The Milky Way Imaging Scroll Painting (MWISP): Project Details and Initial Results from the Galactic Longitudes of 258–497

The DLH telescope is situated at approximately (37°224N, 97°336E), close to Delingha, the third largest city of Qinghai Province in western China. At an altitude of about 3200 m, dry and stable atmospheric conditions make Delingha an excellent site for millimeter astronomy.

The surface accuracy of the main reflector of the telescope is routinely measured and has a typical value of ≈73 μm. The tracking accuracy of the telescope is approximately 1''–3'' in both azimuth and elevation. The pointing accuracy for the whole sky is better than 5'', about one-tenth of the telescope half-power beam width (HPBW) at 115 GHz (∼50'').

A 3 × 3 multibeam sideband-separating Superconducting SpectroScopic Array Receiver (SSAR) system (Shan et al. 2012) was installed on the DLH telescope in the 2010 winter. The centroids of the 3 × 3 beam patterns have a regular spacing of 172'' (see Figure 16 in Shan et al. 2012), which are also measured before every observational season. The SSAR system employs two-sideband-separating superconductor–insulator–superconductor mixers with a typical single sideband noise temperature of 60 K and image rejection ratio above 10 dB over the frequency range of 85–115 GHz. The instrument includes Fast Fourier Transform Spectrometers (FFTSs), digital Local Oscillator (LO) sources, digital bias power supplies, and an independent intermediate frequency (IF) module.

The IF band is 2.64 ± 0.5 GHz. The ${}^{12}\mathrm{CO}$, ${}^{13}\mathrm{CO}$, and C¹⁸O (J = 1–0) lines at 115.271 GHz, 110.201 GHz, and 109.782 GHz, respectively, are all within the 1 GHz band when the LO is set at 112.6 GHz. Eighteen high-resolution FFTS digital spectrum analyzers work at 1 GHz or 200 MHz bandwidth with 16,384 channels. The 1 GHz bandwidth covers a wide velocity range for the CO line survey, while the 200 MHz bandwidth provides higher spectral resolution observations for further study (e.g., see Table 1 in Gong et al. 2018).

Typical system temperatures, which include noises from the receiver, the antenna (the optical system, the dome, and the membrane), and the atmosphere, are ∼250 K for ${}^{12}\mathrm{CO}$ at the upper sideband and ∼140 K for ${}^{13}\mathrm{CO}$ and C¹⁸O at the lower sideband, respectively. Observations were calibrated using the standard chopping wheel method that allows switching between the sky and an ambient temperature load. Details on the telescope can be found at http://www.radioast.nsdc.cn/mwisp.php.

The sky coverage of the MWISP project is divided into 10,941 cells for the region of l = −10° to +250° and $| b| \,\lesssim$ 52. Each cell with 30' × 30' is scanned along the Galactic longitude (l) and the Galactic latitude (b) at least twice to reduce the fluctuation of noise. We chose 30' × 30' as the cell's size based on three factors. One is that the cell can be completed in time with the suitable scanning parameters. This consideration is important to ensure the atmospheric stability during the 1–1.5 hr observation for a full mapping (see below). The other factor is to limit the final data size of the completed cell for one line, e.g., ∼0.5 GB (91 pixels×91 pixels×16,384 channels×4 Byte). Finally, such a size is good to divide the whole mapping sky with a regular form, leading to easy labeling for a cell in Galactic coordinates.

All observations are taken in position-switch On-The-Fly (OTF; see Sun et al. 2018) mode. The scan rate is 50'' (or 75'') per second, with a dump time of 0.3 s (or 0.2 s), depending on a cell's elevations, off-positions, and weather conditions. The sampling interval is 15'' (=scan rate×dump time) and the spacing between scan rows is 10'', fulfilling the oversampling of the 50'' beam of the 13.7 m telescope. A cell thus can be fully mapped in half an hour (or an hour) along l or b. Generally, we used the (scan rate = 50'' s⁻¹, dump time = 0.3 s) mode in observations. Mapping twice (or three hours) can satisfy the expected rms noise level. The (scan rate = 75'' s⁻¹, dump time = 0.2 s) mode is often used for special regions such as, for example, the low elevation cells toward l < 10°−15° (e.g., near the direction of the Galactic center), or cells with far off positions (e.g., the Aquila Rift region and the Cygnus region). Sometimes, scanning twice (e.g., three hours) cannot reach the expected rms noise level, so further mapping with the (scan rate = 75'' s⁻¹, dump time = 0.2 s) mode is also used to reduce the rms noise for the observed cells.

The most important consideration is to maintain uniform sensitivity, which essentially depends on the weather or the total system temperature. During observations, standard sources are thus observed about every two hours in the position-switch mode. The spectral profiles and intensities of standard sources are monitored daily to check the stability of the observations.

Before observations, the off position with an area of $\sim 8^{\prime} \times 8^{\prime}$ (i.e., the sky coverage of the 3 × 3 beams) near the target cell (e.g., ≤2°) is carefully checked to ensure free emission for the ${}^{12}\mathrm{CO}$ line. However, the ${}^{12}\mathrm{CO}$ emission in the range of 0–40 km $\,{{\rm{s}}}^{-1}$ is widely distributed toward the inner Galaxy, especially for the central molecular zone, the Aquila Rift, and the Cygnus region. When we observe these regions, some off-positions with a little ${}^{12}\mathrm{CO}$ emission (T_MB ≲ 0.3 K) are also used as the reference backgrounds. For these regions, ${}^{12}\mathrm{CO}$ lines may show some weak absorption features at certain local standard of rest (LSR) velocities. The relevant information was recorded for all observed cells.

