Presenter: Hajime KAWAHARA
Abstract:
While astronomers have confirmed 4,000 exoplanets so far, it is still difficult to directly compare exoplanets with solar planets because most of the transiting exoplanets discovered so far have an orbital period shorter than one year. Using graphic processing unit (GPU) computing and techniques in machine learning, we surveyed 200,000 stars observed by the Kepler spacecraft for signals of transiting planets whose orbital period is larger than two years. Most of these signals were overlooked because only one or two transits occurred in four-year light curves, and they were difficult to identify through standard periodic analysis of the detection pipelines. We identified dozens of long-period transiting exoplanets and finally published the catalog of these planets including Jupiter-like gas giants. Also, we found that Neptunian-sized planets around the snow line (at a few au) are common around FGK stars. It is difficult to explain this population using the current formation theory.
Known exoplanets near the snow line are located too far to measure their nature even for its
mass. Nearby targets are crucial for further study by future observatories. I will also talk about a conceptual plan of the nanosatellite constellation (LOTUS) to find the long-period planets in nearby stars.