2000 character limit reached
ALMA CO Observations of a Giant Molecular Cloud in M33: Evidence for High-Mass Star Formation Triggered by Cloud-Cloud Collisions
Published 22 Aug 2019 in astro-ph.GA | (1908.08404v4)
Abstract: We report the first evidence for high-mass star formation triggered by collisions of molecular clouds in M33. Using the Atacama Large Millimeter/submillimeter Array, we spatially resolved filamentary structures of giant molecular cloud 37 in M33 using ${12}$CO($J$ = 2-1), ${13}$CO($J$ = 2-1), and C${18}$O($J$ = 2-1) line emission at a spatial resolution of $\sim$2 pc. There are two individual molecular clouds with a systematic velocity difference of $\sim$6 km s${-1}$. Three continuum sources representing up to $\sim$10 high-mass stars with the spectral types of B0V-O7.5V are embedded within the densest parts of molecular clouds bright in the C${18}$O($J$ = 2-1) line emission. The two molecular clouds show a complementary spatial distribution with a spatial displacement of $\sim$6.2 pc, and show a V-shaped structure in the position-velocity diagram. These observational features traced by CO and its isotopes are consistent with those in high-mass star-forming regions created by cloud-cloud collisions in the Galactic and Magellanic Cloud HII regions. Our new finding in M33 indicates that the cloud-cloud collision is a promising process to trigger high-mass star formation in the Local Group.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Explain it Like I'm 14
What this paper is about
This study looks at how very big, hot stars are born in another galaxy called M33 (the Triangulum Galaxy), about 2.7 million light-years away. The authors used a powerful radio telescope (ALMA) to zoom in on a giant cloud of gas and dust and found signs that two clouds crashed into each other. That crash seems to have squeezed the gas enough to spark the birth of up to about ten high‑mass stars.
The main questions the researchers asked
- Can we find clear, detailed evidence that two gas clouds in M33 ran into each other?
- If so, did that collision trigger the birth of high‑mass stars (the kind that shine very brightly and live short lives)?
- Could the motion we see be explained instead by the stars pushing the gas around after they formed, or does a collision fit better?
How they did the study (in simple terms)
- Tool: They used ALMA, a network of radio telescopes in Chile that can “see” cold gas clouds. ALMA can resolve details just a couple of parsecs across (about 6–7 light‑years), which is very sharp for another galaxy.
- What they observed: They mapped a giant molecular cloud (a star‑forming gas cloud) called M33GMC 37 using three versions of carbon monoxide gas:
- 12CO, 13CO, and C18O. Think of these like three kinds of “highlighters” that trace gas of different densities. C18O tends to pick out the densest parts where stars are forming.
- Other clues they used:
- 1.3 mm “continuum” emission (a kind of glow from dust warmed by nearby young stars).
- Hα light (a red glow from hydrogen gas lit up by young, hot stars).
- Spitzer infrared data and Hubble Space Telescope images to spot young massive stars.
- How they looked for a collision:
- Position–velocity diagrams: Graphs showing where gas is on the sky and how fast it’s moving. A cloud collision can leave a “V-shaped” pattern or a “bridge” connecting two speed groups.
- Complementary shapes: If one cloud slammed into another, you might see puzzle‑piece shapes: a dent or hole in one cloud that matches a bump in the other, often shifted a bit on the sky.
What they found and why it matters
- Two clouds moving at different speeds:
- They clearly see two separate clouds along the same line of sight, one “blue” and one “red,” moving at different speeds with about a 6 km/s difference. That’s fast enough (supersonic in these clouds) to cause strong compression when they collide.
- Newborn high‑mass stars in the crash zone:
- Three compact dust sources (seen at 1.3 mm) and bright Hα/infrared emission mark where stars are forming.
- Hubble images reveal around 10 young, massive stars. Using the light they emit, the team estimates their types to be between B0V and O7.5V—these are big, hot stars.
- Classic collision signatures:
- Complementary “puzzle‑piece” shapes: The two clouds fit together best when shifted by about 6.2 pc (around 20 light‑years), as if one gouged a cavity in the other.
- A V-shaped pattern in the position–velocity diagram: This is a known sign of two clouds decelerating as they interact and overlap.
- Dense gas (traced by C18O) and the new stars sit right where the clouds overlap, just where a collision would squeeze gas the most.