The total bandwidth of 1 GHz, with 16,384 channels, provides a channel frequency interval of 61 kHz, resulting in a velocity separation of about 0.16 km s⁻¹ for ${}^{12}\mathrm{CO}$ and 0.17 km s⁻¹ for ${}^{13}\mathrm{CO}$ and C¹⁸O. A first-order (or linear) baseline was fitted for the CO spectra. Most of the bad channels in the spectra were removed. The antenna temperature, ${T}_{{\rm{A}}}^{\star }$, was converted to the main beam brightness temperature, T_MB, with the formula ${T}_{\mathrm{MB}}={T}_{{\rm{A}}}^{\star }/({f}_{{\rm{b}}}\times {\eta }_{\mathrm{MB}})$. The beam-filling factor of f_b is assumed to be 1 for the extended CO emission and the main beam efficiency η_MB varied between 40% and 50% in the past seven seasons. All intensities shown in this paper are on the T_MB scale.

Due to the field rotation, a larger cell size of 455 × 455 was preserved in the data processing, resulting in some overlap between neighboring cells. For each cell named as GLLL.L±BB.B, all CO spectra were summed together with the rms as the weight. Finally, the three-dimensional (3D) FITS data cubes of each cell were made with a grid spacing of 30'' for ${}^{12}\mathrm{CO}$, ${}^{13}\mathrm{CO}$, and C¹⁸O (J = 1–0) lines. All data were reduced using the GILDAS software⁴ (Pety 2005).

The MWISP project is a large-scale, unbiased, and high-sensitivity triple CO isotope line survey for the northern Galactic plane (e.g., l = −10° to +250° and $| b| \,\lesssim$ 52) that uses the 13.7 m single-dish telescope. The new nine-beam array with the OTF observation mode increases the mapping speed by roughly an order of magnitude, compared with the old single-beam system.

The DLH telescope can observe the northern Galactic plane for about 15–16 hr per day from September to April. Generally, each cell with the required rms noise (see Table 1) consumes approximately 3–5 hr (or 2–4 scans along l and b). That is, the MWISP project can map about 240–250 square degrees of the Galactic plane per year. The MWISP survey started in 2011 November and is expected to be completed in 2022.

Table 1.  Parameters of the MWISP Survey Data

Molecular lines	${}^{12}\mathrm{CO}$ (J = 1–0)	${}^{13}\mathrm{CO}$ and C18O (J = 1–0)
HPBW	∼49''	∼52''
Grid spacing	30''	30''
Velocity separation	0.16 km s−1	0.17 km s−1
System temperature	∼250 K	∼140 K
rms (TMB)	∼0.5 K	∼0.3 K
Mapping regions	258 ≲ l ≲ 497, $\left|b\right|\,\lesssim$ 52	258 ≲ l ≲ 497, $\left|b\right|\,\lesssim$ 52

Download table as:  ASCIITypeset image

The characteristics and advantages of the MWISP CO survey are summarized as follows:

• 1.  
  Large-scale CO mapping with a high spatial dynamic range. The MWISP project will provide us with large-scale CO maps of ∼2600 deg², with a moderate angular resolution of ∼50''. The final full-sampling 3D CO data sets have a grid spacing of 30''. Researchers thus can investigate the detailed structures of local MCs (e.g., 1 pixel ≲ 0.15 pc at a distance of ≲1 kpc), the properties of MCs with moderate distances (e.g., ∼3–6 kpc), and the distribution of distant molecular gas (e.g., ≳10 kpc).
• 2.  
  Unbiased CO survey with high sensitivity. The unbiased survey with high sensitivity, wide velocity coverage, and high velocity resolution has a uniform sensitivity, providing a clearer picture of the molecular gas distribution and properties of the Milky Way. Through systematic studies of the CO emission of the molecular gas, many new MCs will be revealed due to the high-quality data of the new survey, which have an rms sensitivity of ∼0.5 (0.3) K, a velocity coverage of ∼2600 km $\,{{\rm{s}}}^{-1}$, and a velocity resolution of ∼0.16 (0.17) km $\,{{\rm{s}}}^{-1}$ for ${}^{12}\mathrm{CO}$ (${}^{13}\mathrm{CO}$ and C¹⁸O) lines.
• 3.  
  Simultaneous observations of ${}^{12}\mathrm{CO}$, ${}^{13}\mathrm{CO}$, and C¹⁸O (J = 1–0) transitions. The ${}^{12}\mathrm{CO}$ emission tracing the total molecular gas can reveal structures and distributions of the diffuse gas with a typical density of 10² cm⁻³. The optically thinner ${}^{13}\mathrm{CO}$ and C¹⁸O lines trace denser molecular gas with a typical density of 10³–10⁴ cm⁻³ because of less abundance with respect to H₂ and therefore less optical depth effects. Therefore, the MWISP survey is adequate for revealing the diffuse gas of MC envelopes, denser molecular gas of giant molecular clouds (GMCs), and changes in abundance ratio between ${}^{13}\mathrm{CO}$ and C¹⁸O due to isotope-selective photodestruction of the rarer CO species.

