We live in an elegant Universe that amply rewards observing and pondering. Trying to understand Nature boosts our intellectual appetite as it raises more questions than we can answer and allows us to reach deep realizations that stretch beyond the time and space we live in. Personally, this gives me intellectual satisfaction and makes life more meaningful.
During my career as a scientist, I had the pleasure to have contributed to the investigation of various problems in astrophysics, including the discovery and characterization of planets beyond our Solar System and the search for dark matter. I build or use statistical methods to test, select or refine hypotheses under the light of observations. Although I am easily attracted to all wonders of our Universe, I find exoplanets and dark matter especially interesting, because they relate to the cosmological origins of our planet Earth, life, and our species. And statistics excites me because inference is the essence of the scientific method, which is the quest to obtain consistent observations and models.
Planets beyond our Solar System are intriguing objects that place our planet Earth in the cosmological context. They allow us to study the formation and evolution of planets and their atmospheres. The Transiting Exoplanet Survey Satellite (TESS) is a spaceborne NASA telescope to survey the sky for transiting exoplanets. As a vetting lead of the TESS mission, I have worked on the delivery of more than 2100 exoplanet candidates and the detection and characterization of hundreds of exoplanets using the TESS data.
Exoplanets have a surprisingly broad range of radii, masses, equilibrium temperatures, and types of host stars. Some exoplanets are detected by the periodic dimming of the host star as they transit our line-of-sight, allowing the inference of the planet radius. Another class of detections are based on the periodic shift of the spectral features of the host star induced by the orbital reflex motion, where the planet's mass can be bounded from below. The former requires a geometrically rare alignment, whereas the latter requires the host star to be bright enough for high-resolution spectroscopy.
Hunting for Exoplanets
Transiting planets hosted by bright stars are especially opportune, as they become amenable to precise radius and mass measurements, yielding insight into the bulk composition of the planet. Furthermore, they also potentially allow atmospheric characterization with broad implications for probing atmospheric escape as well as biosignatures. An even more interesting situation is when multiple such planets transit the same bright star, as the resulting multiplanetary system enables comparative studies of the formation, dynamics, and atmospheric evolution of the planets.
In this context, a significant highlight from the TESS mission is the discovery of four small, transiting exoplanets hosted by the bright, Sun-like star HD 108236, also known as TOI 1233 (Daylan et al. 2021a). The system contains a hot, likely rocky super-Earth and three outer sub-Neptunes with gaseous envelopes.
An interesting class of exoplanets are hot Jupiters, which are much hotter analogs of our Jupiter that are inflated due to the large irradiation from their host star. The very existence of hot Jupiters is intriguing since they likely migrated to their observed orbits long after their formation farther away from their host stars. Their atmospheres are also of interest since they provide an observationally accessible laboratory for learning atmospheric dynamics on exoplanets. Towards this purpose, we analyzed of the TESS phase curve of WASP-121b, an ultra-hot Jupiter, where we measured the nightside temperature of the planet and found its heat redistribution to be very inefficient (Daylan et al. 2021b).
Below is a colloquium I gave remotely in the Department of Physics at Washington University in St. Louis, which provides a broad overview of how exoplanet research relates to our evolving understanding of planet formation, evolution, and migration.
First inferred to explain the mass of the Coma cluster by Fritz Zwicky in 1933, dark matter is a non-luminous matter component of our Universe whose existence is inferred from its gravitational interactions and its effect on the structure of the Universe. It is gas-like and transparent and does not interact with light.
Given the absence of consistent evidence of its particle nature, dark matter continues to elude us despite eight decades of research. There are mainly three ways of searching for dark matter: direct detection, production, and indirect detection. Direct detection relies on searching for rare particle interactions in a low-background particle detector placed underground. Second, high-energy particle colliders search for collision products that invisibly escape the detector. Finally, indirect detection consists of searching for astrophysical signals such as cosmic-ray and high-energy photon fluxes that can be explained by dark matter, but not by the Standard Model of particle physics.
The gamma-ray excess in the inner Milky Way
An enigmatic feature of the gamma-ray sky as surveyed by NASA's Fermi-LAT telescope, is the anomalous emission originating from the center of our galaxy in excess of expectations based on its estimated content of cosmic rays, gas, and dust. The excess can be due to a population of millisecond pulsars in the galactic bulge, although this population would have to be made of a surprisingly large number of and dim millisecond pulsars.
An intriguing feature of this emission is that it is also consistent with the simplest models of dark matter. In a highly cited work (Daylan et al. 2016), we characterized this excess emission and showed that the spatial morphology, amplitude, and spectrum of the excess are consistent with dark matter, where weakly interacting massive particles annihilate with a thermal cross-section to high energy photons that constitute the excess.
The Alpha Magnetic Spectrometer 2 (AMS-02) is a particle detector on the International Space Station (ISS), which measures the cosmic-ray flux from outer space. It has a magnet and a silicon tracker to bend and measure the momentum of charged particles, an electromagnetic calorimeter to measure the particle energy and other auxiliary subdetectors to identify incident particles. AMS-02 measures interesting observables such as positrons and anti-helium, whose production can potentially be a signature of dark matter annihilation. I was an undergraduate research assistant at CERN between 2011 and 2013 and worked on implementing an improved event reconstruction algorithm for the detector.
Transdimensional Bayesian inference
There are many problems in science, where the number of unknowns is unknown. Therefore, a robust inference framework that allows marginalization over the unknown degrees of freedom is desirable. The Probabilistic Cataloger (PCAT; Daylan et al. 2017) is a transdimensional and Bayesian sampler that allows fairs samples to be taken from the posterior probability distribution of a metamodel, i.e., a super-model that contains models with different numbers of parameters. This allows transdimensional uncertainties to be robustly propagated to the marginal posterior probability distribution of interest.
Probabilistic cataloging has been successfully used to model an apriori-unknown number of dark matter subhalos in simulated strong-lenses (Daylan et al. 2018) and points sources in a crowded image (Portillo et al. 2018).
Below is an illustration of probabilistic cataloging. The plot on the left emphasizes the multimodality of the catalog space. The image on the right shows fair samples drawn from the posterior of the catalog space consistent, up to Poisson likelihood, with the gamma ray sky towards the North Galactic Polar Cap, i.e., if you look towards the zenith taking the galactic plane as datum.