- Not explained by stellar pushback:
- Could the stars’ winds have blown the gas apart and created the observed motion? The team compared the “oomph” (momentum) in the stellar winds to the momentum in the clouds. The clouds win by about 100×. So the observed speeds are not mainly caused by the young stars pushing the gas around.
- Timing makes sense:
- Based on the shift between the cloud shapes and their speed difference, the collision seems to have started roughly 1 million years ago—young enough that we still see dense gas and small, bright H II regions (glowing gas around newborn stars).
Why it matters: These combined clues strongly point to a cloud–cloud collision as the trigger for high‑mass star formation in this part of M33. That adds to growing evidence that such collisions are a common way massive stars get started, not just in our Milky Way but in other nearby galaxies too.
What this means going forward
- This is the first strong case in M33 showing high‑mass stars likely formed by two gas clouds crashing together. It suggests cloud collisions are a key, perhaps universal, process for creating massive stars in the Local Group of galaxies.
- Understanding this helps answer a long‑standing question in astronomy: how do the biggest stars form? Collisions can rapidly compress gas, boosting the mass‑feeding rate onto baby stars so they can grow big before their own radiation pushes gas away.
- Future ALMA studies across M33 can test how common and important this process is and compare it with similar regions in the Milky Way and the Magellanic Clouds.
Quick recap
- Two gas clouds in M33 are moving past each other at about 6 km/s.
- Their shapes fit together like puzzle pieces and make a V-shape in speed–position plots—both classic collision signs.
- Right where they overlap, there are dense gas clumps and about 10 newborn high‑mass stars.
- Star winds can’t explain the gas motions; a collision can.
- This supports the idea that cloud collisions can kick‑start the birth of massive stars in other galaxies too.
Knowledge Gaps
Knowledge gaps, limitations, and open questions
The paper presents compelling evidence for cloud–cloud collision (CCC)–triggered high-mass star formation in M33GMC 37, but several aspects remain uncertain or unexplored. The following concrete gaps could guide future work:
- Massive stars and stellar content
- No spectroscopic classification for the ≲10 candidate high-mass stars; spectral types are inferred from integrated Hα and 24 μm luminosities under the simplifying assumption that all stars share the same type. What are the individual spectral types, extinctions, ages, and memberships from color–magnitude diagrams and spectroscopy?
- Extinction corrections for Hα are not explicitly derived for this region; how does internal extinction affect the inferred Lyman continuum and the number of ionizing stars?
- The association of all ∼10 bright HST sources with the CO/continuum/Hα structures is unconfirmed; what fraction are bona fide cluster members versus foreground/background?
- Continuum emission and core properties
- The 1.3 mm continuum detection (∼0.1–0.13 mJy beam⁻¹) is not decomposed into dust vs. free–free; without multi-frequency SEDs, core masses, temperatures, and dust emissivities remain unconstrained.
- No estimates of gas mass, density, or temperature for MMSs 1–3 from continuum or line modeling; are these bound cores, UC/H II regions, or compact clusters?
- H(30α) and SiO(5–4) were non-detections, but sensitivity limits are not tied to expected levels; how deep must observations be to confirm or rule out free–free and shock tracers?
- Kinematic evidence for collision
- The relative velocity between the two clouds is modest (∼6 km s⁻¹); the “bridging feature” is weak/ambiguous and largely inferred from single-peak profiles at the mean velocity. How robust is the CCC interpretation against alternative kinematic models (e.g., shear, streaming motions along a spiral arm, rotation within a single GMC, inflow along filaments)?
- The V-shaped PV signature is shown along one cut; is it persistent across different cuts and robust to spatial averaging? Could similar PV morphologies arise from non-collisional processes?
- The complementary spatial distribution is quantified via a Pearson correlation (minimum −0.44) using selected polygons and 12CO only; how sensitive is the result to choice of region, tracer (13CO, C18O), and spatial filtering? A Monte Carlo or surrogate-data test for chance anti-correlation is missing.
- Geometry, timescales, and projection effects
- The collision timescale (~1 Myr) assumes a 45° inclination; the true 3D geometry is unknown. What are the allowed ranges of timescale and displacement under plausible angles?
- Ages of the massive stars and H II regions are not derived; can stellar ages (from spectroscopy/photometry) be reconciled with the inferred collision timeline?
- The head-on vs. oblique collision distinction is inferred from PV morphology, but not tested with tailored synthetic observations spanning viewing angles and cloud morphologies.