With a spatial resolution of ∼50'' and a grid spacing of 30'', the full-sampling MWISP survey provides a rich CO data set for b > 1° regions, which are less covered by other CO surveys, excluding the CfA 1.2 m complete ${}^{12}\mathrm{CO}$ (J = 1–0) survey with a beam size of ∼8' (Dame et al. 2001). Figure 1 gives a comparison between the CfA 1.2 m CO map and our MWISP map for the Serpens/Aquila Rift MC complex. Note that the MWISP integrated map reveals more detailed structures than that of the 1.2 m CO data. Furthermore, the MWISP survey has a wider velocity coverage than that of the previous CO survey (e.g., −5 to +135 km $\,{{\rm{s}}}^{-1}$ for the GRS ${}^{13}\mathrm{CO}$ survey, Jackson et al. 2006), leading to more completed velocity coverage for MCs in the first quadrant of the Milky Way (e.g., the most distant MCs beyond the solar circle).

Zoom In Zoom Out Reset image size

The ongoing FUGIN survey plans to investigate the distribution and properties of molecular gas in the Galaxy with the ${}^{12}\mathrm{CO}$, ${}^{13}\mathrm{CO}$, and C¹⁸O (J = 1–0) lines (Umemoto et al. 2017). The region of l = +10° to +50° and b = +1° to +1° is fully covered using the multibeam (2 × 2), dual-polarization, two-sideband receiver installed in the Nobeyama 45 m telescope. Compared with the FUGIN project, the MWISP CO survey has the larger coverage (2600 deg² versus 160 deg²) and the higher velocity resolution (0.16 km $\,{{\rm{s}}}^{-1}$ versus 0.65 km $\,{{\rm{s}}}^{-1}$). Meanwhile, the sensitivity of the MWISP survey is better than that of the FUGIN project (e.g., T_rms(T_MB) ∼ 1.5 K for ${}^{12}\mathrm{CO}$ and ∼ 0.7 K for ${}^{13}\mathrm{CO}$/C¹⁸O at 85 × 85 × 1.3 km $\,{{\rm{s}}}^{-1}$ in the firt quadrant regions, Umemoto et al. 2017). On the other hand, the spatial resolution of the Nobeyama 45 m telescope is about 2.5 times higher than that of the DLH 13.7 m telescope. Therefore, the FUGIN data with the final grid spacing of 85 can resolve more detailed MC structures in the Galactic plane of $| b| \lt 1^\circ$.

In summary, the MWISP CO survey gives us a good opportunity to study the Galactic structures, the MC properties and the star formation, and the associations between the molecular gas and the extended radio sources such as, for example, H ii regions and supernova remnants (SNRs). Much work has been done based on the MWISP data, such as Galactic arms traced by the CO emission (Sun et al. 2015, 2017; Du et al. 2016, 2017; Su et al. 2016), MCs and star formation (Li et al. 2013, 2018a, 2018b; Zhang et al. 2014b; Su et al. 2015; Chen et al. 2016, 2017b; Gong et al. 2016, 2017; Zhan et al. 2016; Wang et al. 2017; Xiong et al. 2017; Sun et al. 2018), and interactions between SNRs and MCs (Su et al. 2014b, 2014c, 2017a, 2017b, 2018; Zhou et al. 2014, 2016; Chen et al. 2017a). The MWISP project also provides an invaluable data set for studying the distribution and kinematics of the MCs in the Milky Way (Section 3). The large-scale CO data are important in delineating the spiral structure of the Milky Way (e.g., see the recent review in Xu et al. 2018). More studies will be presented in the immediate future.

The MWISP survey is still ongoing. At the time of this writing the survey has completed about 60% of the planned 2600 square degree coverage (see the details at http://www.radioast.nsdc.cn/viewallobsed.php). The complete set of the survey data generated by the MWISP project will be made publicly available via an online archive. Early access to the 3D data set for specific regions of the sky is also possible through mutual collaboration. People who are interested in the MWISP data can request the datacube by contacting the email address dlhproposal@pmo.ac.cn. Many mosaic data cubes have already been provided to researchers through the FTP server.

Very recently, Su et al. (2016) and Sun et al. (2017) presented the results of distant MCs traced by CO emission between l = 3475 and l = 4525. These MCs with negative velocities are divided into the distant Outer Arm and the Extreme Outer Galaxy (EOG), respectively. These two parts of the molecular gas are very likely the extension of the Norma–Cygnus Arm and the Scutum–Centaurus arm in the first quadrant (i.e., molecular gas structures from the fourth quadrant of the inner Galaxy to the first quadrant of the outer Galaxy). Actually, the more negative velocity feature at V_LSR < 0 km s⁻¹ was proposed to be from the Outer Scutum–Centaurus arm (Dame & Thaddeus 2011).

With the progress of the MWISP project, larger mapping was completed for the range of +258 ≲ l ≲ +497 and $| b| \,\lesssim$ 52. Figure 4 shows the spatial distribution of the CO gas with V_LSR < 0 km s⁻¹, together with the corresponding LSR velocity information. A large amount of new ${}^{12}\mathrm{CO}$ emission is revealed because of the high sensitivity and the large coverage of the unbiased survey. Some of the MCs with relatively strong ${}^{12}\mathrm{CO}$ emission have corresponding ${}^{13}\mathrm{CO}$ emission, but none of them show significant C¹⁸O emission under the improved rms level of ∼ 0.16 K at the smoothed velocity resolution of 0.5 km s⁻¹. Higher resolution and sensitivity observations are expected to detect the weak emission of the dense gas far away from us.

