intensity map of Anomaly C at V_LSR = 77.5 km s⁻¹ in the CS J = 2–1 emission line (upper part) and the horizontal arm at V_LSR = 67.5 km s⁻¹ in the H42α recombination line (lower part). The central velocities are shown at the upper-left and lower-right corners of the upper and lower parts, respectively. The velocity widths of the two parts are 10 km s⁻¹. The red dashed ellipse shows the elliptical outline of the CND. The white contours show the continuum emission at 100 GHz of the GCMS at 100 GHz for comparison (Tsuboi et al. 2016). Upper-right panel: P–V diagram along Anomaly C in the CS J = 2–1 emission line. The sampling area is shown by the upper half of the red rectangle in the left panel. Lower-right panel: P–V diagram along the horizontal arm in the H42α recombination line. The sampling area is shown by the lower half of the red rectangle in the left panel. (Color online)

Open in new tabDownload slide

5 Conclusions

We have presented high angular resolution and high sensitivity images of the CND and the surrounding region in the CS J = 2–1, SiO v = 0 J = 2–1, H¹³CO⁺J = 1–0, C³⁴S J = 2–1, and CH₃OH |$J_{K_a, K_c}=2_{1,1}$|–1_{1, 0}A₋₋ emission lines using ALMA.

• The CND is depicted as a torus-like collection of molecular gas with an inner radius of R_in ∼ 1.5 pc and an outer radius of R_out ∼ 2 pc in these emission lines, except for the CH₃OH emission line.

• The kinematics of the CND roughly obey rotation with some radial motion. The rotation velocity and radial motion of the CND are estimated to be V_rot ∼ 115 km s⁻¹ and V_rad ∼ 23 km s⁻¹, respectively.

• The LTE molecular gas mass of the CND is estimated to be M_LTE ∼ 3.1 × 10⁴ M_⊙ from the observations of the CS J = 2–1 and C³⁴S J = 2–1 emission lines.

Our derived size, kinematics, and mass for the CND are consistent with those in previous observations.

We also found some anomalous molecular clouds in the surrounding region, which have filamentary appearances prominent in the CS J = 2–1 emission line.

• AA is positionally connected with the northwestern sector of the CND, which is adjacent to WA. This is likely being disrupted by the tidal shear of Sgr A*. The cloud seems to rotate around Sgr A* in the opposite direction to the CND. The molecular cloud would be recently falling from the outer region to the CND because relaxation has not yet occurred.

• AB is continuously connected with the eastern tip of the “eastern arm” of the GCMS. The velocity of the cloud is consistent with that of the ionized gas in the EA. These facts suggest that the molecular cloud is falling from the outer region to the vicinity of Sgr A*, being disrupted by the tidal shear, and ionized by strong UV emission from the central cluster because the impact parameter of the cloud is smaller than that of AA.

These clouds would play an important role in transferring material from the outer region of the Sgr A complex to the CND and/or the vicinity of Sgr A*.

Acknowledgments

This work is supported in part by Grant-in-Aid from the Ministry of Education, Sports, Science and Technology (MEXT) of Japan No. 16K05308. This work makes use of the following ALMA data: ADS/JAO.ALMA#2012.1.00080.S and ALMA#2012.1.00543.S. The later data are retrieved from the JVO portal 〈http://jvo.nao.ac.jp/portal〉 operated by the National Astronomical Observatory of Japan (NAOJ). The National Radio Astronomy Observatory (NRAO) is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc. ALMA is a partnership of ESO (representing its member states), NSF (USA), and NINS (Japan), together with NRC (Canada), NSC and ASIAA (Taiwan), and KASI (Republic of Korea), in cooperation with the Republic of Chile. The Joint ALMA Observatory is operated by ESO, AUI/NRAO, and NAOJ. This work has made use of NASA’s Astrophysics Data System. This work also has made use of the JCMT Science Archive (http://www.cadc-ccda.hia-iha.nrc-cnrc.gc.ca/en/jcmt/). The JCMT has historically been operated by the Joint Astronomy Centre on behalf of the Science and Technology Facilities Council of the United Kingdom, the NRC, and the Netherlands Organisation for Scientific Research.

Footnotes

References