What DUVET aims to do: DUVET is a survey of 30 galaxies in which we make resolved (0.1-1 kpc) observations of outflows in galaxies with extreme star formation rates. The DUVET team aims to directly test subgrid physics related to star formation feedback. Feedback based models of star formation are known to have difficulties in the limit of high star formation, and our observations are therefore designed specifically to learn more about this regime, in which most cosmic star formation takes place. We simultaneously probe interstellar gas and stellar populations through strong emission lines [OII], [Ne III], Hβ and [OIII], as well as utlra-faint emission lines like [OIII] 4363 and He II. We can directly compare physical properties to outflow properties in thousands of lines-of-sight. Using DUVET observations we also push strong line detections well beyond the edge of the stellar disk. There is a wealth of science possible with this data set, and we are excited to get started.

DUVET will measure both the kinematics of the gas expelled by supernova and the details of the source of that outflow in 10,000 lines-of-sight in 30 galaxies.

DUVET is now beginning to produce results.

You can see an overview of the survey in this award winning talk by our excellent PhD student Bron Reichardt-Chu.

You can also read Bron’s recently accepted paper on the outflows in one of our pilot targets here.

The DUVET team has recently published the first ever mapping of metal enrichment starting from gas inflowing on the minor axis, then in the galaxy, and finally into the outflow. The reason we can do this is because our observations are extremely sensite and thus able to track the robust auroral line tracers of metallciity. The figure on the left shows our metallicity map and a schematic of how we interpret these observations.

Read Cameron et al. (2021) for more on metal enrichment of outflows.

Physical properties of gas in the galaxy and the flow: We first measure the outflow on the [OIII] 5007 line and then carry out more informed fits to both Hβ and [OII] doublet. We do this because the later lines have either absoprtion features or blending that makes deriving fits independant of information more challenging. We can then make estimates of the physical properties of both the gas in the outflow and the galaxy in each spaxel.

To measure outflows in edge-on galaxies we measure a set of emission lines extending to roughly 10 kpc above the plane of disk. We can then determine the surface brightness of gas, the metallicity, the electron density and ionization state of gas in the outflow. The outflow component of galaxies, in this case, is identified as regions of the disk with larger line width. The large line-width is due to both the turbulence of the gas and the expanding biconical structure of the outflow. DUVET observations will therefore make a systematic study of the opening angle and radial extent of outflows. The image on the right shows the outflowing gas from MRK1486, one of the DUVET galaxies, we can directly see the outflow in this HST, with DUVET observations we cover a much larger distance from the galaxy and measure the kinematics of the expanding outflow cone.

How do we measure outflows?

We use two methods to detect outflows (1) decomposing the spectral lines in face-on galaxies and (2) direct imaging of the gas above the plane of the disk in edge-on galaxies. Together these give us a holistic view of outflow properties for given galaxy mass and star formation rates.

In face-on galaxies outflows are fit by decomposing spectral features into multiple components. In galaxies, when supernova explode the gas will expand away from the supernova, and entrain more gas within them as they move. If something is moving toward you at high velocity the well known Doppler effect will cause this gas to have a bluer wavelength than the gas that is stationary with respect to you. The total spectral line is therefore a combination of both galaxy emission and outflow emission.

A challenge for DUVET is that outflows are have hystorically been made with a single measurement for each galaxy, which can be monitored and tuned by the user. However, for DUVET we intend to measure thousands of outflows, with hunndreds in each galaxy. By hand methods cannot be used in our analysis. PhD student Bron Reichardt Chu has developed a code, called KOFFEE, implementing a series of tests that automatically determine if an extra outflow component is needed in the fit, then derive the appropriate functional fit to each spaxels. We also consider the impact of beam smearing on fits, as well as noise structure. KOFFEE is currently still in development (mid-2020), but will be made publically available soon.