Zoom In Zoom Out Reset image size

Some MCs with V_LSR ≲ 0 km s⁻¹ are likely the local molecular gas due to their ∼0 km s⁻¹ LSR velocities and relatively high $| b|$ values (e.g., $| b| \,\gtrsim$ 25; see the Aquila Rift region near l ∼ 26°–32° in the lower panel of Figure 4). However, most V_LSR < 0 km s⁻¹ MCs in Figure 4 are believed to lie beyond the solar circle. Obviously, CO emission is mainly concentrated near the Galactic plane, excluding the possible local gas in the Aquila Rift region at high $| b|$. In spite of this, the distant CO gas seems to also be slightly displaced from b = 0°, which was explained by the warped plane at larger Galactocentric distances and the apparent tilted structure caused by the Sun's z-height above the physical midplane of the Galactic disk (see Su et al. 2016; Sun et al. 2017).

Due to the large distances, many MCs have small angular sizes and weak CO emission. These MCs are relatively isolated in l–b–v space. Despite these properties, some interesting distant MCs with brighter CO emission also display extended concentrations, which usually have special morphologies such as filaments, arcs/shells, and irregular cavity-like structures. These extended concentrations could be related to the massive star-forming activities in the surroundings. Very recently, Wenger et al. (2018) investigated the CO properties toward star-forming regions in the the Outer Scutum–Centaurus arm using a single pointing observation of the 12 m telescope located at Arizona Radio Observatory. Benefiting from the full CO map and the high sensitivity of the MWISP data, we will systematically study the relationship between the distant MCs and the surrounding star formation activities. Detailed identification and analyses of these MCs will be presented in future papers. Section 3.3.1 shows some initial results for the CO gas at the edge of the Galaxy.

Figure 5 shows the ${}^{13}\mathrm{CO}$ (J = 1–0) intensity map in the 0–130 km s⁻¹ interval, overlaid with radio continuum contours from the Effelsberg 11 cm survey (Reich et al. 1990). Some bright and/or extended radio sources discussed below are labeled on the guide map.

Zoom In Zoom Out Reset image size

Figures 6 and 7 show ${}^{12}\mathrm{CO}$ and ${}^{13}\mathrm{CO}$ channel maps, respectively. The channel maps are made by integrating emission over a 10 km s⁻¹ velocity bin from 0 km s⁻¹ to 120 km s⁻¹. The 120–130 km s⁻¹ ${}^{12}\mathrm{CO}$ and ${}^{13}\mathrm{CO}$ gas, which displays emission only near l ∼ 26°–31°, is presented in Figure 8. A large number of MC structures and features are seen in both of the ${}^{12}\mathrm{CO}$ and ${}^{13}\mathrm{CO}$ channel maps.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

For the 0–20 km s⁻¹ maps, the most prominent features are the Aquila Rift (e.g., see Figure 3 in Dame & Thaddeus 1985), which displays large-scale, diffuse, and enhanced CO emission in the field of view (FOV). The MWISP survey, with wide spatial dynamic range, reveals lots of detailed structures for the large-scale molecular gas not far from us (i.e., d ≲ 450 pc; Ortiz-León et al. 2017). Large-scale extended structures are also discernible from the PV diagrams of ${}^{12}\mathrm{CO}$and ${}^{13}\mathrm{CO}$ emission (see the region of V_LSR ∼ 0–20 km s⁻¹ and l ∼ 26°–40° in Figure 9). The ${}^{12}\mathrm{CO}$ J = 2–1 and ${}^{13}\mathrm{CO}$ J = 2–1 emission from the Aquila Rift region was investigated by Nakamura et al. (2017) using the 1.85 m telescope. Combinations of our CO J = 1–0 data and the 1.85 m CO J = 2–1 data will help us understand the detailed molecular gas properties of such local regions.

Zoom In Zoom Out Reset image size

Many of bright CO concentrations in the velocity interval of ∼0–30 km s⁻¹ are found to be near the Galactic plane of b ∼ − 04 to +06, excluding the diffuse and extended emission from the local MCs (e.g., the Aquila Rift region discussed above). These concentrations, which have typical angular sizes of several arcminutes, are likely related to the distant Perseus Arm in the first quadrant of the Galaxy.

For 20–40 km s⁻¹ gas, the ${}^{12}\mathrm{CO}$ and ${}^{13}\mathrm{CO}$ emission is found to be extended over a large region from l ∼ 34°–44° and $| b| \lesssim 5^\circ$ in the channel maps. This GMC complex is less studied in the literature. A 170 pc long giant molecular filament of G40.82−1.41, which is at ∼2 kpc (e.g., the kinematic distance and the extinction distance, Su et al. 2018), is probably a part of the GMC complex. The extended molecular gas at b ≳ 0° of the GMC is related to the H ii regions Sh 2–75 (at l = 4012, b = 150) and Sh 2–76 (at l = 4044, b = 245). One of the H ii regions, Sh 2–76, is exactly located at a distance of 1.92${}_{-0.08}^{+0.09}$ kpc from the parallax measurements of 0.521 ± 0.024 mas (Chibueze et al. 2017).

At 30–60 km s⁻¹, significant CO emission appears at l ∼ 34°–36° and b ∼ −3° to +1°, which is roughly perpendicular to the Galactic plane. Enhanced CO emission in the GMC complex displays complicated morphologies, such as multi-shells, bubbles, and cometary bright-rimmed structures. The GMC complex is associated with the high-mass star-forming region W48 (G35.2−1.8) at a parallax distance of 3.27${}_{-0.42}^{+0.56}$ kpc (Zhang et al. 2009). The very bright SNR G34.7−0.4, also known as W44 at a near kinematic distance of ∼3.3 kpc (e.g., Su et al. 2014a), may be associated with the GMC complex.

