About Me

I am a former astronomer with a PhD in Physics and Computer Science from the University of Missouri—Kansas City. I'm now a data scientist working in the geospatial industry. I'm currently working for Alynix LLC on a project also called DECKER (no relation to me, it was named before I was hired!) that uses infrared images from drones to find areas of concrete delamination in highway bridges. It is a fascinating application of basic scientific principles, and I have thoroughly enjoyed using my background in astronomy and infrared imaging to guide the research stage of the project from a disjointed series of coding libraries to a full data reduction and analysis pipeline. I like to joke that I have gone from infrared cameras on the ground looking at the sky to infrared cameras in the sky looking at the ground!

Between academia and industry I now have ten years of experience working directly with imaging data as well as analysing and modelling large and often disparate data sets. I work primarily in python and I have over a decade of experience doing scientific computing, including implementing standard machine learning models in python. Part of my work with Alynix has been to migrate our pipeline to a cloud-based approach using AWS S3 and EC2, and I am also the owner of the Alynix GitHub repositories.

As an academic, I was part of the Massive and Distant Clusters of WISE Survey (MaDCoWS) collaboration, studying high-redshift infrared-selected galaxy clusters. MaDCoWS is a large-area infrared survey, using all-sky WISE data and optical survey data to identify overdensities of galaxies at z~1. MaDCoWS is the only all-sky cluster survey at this redshift, and as such is sensitive to some of the most massive objects in the universe at this redshift.

My dissertation work focussed on the stellar mass properties of a subset of MaDCoWS clusters at z~1.1. Using a combination of mid-infrared data from the Spitzer Space telescope and optical data from the GMOS cameras on the Gemini telescopes, I measured stellar mass fractions and deep luminosity functions for this subset of infrared-selected MaDCoWS clusters. I compared the stellar mass fractions as a function of total mass to the same measurements of ICM-selected clusters from the South Pole Telescope survey. Because infrared luminosity is directly related to stellar mass, it was possible that MaDCoWS clusters would be biased toward finding clusters with more stellar mass than ICM-selected clusters. I did not find a large difference in the stellar mass fractions, however, with only a hint that there might be more scatter in the SPT clusters. My study of the luminosity functions showed a similar consistency between MaDCoWS clusters and previous surveys. Using an algorithm to jointly fit the properties of the Schechter function, I found there was significant evolution, but that the evolution was consistent with passive evolution. Importantly, this evolution was only detectable when performing a simultaneous fit to the Schechter function parameters.

I'm an expert in both optical and radio data reduction, having done both extensively for MaDCoWS. I developed the reduction pipeline for the MaDCoWS Gemini observations, writing a script to reduce the raw GMOS imaging, apply the necessary astrometric corrections, and create a multi-band optical catalogue. This has evolved considerably over time, going from IRAF to Python, and also calling SCAMP, SWarp and SExtractor. I have also travelled to Las Campanas Observatory twice to take data for MaDCoWS with the Magellan telescopes. I have been heavily involved in applying for and reducing interferometric radio data for MaDCoWS. I used CARMA SZ data to measure the masses of several of the clusters in my 2019 paper, as well as the mass of MOO J1142+1527, the most massive cluster yet discovered at z > 1.15. I was also a part of the successful MaDCoWS ALMA proposal, and performed the initial reduction of those data. Since then, I have been to the single-dish Green Bank Observatory twice and am a certified remote observer. I have been one of the observers on all of the MaDCoWS GBT projects.

As is hinted at above, I am an extremely talented programmer. In addition to the aforementioned Gemini reduction pipeline, I wrote a Bayesian fitting algorithm to more quickly perform the joint fit of the Schechter function required for my dissertation work. There are a huge number of other scripts that go into my work as well. By the time I finished it, the entire analysis and output plots for my second paper could be done and generated with barely more than one line of input at the terminal! I am familiar with many computer learning algorithms as well; some of my initial analyses have made use of them and they are very promising for future classification work. In my last year in academia, I also taught the Department of Physics and Astronomy Research Training Skills course. This is a course designed to cover the basics of Python programming and data analysis for aspiring undergrad researchers.

I'm originally from the Kansas City area and I got my Bachelor's degree in physics from Truman State University in Kirksville, MO. Although it is very hard to find a suitably dark sky in Kansas City, I enjoy talking about astronomy and host occasional outreach events. Outside of astronomy I have a great love of sport, in particular baseball and cricket. I am a lifelong (and intermittently-suffering) Royals fan as well as a supporter of Lancashire and England. I don't have a lot of spare brain-power after work, but I love digging into baseball statistics, and I am trying to get back into the habit of blogging statistical analyses. I find a lot of discourse around advanced metrics could use a scientific approach! I also do some recreational Python coding, and enjoy hiking, playing golf, and generally spending time outdoors, particularly with my significant other.