Data Processing and Algorithm Design for Observations of Star-forming Galaxies Using DESHIMA 2.0

Example research questions:

• How can we separate weak astronomical signals from noise and systematics, such as the detector and atmosphere?

• What can the emission lines and continuum emission of distant galaxies tell us about their properties?

Background

DESHIMA 2.0 is an integrated superconducting spectrometer designed for high-redshift astronomy. This means we can use it to look back in time and study the early Universe. It operates in the 200-400 GHz band, giving it the widest bandwidth in the world for any (sub)millimeter astronomical spectrometer. The astronomical signals of interest, primarily dusty star-forming galaxies, are extremely weak after their long journeys through the Universe. When they arrive at the Earth, their signals are further attenuated as they propagate through the atmosphere, and the atmosphere itself emits radiation that is orders of magnitude stronger than the target sources. Additional error contributions arise from photon noise, noise sources in the detector, and systematic effects within the telescope, such as optics misalignment. These error sources exhibit different temporal correlations (e.g., atmosphere and detector noise are time-correlated, whereas photon noise is not) and different spectral behaviours (the atmosphere has a distinct spectrum, whereas the detector noise is spectrally uncorrelated). Robust data-processing algorithms exploit these correlations to isolate the astronomical signal and uncover information about the early Universe.

Your project

After a successful observing campaign at the end of 2024, we have obtained a wealth of astronomical data. You will work with real DESHIMA data and simulations, investigate the statistical properties of different noise sources, and design or improve algorithms that make optimal use of their spectral and temporal signatures. DESHIMA 2.0 is an instrument developed with the help of many students, so we are eager to incorporate your interest. You are free to steer this project into a direction that interests you: be it the development of a dedicated data-processing pipeline capable of separating atmospheric fluctuations from the astronomical signals, a deeper analysis of the various noise sources present in the data, or a focus on the astrophysical interpretation of DESHIMA observations.

Contact

Last modified: 2025-11-28