For the V_LSR ≳ 50 km s⁻¹ molecular gas, CO emission is mainly confined within $| b| \lesssim 1^\circ$. In addition to the CO emission close to the Galactic plane, considerable MCs are detected in $1^\circ \lesssim | b| \lesssim 2^\circ$, even for velocities near the tangent point (see 80–120 km s⁻¹ maps in Figures 6 and 7). These features have not been revealed by previous CO surveys (e.g., FCRAO GRS project, Jackson et al. 2006) due to the limited Galactic latitude coverage of $| b| \lesssim 1^\circ$ toward the first quadrant of the Milky Way. From Figure 9, we find that the PV diagram of the MWISP ${}^{13}\mathrm{CO}$ data is consistent with the previous PV map of the GRS ${}^{13}\mathrm{CO}$ survey (see Figure 3 in Jackson et al. 2006). Furthermore, some new features in the ${}^{13}\mathrm{CO}$ PV diagram are unveiled for the V_LSR ≲ 50 km s⁻¹ range (e.g., see structures at l ∼ 38°–42° and V_LSR ∼ 25–45 km s⁻¹, the rectangle in the diagram) because of the larger latitude coverage of the MWISP survey. The CO emission of the structure is from the molecular gas associated with H ii regions Sh 2–75 and Sh 2–76.

Figure 10 displays the spatial and velocity distribution of the C¹⁸O (J = 1–0) emission in the velocity range of 0–40 km s⁻¹. Enhanced C¹⁸O emission is clearly seen near l ∼ 287 and b ∼ 37, which is probably associated with the feedback of massive stars in the Serpens south and the W40 complex (see Section 3.3.2). Excluding C¹⁸O emission from the local MCs such as the Aquila Rift, the rest of C¹⁸O gas is mainly confined in b ∼ −2° to b ∼ 3°. In the figure, two other prominent C¹⁸O features, which are located at l ∼ 35° and l ∼ 40°, are found to be away from the b = 0° plane. The two C¹⁸O features are related to the massive star-forming region W48 and H ii regions Sh 2–75 and Sh 2–76, respectively.

Zoom In Zoom Out Reset image size

Figure 11 shows the C¹⁸O emission in the velocity interval of 40–80 km s⁻¹ and 80–120 km s⁻¹, which is mainly within $| b| \lesssim 2^\circ$ and $| b| \lesssim 1^\circ$, respectively. For the 40–80 km s⁻¹ maps, the W48 GMC complex at l ∼ 35°, which is roughly perpendicular to the b = 0° plane, is clearly seen in Figures 11(a) and (b). On the other hand, the W51 GMC complex at a distance of 5.41${}_{-0.28}^{+0.31}$ kpc (e.g., the trigonometric parallax from Sato et al. 2010) also can be seen at l ∼ 49°. The massive star-forming region W51 was recently reviewed by Ginsburg (2017, and references therein).

Zoom In Zoom Out Reset image size

For the V_LSR = 80–120 km s⁻¹ maps (see Figures 11(c), (d)), the most enhanced C¹⁸O emission with ∼10 K km s⁻¹ is associated with the mini-starburst region W43 at (l ∼ 307, b ∼ 00). The massive star-forming region, which is at a parallax distance of 5.49${}_{-0.34}^{+0.39}$ kpc (Zhang et al. 2014a), is probably at or close to the near end of the Galactic long bar. The ongoing star-forming activity of the global mini-starburst region is likely the result of the massive gas clouds accumulated by the bar potentials and kinematics. The gas property of the interesting GMC complex W43 was studied by many groups (e.g., Nguyen Luong et al. 2011; Carlhoff et al. 2013; Motte et al. 2014; Sofue et al. 2018).

For the survey data presented here, the C¹⁸O emission is relatively discrete in comparison with the extended ${}^{12}\mathrm{CO}$ and ${}^{13}\mathrm{CO}$ emission. Researchers thus can identify the main structures of MCs using C¹⁸O emission due to the less velocity crowding and line blending (e.g., see the W40 region in Section 3.3.2).

As shown in Figure 11, the distribution of the C¹⁸O emission is below the plane of b = 0°, especially for the 40–80 km s⁻¹ MCs. Near the tangent point, where the dense gas is more concentrated in the plane, the dominant part is also below the plane (e.g., the 80–120 km s⁻¹ maps). This interesting feature is explained by the Sun's offset above the physical midplane of the Milky Way (e.g., ${z}_{\mathrm{Sun}}\,\sim$ 17.1 pc; see discussions and Figure 7 in Su et al. 2016). We further investigate the phenomena using MC samples close to the tangent point, e.g., the CO gas with V_LSR > 60 km s⁻¹ (Section 3.3.3).

3.3.1. CO Gas at the Edge of the Milky Way

The CO emission of the distant molecular gas is very weak. Due to the warping and flaring of the outer gas disk, the distant MCs with negative velocities are concentrated at somewhat higher latitudes, which is different from the molecular gas in the first quadrant of the inner Galaxy. The unbiased MWISP CO survey has a wide velocity and areal coverage, meanwhile there is also a high sensitivity, allowing us to systematically investigate weak CO emission far away from the Sun (see Figure 4). Compared with the CfA 1.2 m CO survey (Dame et al. 2001), the MWISP CO data from the 13.7 m telescope reveal detailed distributions and structures of the distant MCs.

