Exoplanet Transit Spectroscopy with JWST NIRSpec: Diagnostics and Homogeneous Case Study of WASP-39 b

TL;DR

This study evaluates JWST NIRSpec modes for exoplanet spectroscopy using WASP-39 b data, developing a new reduction framework to compare spectral quality, stability, and noise characteristics across modes and detectors.

Contribution

The paper introduces JexoPipe, a new data reduction framework, and provides a homogeneous assessment of NIRSpec modes for exoplanet spectroscopy with JWST.

Findings

G395H spectrum has higher precision than Prism at the same resolution.

Detected an inter-detector offset of 40-50 ppm in G395H.

Identified correlated noise in the Prism white light curve.

Abstract

The JWST has ushered in a new era of exoplanet transit spectroscopy. Among the JWST instruments, the Near-Infrared Spectrograph (NIRSpec) has the most extensive set of configurations for exoplanet time series observations. The NIRSpec Prism and G395H grating represent two extremes in NIRSpec instrument modes, with the Prism spanning a wider spectral range (0.6-5.3 $\mu$m at lower resolution (R$\sim$100) compared to G395H (2.87-5.14 $\mu$m; R$\sim$2700). In this work, we develop a new data reduction framework, JexoPipe, to conduct a homogeneous assessment of the two NIRSpec modes for exoplanet spectroscopy. We use observations of the hot Saturn WASP-39 b obtained as part of the JWST Transiting Exoplanets ERS program to assess the spectral quality and stability between the two instrument modes at different epochs. We explore the noise sources, effect of saturation, and offsets in transmission spectra between the different instrument modes and also between the two G395H NRS detectors. We find an inter-detector offset in G395H of $\sim$ 40-50 ppm, consistent with recent studies. We find evidence for correlated noise in the Prism white light curve. We find the G395H spectrum to be of higher precision compared to the Prism at the same resolution. We also compare the JexoPipe spectra with those reported from other pipelines. Our work underscores the need for robust assessment of instrument performance and identification of optimal practices for JWST data reduction and analyses.