Published July 3, 2026 | Version v1

Optex: An optimal estimation retrieval code

  • 1. ROR icon Observatoire de Paris
  • 2. CNRS

Description

The interpretation of high-SNR spectroscopic observations from JWST and the upcoming ELT era requires the

development of methods to shift from static, grid-based modeling towards flexible, computationally efficient

retrieval frameworks. We present Optex, a novel atmospheric retrieval tool based on Optimal Estimation

(Rodgers, 2000). By integrating Earth-science methodologies with the Exo_k opacity library (Leconte, 2021),

Optex achieves high vertical resolution and superior computational speed compared to traditional sampling

methods (e.g., MCMC and nested sampling), providing a rigorous framework for uncertainty quantification

and mitigating prior-driven biases. In this talk, we demonstrate Optex’s unique capabilities to resolve complex,

non-standard atmospheric structures. We focus on the isolated T-dwarf W1935, for which JWST observations

revealed unexpected methane emission (Faherty et al., 2024), a signature of auroral heating in the upper

atmosphere. Using Optex, we robustly characterize the thermal inversions and non-equilibrium chemistry

required to reproduce this emission. To conclude, we also highlight Optex’s scalability to multidimensional

modeling, offering a high-performance pathway to move beyond static 1D snapshots and promising to map the

dynamic climates of substellar objects effectively.

Files

Poster_exoplanets6.pdf

Files (7.7 MB)

Name Size
md5:ab5ce43c419a503f4637254a3e028815
7.7 MB Preview Download