Figure 12 shows the longitude–velocity (l–v) diagram of the ${}^{12}\mathrm{CO}$ emission for the V_LSR ≲ 0 km s⁻¹ molecular gas. Note that the CO intensity is multiplied by a factor of 100 for the corresponding signal velocity range but not the whole velocity range. The noise is therefore suppressed and the tiny features of weak CO emission are enhanced in the l–v diagram. Based on the diagram, we find that the bright emission (thick blue parts in the figure) exhibits a large-scale molecular gas structure outside the solar circle. On a large scale, the structure shows a velocity gradient of ∼ 2.5–3.0 km $\,{{\rm{s}}}^{-1}$ degree⁻¹ and has relatively enhanced ${}^{12}\mathrm{CO}$ emission, which is likely from MCs within the Outer arm (or the Norma–Cygnus arm) in the first quadrant. Moreover, the large-scale structure is comprised of several interesting substructures in the l–v space, which themselves are worthy of further investigation.

Zoom In Zoom Out Reset image size

In addition to the enhanced ${}^{12}\mathrm{CO}$ emission, considerable molecular gas with weaker CO emission at more negative LSR velocities is also discerned from Figure 12 (see the regions between the red lines, i.e., ${V}_{\mathrm{LSR}}=-1.57\times l+4.34\,\pm 12$ km s⁻¹). The most negative velocity is at ∼ −75 km s⁻¹, with low surface brightness. These MCs are probably in a distant section of the Scutum–Centaurus arm (designated as the OSC by Dame & Thaddeus 2011), which is referred to as the EOG region by Sun et al. (2017). The trend of the molecular gas in the l–v space is approximately described by a linear fit of V_LSR = −1.57 × l + 4.34 km s⁻¹, which is in good agreement with previous studies (see Dame & Thaddeus 2011; Sun et al. 2017). The velocity gradient of ∼1.6 km $\,{{\rm{s}}}^{-1}$ degree⁻¹ for the EOG gas is smaller than that of the molecular gas of the Outer arm in the l–v space (e.g., ∼2.8 km $\,{{\rm{s}}}^{-1}$ degree⁻¹ in Su et al. 2016). The uncertainty of the tentative fit is largely due to the limited longitude coverage. On the other hand, the OSC arm may consist of several broken lines with somewhat different slopes in the l–v space. Further MWISP data from l = 0° to 26° and l = 50° to 100° can give us a firmer conclusion (e.g., see Figure 3 in Sun et al. 2015).

Figure 13 displays the distribution of the EOG ${}^{12}\mathrm{CO}$ emission. Samples in the figure are selected from the MCs with V_LSR ≲ −1.57 × l + 4.34 km s⁻¹, which reduces possible contamination from the Outer arm gas. We find that these most distant MCs are quite sparse in the l–b space and indeed located above the b = 0° plane. For regions of l ∼ 40°–50°, the lack of EOG CO emission in the map indicates that the slope of −1.57 in our fitting is probably too steep for the distant MCs at larger longitudes (e.g., a slightly flat slope of ≳−1.4 for l ≳ 40° CO gas in the l–v space). Two intensity peaks are located at b ∼ 12 and b ∼ 28, or ∼370 pc and ∼870 pc above the b = 0° plane, assuming a median heliocentric distance of 17.8 kpc (Sun et al. 2017). These values are several times larger than that of the Outer arm gas (e.g., 042 or ∼110 pc at a heliocentric distance of 15 kpc; Su et al. 2016), indicating the different distributions of the two MC groups for the distant V_LSR < 0 km s⁻¹ gas. Indeed, the more negative the value of the LSR velocity is, the higher the Galactic latitude of the distant molecular gas.

Zoom In Zoom Out Reset image size

3.3.2. Bubble-like ${{\rm{C}}}^{18}{\rm{O}}$ Structure near the H ii Region W40

At similar distances, there may be multiple MCs with different velocity fields in the ISM, leading to complicated molecular gas structures. Simultaneous ${}^{12}\mathrm{CO}$, ${}^{13}\mathrm{CO}$, and C¹⁸O (J = 1–0) line observations can provide a more complete picture of the molecular gas properties, such as structures, kinematics, and dynamics. Star formation is associated with molecular gas. The MWISP survey is helpful for studying the relationship between MCs and star-forming activity in the Milky Way.

An intriguing case is the W40 region, which is part of the Serpens/Aquila Rift MC complex. The region is located at a distance of 436.0 ± 9.2 pc according to the measurements of proper motions of seven stars across the MC complex (Ortiz-León et al. 2017). Nakamura et al. (2017) covered the MC complex using ${}^{12}\mathrm{CO}$, ${}^{13}\mathrm{CO}$, and C¹⁸O (J = 2–1) lines with the 1.85 m telescope. The gas toward W40 within an area of ∼ 1 square degree was further investigated by the Nobeyama 45 m telescope (Shimoikura et al. 2018). Based on the MWISP survey on the 13.7 m telescope, a prominent bubble-like structure is unveiled in the C¹⁸O intensity map (e.g., see the upper right corner of Figure 10).

