Conversations
New Webb starlight-removal method boosts sensitivity for faint exoplanet detection by 2 to 15 times
A research team has proposed a physics-based starlight subtraction method for processing coronagraph observations from the Webb telescope’s NIRCam. Tests show that its sensitivity to faint exoplanets is 2 to 15 times higher than standard techniques, and it detected the giant planet 51 Eridani b at the 4.3-micron wavelength band for the first time.
Direct imaging of exoplanets first requires suppressing the bright light of the host star and then distinguishing the faint light emitted by the planet. Existing commonly used algorithms rely on reference images to build a model of starlight, but instabilities such as telescope pointing errors and thermal drift can create differences between reference observations and science observations, thereby limiting detection capability.
The new method combines multi-step optimization with a flexible physical model to model and compensate for instrument instability, and uses automatic differentiation to optimize millions of model parameters. The team validated the framework on coronagraph data from the Webb telescope’s Near-Infrared Camera and says its sensitivity is 2 to 15 times higher than standard techniques.
The tests also detected the young giant planet 51 Eridani b for the first time at a wavelength of 4.3 microns. The paper’s authors say this is, to their knowledge, the first time such a method has demonstrated a significant sensitivity improvement on real space-based coronagraph images. The framework could potentially be extended in the future to facilities such as the Roman Space Telescope and the Habitable Worlds Observatory, reducing requirements for extreme stability.