- Physical conditions and shock diagnostics
- No non-LTE radiative transfer (e.g., LVG/RADEX) modeling to derive kinetic temperatures, densities, and optical depths from 12CO/13CO/C18O; how do gas conditions vary across the putative collision layer?
- No detection (or systematic search) of shock-/dense-gas tracers (e.g., SiO, HCN, HCO⁺, CS), which could directly test for shocked, compressed gas at the interface.
- Magnetic fields are invoked in the effective sound speed but not measured; what is the B-field strength and topology (e.g., via dust polarization or Zeeman), and how do they modify fragmentation and accretion?
- Feedback and alternative triggering mechanisms
- The momentum budget analysis considers stellar winds but not photoionization pressure, radiation pressure, or the cumulative effect of multiple B/O stars; could these alter cloud kinematics locally?
- The region contains/abuts M33SNR 35; possible SNR or older H II-region expansion effects on the gas dynamics are deferred to future work and not ruled out here.
- Masses, conversion factors, and calibration
- Gas masses rely on a fixed X_CO appropriate for M33 and a uniform 2–1/1–0 ratio (0.7); metallicity and excitation vary across M33, and X_CO may be spatially variable. How do these assumptions bias cloud masses and column densities?
- Masses and virial parameters of the individual blue/red sub-clouds are not assessed; are the sub-clouds gravitationally bound and consistent with the collision scenario energetically?
- Interferometric limitations and spatial filtering
- Maximum recoverable scale (∼3.7″ ≈ 15 pc) may filter diffuse emission; missing-flux checks were performed for 12CO only at 12″ smoothing, not for 13CO/C18O. Could filtered extended emission affect PV morphology and complementarity?
- Single-pointing ALMA coverage may not encompass the full extents of the two clouds; without mosaics, the global pre-/post-collision morphology is incomplete.
- Star formation activity and youth indicators
- Outflow signatures are tentative (a small CO wing/tiny clump); no systematic search or mapping of high-velocity gas or masers (H₂O/CH₃OH) to confirm ongoing massive star formation.
- No cm-wave radio continuum analysis to constrain free–free emission and ionizing flux independent of extinction.
- Environmental and statistical context
- The broader HI/CO environment and spiral-arm dynamics at sub-10 pc scales are not incorporated; how do large-scale flows feed or bias the local collision?
- This is a single case study; how frequent are CCC-induced massive star-forming sites in M33, and how do their properties compare statistically to Milky Way analogs?
- Model–data consistency and predictive tests
- The schematic scenario is not confronted with dedicated MHD simulations tailored to observed masses, velocities, and morphologies; can simulations reproduce the observed PV, hole, and displacement simultaneously?
- Predictions for future tests (e.g., spatial distribution of shock tracers, polarization patterns, age gradients across the interface, proper motions of ionized gas) are not articulated.
Addressing these gaps will require multi-frequency continuum (cm–mm) and spectral-line follow-up (dense/shock tracers), stellar spectroscopy and photometry (ages, types), magnetic field measurements, HI context, wider ALMA mosaics, and tailored MHD simulations with synthetic observations to test the CCC hypothesis against alternatives.
Glossary
- Alfven speed: The speed at which magnetic disturbances travel in a magnetized plasma. "Here, the effective sound speed is defined as , where is the sound speed, is the Alfven speed, and is the velocity dispersion."
- ALMA Band 6: A specific frequency band (211–275 GHz) of the ALMA telescope used for observations. "We carried out ALMA Band~6 {(211--275~GHz)} observations toward M33GMC~37 in Cycle~6..."
- Atacama Large Millimeter/submillimeter Array (ALMA): A large radio interferometer for millimeter/submillimeter astronomy. "Using the Atacama Large Millimeter/submillimeter Array, we spatially resolved filamentary structures of giant molecular cloud 37 in M33..."
- Atacama Submillimeter Telescope Experiment: A single-dish submillimeter telescope used for CO observations. "These values are roughly consistent with the previous study by \citet{2012ApJ...761...37M} using the Atacama Submillimeter Telescope Experiment..."
- Bandpass calibrator: A source observed to calibrate the frequency response of a radio interferometer. "Two quasars, J22531608 and J02372848 were observed as bandpass and flux calibrators."
- Baseline length: The distance between antennas in an interferometric array, affecting resolution and spatial scales. "The baseline length ranges from 15.06 to 1397.85 m..."