Figure 14 displays the close-up view toward the W40 region, where the blue, green, and red colors represent the ${}^{12}\mathrm{CO}$, ${}^{13}\mathrm{CO}$, and C¹⁸O emission, respectively. The circular morphology of the dense gas in the three-color image is centered at (l = 2907, b = 382), with a radius of 045. Obviously, the emission of the dense gas traced by ${}^{13}\mathrm{CO}$ and C¹⁸O lines is mainly concentrated in the western half of the bubble-like structure, while the C¹⁸O structure in the eastern half is broken. The LSR velocity of C¹⁸O emission peaks at V_LSR ∼ 7.5 km s⁻¹ on the bubble-like structure, where both the ${}^{12}\mathrm{CO}$ and ${}^{13}\mathrm{CO}$ lines show striking self-absorption features (see the typical spectra to the right of the figure).

Zoom In Zoom Out Reset image size

The strongest ${}^{12}\mathrm{CO}$ emission of T_MB ∼ 37 K at V_LSR ∼ 4.5 km s⁻¹ is located at (l = 2875, b = 351), which is close to the center of the H ii region W40 (see the purple circle in Figure 14). The typical peak temperature of the optically thick ${}^{12}\mathrm{CO}$ line is nearly 10 K in some bright regions, while considerable ${}^{12}\mathrm{CO}$ emission has a lower peak temperature of ∼4–6 K in other faint regions. Huge regions of molecular gas with relatively weak CO emission extend several tens of square degrees over the FOV (e.g., see 0–10 km s⁻¹ and 10–20 km s⁻¹ maps in Figure 6).

Figure 15 displays the gas distribution in velocity intervals of 3–5, 5–7, 7–9, and 9–11 km s⁻¹ toward the complex. For the 3–5 km s⁻¹ gas, the very strong CO emission is found to surround the W40 H ii region (the 9' purple circle in the map), while CO emission seems to be lacking toward the center of the H ii region. The enhanced dense gas, which is represented by the bright CO emission, exhibits a bubble-like structure in the 5–7 and 7–9 km s⁻¹ maps. We use a large circle with a diameter of 09 to show the bubble-like structure.

Zoom In Zoom Out Reset image size

In the northeastern region of the broken C¹⁸O bubble-like structure (i.e., near l ∼ 295 and b ∼ 40), several finger-like protrusions traced by ${}^{12}\mathrm{CO}$ and ${}^{13}\mathrm{CO}$ emission are found to point to the W40 H ii region (see the V_LSR ∼ 9–11 km s⁻¹ map in Figure 15). Moreover, the CO data reveal many interesting features, including rim-bright MCs, molecular gas bubbles/cavities, and shells/arcs in the complicated region. These features may be related to the strong feedback from the H ii region W40.

Figure 16 exhibits the velocity distribution of the C¹⁸O gas toward the W40 region. Three main velocity components are recognized, such as the partial shell structure surrounding W40 at V_LSR ∼ 5–6 km s⁻¹ (thick blue), the western half bubble-like structure at V_LSR ∼ 7–8 km s⁻¹ (thin blue to green), and the eastern broken bubble-like structure at V_LSR ∼ 9 km s⁻¹ (red). Some line-broadening features of ${}^{12}\mathrm{CO}$ gas (e.g., V_LSR ≲ 2 km s⁻¹ and V_LSR ∼ 12 km s⁻¹) are also revealed in the whole region, indicating the potential outflow candidates. Detailed analysis of the nearby region will be helpful for understanding the relationship between the molecular gas and ongoing star formation.

Zoom In Zoom Out Reset image size

3.3.3. CO Gas Distribution Perpendicular to the Galactic Plane

Dame & Thaddeus (1994) suggested that a thick molecular disk traced by faint CO emission is well above the central thin CO disk in the inner Galaxy. Their results are based on CO observations of the 1.2 m telescope for narrow regions of Galactic longitudes l = 30°, 40°, and 50° and Galactic latitudes $| b| \lt 4^\circ$. Benefiting from the unbiased, high-sensitivity, and large-area MWISP CO survey, we have a pretty good chance of studying the second, faint, and thick disk component of the molecular gas toward the inner Galaxy.

Channel maps of ${}^{12}\mathrm{CO}$ and ${}^{13}\mathrm{CO}$ (Figures 6 and 7) show that there are many MCs near the terminal velocity extending up to ∼1°–3° from the plane. Some MCs have extremely b values of >4°. These MCs at $| b| \gtrsim 1^\circ$ and V_LSR ≥ 60 km $\,{{\rm{s}}}^{-1}$ are relatively isolated with respect to the molecular gas near the Galactic plane. The MCs with weak CO emission, which are often associated with the enhanced H i emission (e.g., to compare with the GALFA data, Peek et al. 2011), have little velocity crowding at regions of relatively high b.

Peak ${}^{12}\mathrm{CO}$ velocities of V_LSR ≥ 60 km $\,{{\rm{s}}}^{-1}$ molecular gas are extracted pixel by pixel for the whole region of +258 ≲ l ≲  +497 and $| b| \,\lesssim$ 52. The pixels with three consecutive channels ≥3 × rms (or ∼0.8 K for the smoothed velocity resolution of 0.5 km s⁻¹) would be considered to be the valid signal. The near kinematic distances of the CO signal are calculated assuming the flat Galactic rotation curve model (e.g., Reid et al. 2014). Then, the distances from the b = 0° plane (i.e., z) are estimated for each pixel. To reduce sample fluctuations, z values are combined every 10 pc per bin from −500 pc to +400 pc. CO intensities are averaged for all valid CO signal per bin. The error bars are estimated from the express of I_mean(CO)/(number of pixels per bin)^0.5. As a consequence, the error bars of the high-z samples are obviously larger than those of points near the Galactic plane.

