I am a Ph.D. candidate in Astrophysics at the University of California, Berkeley. I work at the interface of theory and radio observations of Cosmic Dawn and the Epoch of Reionization. I received my B.S. at the University of Michigan, Ann Arbor in Physics and Astrophysics.
nkern@berkeley.edu | github.com/nkern
See below for some research highlights.
I am leading the effort to analyze HERA Phase I data and produce its first upper limit on the 21cm power spectrum at the EoR (HERA Collaboration et al. in prep.). Data analysis pipelines for 21cm science require ultra-precise control of systematics. I have worked on developing hybrid calibration schemes for HERA given its observational constraints (Kern et al. 2019c), and have worked on instrumental systematic identification, modeling and removal (Kern et al. 2019a, Kern et al. 2019b).
Connecting data to theoretical predictions from sophisticated and compute-intensive simulations is often prohibitively expensive using Markov Chain Monte Carlo (MCMC) samplers. A solution is to use machine learning methods to emulate the mapping of the simulation's input parameters to its output products with a less computationally expensive surrogate model. We explored this in the context of a cosmological 21cm simulation, and demonstrated a parameter constraint forecast jointly across cosmological LCDM, EoR star formation and Cosmic Dawn X-ray heating parameters using a full MCMC sampler and a 1000-hour mock HERA observation (Kern et al. 2017).