- Brightness temperature: A measure of specific intensity expressed as an equivalent temperature. "The peak brightness temperatures of the three sources are comparable..."
- Bridging feature: An intermediate-velocity CO component connecting two clouds, created by collisional deceleration. "The colliding clouds have a supersonic velocity {difference} with an intermediate velocity component---bridging feature---{created} by the collisional deceleration."
- CO( = 2--1): A rotational transition of the C18O isotopologue used to trace dense molecular gas. "CO( = 2--1) line emission at a spatial resolution of pc."
- CASA (Common Astronomy Software Application): Software used for calibration and imaging of radio astronomy data. "The data reduction was performed using the Common Astronomy Software Application (CASA; \cite{2007ASPC..376..127M}) package version 5.5.0."
- Cloud-cloud collisions: Interactions between molecular clouds that can compress gas and trigger star formation. "We report the first evidence for {high-mass} star formation triggered by collisions of molecular clouds in M33."
- CO-to-H conversion factor (X): A factor converting CO integrated intensity to H2 column density. "where is the CO-to-H conversion factor in units of (K km s) cm..."
- Column density: The amount of material along the line of sight, usually in cm. "The peak column density of molecular hydrogen is cm for the blue-cloud..."
- Complementary spatial distribution: Spatial anti-correlation where one cloud’s structures fit into cavities of another, indicating interaction. "The two molecular clouds show a complementary spatial distribution with a spatial displacement of {} pc..."
- Continuum sources: Emission sources without spectral lines, often tracing dust or compact objects. "There are three sources in 1.3 mm continuum {as shown in Figure \ref{fig2}g}---{GMC37-}MMSs~1--3..."
- Declination--velocity diagram: A slice of data showing velocity versus declination, used to analyze kinematics. "Declination--velocity diagram of CO( = 2--1)."
- Effective Jeans mass: The mass scale for gravitational collapse considering thermal, magnetic, and turbulent support. "cloud-cloud collision increases the {effective Jeans mass} {so high-mass stars can} form {the high-mass stars} in the shock-compressed layer..."
- Effective sound speed: Combined speed accounting for thermal, magnetic, and turbulent motions in gas. "Here, the effective sound speed is defined as ..."
- Field of view (FoV): The area of the sky covered by an observation. "The blue circle indicates the ALMA FoV."
- Flux calibrator: A source observed to set the absolute flux scale. "Two quasars, J22531608 and J02372848 were observed as bandpass and flux calibrators."
- Forward efficiency: A telescope efficiency term used to convert measured temperatures to main-beam scale. "we applied a forward efficiency of 0.92 and a main beam efficiency of 0.56..."
- FWHM resolution: The full width at half maximum used to characterize spatial resolution. "smoothed to match the FWHM resolution of $12\"$."
- Giant Molecular Cloud (GMC): A large complex of molecular gas where stars form. "{A giant molecular cloud} (GMC)---M33GMC~37---is mainly located on the western-half of the ALMA FoV..."
- H(30): A hydrogen radio recombination line used to trace ionized gas. "{Although these continuum bands contain line emission {of} H(30) and SiO( = 5--4)...}"
- H{\sc i} disk: The distribution of neutral atomic hydrogen in a galaxy. "the systemic velocity of the M33GMC~37 region which was derived by a rotation model of the H{\sc i} disk..."
- H{\sc ii} region: A region of ionized hydrogen around young, massive stars. "{The ALMA FoV includes several H{\sc ii} regions located in a spiral arm near the galactic center.}"
- H emission: Optical emission line from ionized hydrogen, tracing star formation. "The superposed {white and black} contours indicate the H intensity obtained by the Kitt Peak National Observatory..."
- IRAM 30-m radio telescope: A single-dish millimeter telescope used for CO observations. "To estimate the missing flux {density}, we used the CO( = 2--1) dataset obtained with the IRAM 30-m radio telescope..."
- Integrated intensity: The velocity-integrated line brightness, typically in K km s. "Maps of moment 0 (integrated intensity: Figures \ref{fig2}a, \ref{fig2}d, and \ref{fig2}g)..."
- Kitt Peak National Observatory (KPNO): An observatory providing Hα data used in the study. "The superposed {white and black} contours indicate the H intensity obtained by the Kitt Peak National Observatory (KPNO, \cite{2006AJ....131.2478M})."
- Largest angular scale (maximum recoverable scale): The largest structure an interferometer can image without missing flux. "The maximum recoverable scale (a.k.a. largest angular scale) is calculated to be 3\farcs73."
