I’m an SMA Fellow at the Harvard-Smithsonian Center for Astrophysics. My research, which is described in more detail below, focusses on understanding the planet formation environment, the protoplanetary disk, and involved processes, using radio observations of simple molecular species. As part of this, I write a lot of code which I try to make open source. Details of the main packages can be found below, or by exploring my GitHub.
Before my position at the CfA, I was a postdoctoral researcher at the University of Michigan working in the group of Prof. Ted Bergin. Before that, I studied at the Unversity of Edinburgh for my MPhys in Astrophysics, then at MPIA in Heidelberg where I obtained my PhD in Astronomy, supervised by Prof. Thomas Henning and Dr. Dmitry Semenov.
My research can be can be broadly broken down into a few major themes, but all involve the study of protoplanetary disks using primarily sub-mm osbervations. My work has also been feature in several press releases, which I have linked to below.
Withnessing the Formation of Planetary Systems
If we are to really understand the planet formation process, we must first find the still-forming planets. We have recently made the first detection of a circumplanetary disk, a disk around a still-forming planet believed to be the source for moons. However, this approach is only applicable to the inner disk where large, mm sized grains exist. Therefore I am developing methods to understand the limits of what we are able to detect from the gas, by searching for local disturbances in the rotation speed of the gas. This allowed us to infer the presence of unseen planets in HD 163296, a project which we hope will lay the foundations for more detections of still-forming planets.
Dynamial Structure of Protoplanetary Disks
By studying the motions of the gas in the protoplanetary disk, we also can gain unique insights into the physical processes which are present in the disk. In addition to searching for the signs of embedded planets, we can search for hydrodyanamical proccesses which transport material and angular moment throughout the disk. Using a newly developed technique, we have been able to infer the first 3D velocity map of a disk, confirming the presence of 'meridional flows', demonstrating an efficient way to transport material from the disk atmosphere to the planet-forming midplane.
While protoplanetary disks do not show the chemical complexity found in earlier phases of star formation, the molecules which are detected provide a tremendous amount of information about the underlying physical and chemical structures. I am interested in using molecular excitation to infer physical properties of the disk, such as the gas temperature, density or ionization level. By exploiting our astrochemical knowledge, we can use different molecules to trace different regions in the disk and unravel the full structure of the disk. We have used this technique to trace perturbations in the disk of TW Hya, connecting to features observed with the scattered light.
Open Source Analysis Tools
A suite of tools written in Python to extract kinematical information from spectrally resolved line emission in protoplanetary disks. eddy has been used to infer the presence of unseen planets and make the first confirmation of grain trapping in pressure maxima. This currently contains methods to fit Keplerian rotation patterns to first-moment maps including 3D flared surfaces, as well as measuring precise 3D velocities of the gas.
The code is described mainly in this paper.
Tired of having to play around with masks and clipping thresholds to get a nice looking rotation or emission map? bettermoments is here to help! This is a simple package to quickly make moment maps from position-position-velocity cubes without the need for clipping. And, as a bonus, returns uncertainties!
A short article about this method used can be found at the Research Notes of the AAS.
Exploit the known rotation of a protoplanetary disk to stack emission lines from an annulus to significantly boost the signal to noise of the detection (or to detect the line at all!). This also has the added bonus that we can resample the line profile to better distinguish any deviations from a commonly assumed model.
The code is described in this Joss article, while the documentation gives examples of how to use it with your data.