Figure 17 displays the vertical distribution of the CO gas. We find that one Gaussian model produces a poor fit to the histogram of the CO distribution from the Galactic plane (e.g., χ²/dof = 94.8/86). The FWHM of the one Gaussian component is 162.7 pc, which is larger than results from other studies (e.g., FWHM ∼ 90–110 pc for the inner Galaxy, e.g., Malhotra 1994a; Nakanishi & Sofue 2006, 2016; Heyer & Dame 2015; Roman-Duval et al. 2016). Moreover, the mean emission peak at z ∼ −11 pc and the prominent broad wing at $| z| \gtrsim 100$ pc cannot be fitted by the one Gaussian model.

Zoom In Zoom Out Reset image size

On the contrary, the model of two Gaussian components works better (χ²/dof = 60.5/84, also see the red-thick line in Figure 17). Table 2 shows the parameters of the best fit. The best fit of the zero-point is at ∼3.8 K km $\,{{\rm{s}}}^{-1}$, which is roughly 5σ of the CO integrated intensity. The value of the zero-point depends on the current rms level of the MWISP survey, in which samples with three consecutive channels greater than 3 × rms (∼0.8 K) are chosen as the valid ${}^{12}\mathrm{CO}$ signal of the MCs. A few points near the zero-point are from the noise (e.g., bad channels and baseline fluctuations in the velocity axis of the 3D data cubes). These bad points, which are randomly distributed in the area, are readily discerned by checking their spectra.

Table 2.  Two Gaussian Components for the Inner Galactic CO Plane

Narrow			Broad
z0 (pc)	FWHM (pc)	I0 (K km $\,{{\rm{s}}}^{-1}$)		FWHM (pc)	I0 (K km $\,{{\rm{s}}}^{-1}$)	χ2/dof
−11.3	88.5	40.9		276.8	20.8	60.5/84a

Notes. z₀ is the displacement of the CO gas from the Galactic plane of b = 0°; FWHM is the thickness of the CO layer; I₀ is the peak of the best fit. Note that the zero-point is at 3.8 K km $\,{{\rm{s}}}^{-1}$, i.e., I₀ = I_peak = I(z₀) + 3.8 K km $\,{{\rm{s}}}^{-1}$ (see Figure 17 and the text). Note that the z₀ value of the broad component is fixed to that of the narrow one. Samples are from the CO emission near the tangent point, i.e., V_LSR >  60 km s⁻¹.

^aThe residual is calculated with a weight of the mean CO intensity per bin.

Download table as:  ASCIITypeset image

Samples away from the Galactic plane with values above 3.8 K km $\,{{\rm{s}}}^{-1}$ are actually from the high-z CO emission. An example is the unusual peak at z ∼ −435 pc (Figure 17). CO emission of this feature is mostly from MC G40.331−4.302 at V_LSR ∼ 77–84 km s⁻¹. The high $| b|$ MC is associated with the W50 nebula, which has a near kinematic distance of 4.9 kpc, as estimated from the large-scale gas at V_LSR = 77 km s⁻¹ (see the detailed H i and CO analysis in Su et al. 2018). At the distance of 4.9 kpc, the true displacement from the b = 0° plane is z ≈ −370 pc for the b = −4302 molecular gas, which is likely related to the energetic jets of the microquasar SS 433 in W50 (i.e., jet–ISM interactions discussed in Su et al. 2018).

The two Gaussian components correspond to the central thin molecular gas disk (FWHM = 88.5 pc) and the extended thick CO disk (FWHM = 276.8 pc), respectively. The thickness ratio of the two Gaussian components is FWHM(Thick plane)/FWHM(Thin plane) ∼ 3.1 for the region of Galactocentric distances of R_GC ∼ 3.9–6.4 kpc (mean at ∼5.4 kpc for all samples) toward the inner Galaxy. These results are consistent with previous studies (Dame & Thaddeus 1994; Malhotra 1994b). The total intensity ratio between $| z| \gtrsim 100$ pc and $| z| \lesssim 100$ pc is about 0.03, indicating very faint CO emission for the second (or thick) molecular gas component.

In the above calculation, we used the A5 rotation curve model of Reid et al. (2014), in which the values of R₀ = 8.34 kpc and V₀ = 240 km s⁻¹ are adopted. The thickness of the molecular gas disk will increase by a factor of ∼1.1 for the parameters of R₀ = 8.5 kpc and V₀ = 220 km s⁻¹.

We have shown that CO emission in the first quadrant of the inner Galaxy is somewhat below the Galactic plane based on channel maps in Figure 6. Further quantitative analysis indicates that the distribution of molecular gas is actually below the b = 0° plane for the V_LSR ≳ 60 km s⁻¹ CO emission (Figure 17 and Table 2). Assuming that the inner disk of our Galaxy is flat and tilted, we derive the Sun's offset from the Galactic physical midplane based on the molecular gas distribution along the Galactic latitude. Considering $\displaystyle \frac{{z}_{\mathrm{Sun}}}{{R}_{0}}\approx \displaystyle \frac{{{z}}_{\mathrm{peak}}}{{R}_{\mathrm{GC}}(\mathrm{mean})}$, the Sun is about $8.34\times \displaystyle \frac{11.3}{5.4}\approx 17.5$ pc above the physical midplane of the Milky Way, which agrees well with our previous study of ∼17.1 pc (Su et al. 2016).