- Local Group: The galaxy group containing the Milky Way and M33. "cloud-cloud collision is a promising process to trigger {high-mass} star formation in the Local Group."
- Local Standard of Rest (V): A velocity reference frame used in Galactic astronomy. "We hereafter refer to the component at = {}$143.0$-- km s as the ``blue cloud''..."
- Lyman continuum luminosity: The number of ionizing photons inferred from Hα luminosity. "Lyman continuum luminosities (in units of photons) that are derived from extinction corrected H luminosities ..."
- Magnetohydrodynamical numerical simulations: Simulations of magnetized gas dynamics used to model star formation. "According to magnetohydrodynamical numerical simulations, the effective {J}eans mass in the shock-compressed layer is proportional..."
- Main beam efficiency: The fraction of power in a telescope’s main beam used in calibration. "we applied a forward efficiency of 0.92 and a main beam efficiency of 0.56..."
- Mass accretion rate: The rate at which mass is accumulated onto a forming star. "A supersonic velocity separation {of} at least a few km s therefore produces a large mass accretion rate on the order of -- yr..."
- Milliarcsecond (mas): An angular unit equal to one-thousandth of an arcsecond. "The size of a pixel is 84.5 mas."
- Moment 0 (integrated intensity): The zeroth spectral moment; velocity-integrated line emission. "Maps of moment 0 (integrated intensity: Figures \ref{fig2}a, \ref{fig2}d, and \ref{fig2}g)..."
- Moment 1 (peak velocity): The first spectral moment indicating intensity-weighted mean velocity. "{Figures \ref{fig2}b, \ref{fig2}e, and \ref{fig2}h show the maps of moment 1 (peak velocity).}"
- Moment 2 (velocity dispersion): The second spectral moment indicating velocity spread. "Figures \ref{fig2}c, \ref{fig2}f, and \ref{fig2}i show the maps of moment 2 (velocity dispersion)."
- Multiscale CLEAN algorithm: A deconvolution technique for interferometric imaging. "We used the ``multiscale CLEAN'' algorithm implemented in the CASA package..."
- Nobeyama Radio Observatory 45-m telescope: A radio telescope used to obtain CO(J=1–0) data. "an archival CO( = 1--0) cube data obtained with the Nobeyama Radio Observatory 45-m telescope..."
- Position angle (P.A.): The orientation of the synthesized beam or structure measured on the sky. "The synthesized beam of final dataset is $0\farcs59 \times 0\farcs42$ with a position angle (P.A.) of $0\fdg4$..."
- Position--velocity diagram: A plot showing spatial position versus velocity to analyze kinematics. "a V-shaped structure can be seen in the position-velocity diagram due to the deceleration and hollowed-out structure."
- Shock-compressed layer: Dense gas region formed by compression during a collision. "form {the high-mass stars} in the shock-compressed layer..."
- SiO( = 5--4): A rotational transition of silicon monoxide used to trace shocks/outflows. "{Although these continuum bands contain line emission {of} H(30) and SiO( = 5--4), we could not detect the two lines significantly.}"
- Spectral windows: Defined frequency ranges in radio observations for lines or continuum. "There were two spectral windows including the CO( = 2--1), CO( = 2--1), and CO( = 2--1) line emission..."
- Supersonic velocity difference: Relative cloud speeds exceeding the sound speed, enhancing compression. "The colliding clouds have a supersonic velocity {difference}..."
- Synthesized beam: The effective point-spread function of an interferometric image. "The synthesized beam of final dataset is $0\farcs59 \times 0\farcs42$..."
- Systemic velocity: The characteristic velocity of a region relative to the observer. "corresponding to the systemic velocity of the M33GMC~37 region..."
- u--v distances: Baseline lengths expressed in spatial-frequency units for interferometry. "corresponding to {\it{u--v} distances from 11.6 to 1074.9 ."
- Velocity dispersion: The spread of velocities within a gas component. "Figures \ref{fig2}c, \ref{fig2}f, and \ref{fig2}i show the maps of moment 2 (velocity dispersion)."
- Virial mass: An estimate of mass assuming gravitational equilibrium. "virial mass of ."
- V-shaped structure: A characteristic pattern in position–velocity space indicative of collisional dynamics. "a V-shaped structure can be seen in the position-velocity diagram..."
Collections
Sign up for free to add this paper to one or more collections.