PhD Student (R1) per el Barcelona Supercomputing Center

The study of transient phenomena (e.g., supernova), so-called time-domain astronomy, is currently addressed by realizing full surveys of the sky (e.g., Pan-STARRS, SkyMapper, CRTS). The main challenge is to define the proper cadence, i.e., when and where to look in the sky to catch new events early and so observe as much of the phenomena as possible. The upcoming commissioning of the Vera C. Rubin Observatory will allow to cover the entire available sky over a few nights. While observing from the ground provides great levels of sensitivity, it may involve complications like weather at the observing site or a particular region of the sky, hence limiting the possibility of evenly sampling all regions of the accessible sky.

The PErsistent All-sky Survey from Space (PEASS) project will mitigate such issues considering small scale in orbit telescopes to complete a survey of the sky in the optical wavelengths at high cadence. The concept relies on the ability to deploy a small constellation of nano-satellites equipped with a high sensitivity COTS camera and a compact refracting telescope to perform wide field imaging of patches of the observable sky. Subpercent level photometric accuracy on each image using such apparatus have already been demonstrated for detecting exoplanets in project Panoptes. PEASS will further mitigate the limited pointing stability of a nanosatellite to enable long enough exposures and maximize sensitivity by integrating energy efficient processing units within the nanosatellite. Furthermore, to accomplish the main goal of searching for rare and weak signals at the detectability limit, the data produced at high cadence (typically one image every minute maximizing single exposure time) has to be processed on-board (overlapping several unit fields of view to minimize false alarms) and compressed to deliver science products that can be sent to the ground stations for further analysis and exploitation.

The PEASS project is a collaboration between the Barcelona Supercomputing Center (BSC), with wide expertise in computing systems, and the Observatoire de Paris (OdP), with wide expertise in the astronomy domain. The project is envisioned as a PhD thesis, and will count on the cooperation of advisors from BSC, OdP and the European Space Agency (ESA).

Key Duties

The PhD thesis will:

i. Select the data-analytics pipelines (e.g. instantaneous photometry -computation bounded- versus access to catalogs -IO intensive-), and so the payloads needed to address the selected science topics, which are (1) exoplanet identification (computation bound), (2) rare transient events such as supernova (IO bound), and (3) NEO detection (combination of both).

ii. Down-select the technology in terms of compute - I/O, storage, and energy budgets capacity according to nano-satellite limitations to be integrated in each analytics payload.

iii. Analyze and adapt/parallelize the data processing algorithms on the selected computing units.

iv. Implement the runtime techniques needed for the execution of flexible and/or reconfigurable pipelines on-board, depending on the targeted processor technology.

v. Evaluate the performance of the data processing algorithms to asses mission profiles and narrow science cases, in collaboration with the astronomy community.

Termini per aplicar: Diumenge, 16 Juliol, 2023

Més informació

Més entrades de Pràctiques