Exoplanet atmospheric retrievals through different JWST/NIRISS data reduction pipelines

Since the James Webb Space Telescope (JWST) became available in 2022, its observations of exoplanet transit spectra have revolutionized the field of exoplanet science. Its observations have enabled significant new discoveries, in particular, contributing to the ongoing effort of characterizing a wide array of exoplanet atmospheres [1][2][3]. Distinct data reduction pipelines have been produced to process JWST observational spectra [4][5] – which for some cases have resulted in slightly distinct transmission spectra, which may consequentially lead to distinct interpretations when characterizing said planetary atmospheres.

The goal of this study is to have a grasp of the extent of how distinct data reduction pipelines affect the extraction of transmission spectra and the consequent characterization of exoplanet atmospheres. Here we present an analysis of a small array of hot Jupiter transit spectra by the JWST/NIRISS instrument [6], whose 2 major spectral orders cover the wavelength range 0.7 μm to 2.5 μm. The same dataset of transit observations for this small exoplanet population were processed through multiple pipelines [4][5].

Here we explore how distinct pipelines extract distinct exoplanet transmission spectra from the same observational dataset – across a small sample of 5 hot Jupiters. We then apply the exoplanet atmospheric retrieval code TauREx 3 [7] to the distinct transmission spectra extracted by the several distinct data reduction pipelines. This allows to retrieve the distinct set of parameters that characterize these planetary atmospheres – and to compare how the distinct pipelines affect the retrieved parameters that characterize the planetary atmospheres across this small sample of hot Jupiters. This may provide a useful cross-validation between distinct data reduction approaches for this JWST instrument – increasingly relevant as new dedicated missions to the study of exoplanet atmospheres through transit spectroscopy – such as the ESA Ariel mission [8] - are expected to come online on the coming years.

Figure 1: WASP-39b JWST/NIRISS transmission spectrum extracted by 6 distinct data reduction pipelines, including our group’s IRACLIS pipeline. Following the data release from Feinstein et al, 2022 [4].

References:

[1] – Tsai S. et al, 2023, Nature, 617, https://doi.org/10.1038/s41586-023-05902-2, [2] – Taylor. J., et al, 2023, MNRAS, 524,  https://doi.org/10.1093/mnras/stad1547; [3] – Wellbanks, L., et al, 2024, Nature, 630, https://doi.org/10.1038/s41586-024-07514-w; [4] – Feinstein, A., et al, 2022, Nature, 614, https://doi.org/10.1038/s41586-022-05674-1; [5] - Fournier-Tondreau, M., et al, 2025, MNRAS, 539, https://doi.org/10.1093/mnras/staf489; [6] – Doyon, R., et al, 2012, Proceedings V. 8442, Space Telescopes and Instrumentation 2012: Optical, Infrared, and Millimeter Wave; 84422R, https://doi.org/10.1117/12.926578; [7] – Al-Refaie, A., et al, 2021, ApJ, 917 37, DOI: 10.3847/1538-4357/ac0252; [8] – Tinetti, G., et al, 2018, Experimental Astronomy, 46, https://doi.org/10.1007/s10686-018-9598-x;