Combining high-contrast imaging with high-resolution spectroscopy: Actual on-sky MIRI/MRS results compared to expectations

Context. Combining high-contrast imaging with high-resolution spectroscopy represents a powerful approach to detecting and characterizing exoplanets around nearby stars, despite the challenges posed by their faintness. Instruments like VLT/SPHERE represent the state of the art in high-contrast imaging; however, their spectral resolution (R ≈ 50) limits them to basic characterization of close companions. These instruments can observe planets with masses as low as 5–10 MJup at distances of around 10 AU from their stars. Detection limits are primarily constrained by speckle noise, which dominates over photon and detector noise at short separations around bright stars, even when advanced differential imaging techniques are used. Similarly, image stability also limits space-based high-contrast imaging capability. This speckle noise can, however, be largely mitigated by molecular mapping, a more recent method that leverages information from high-resolution spectroscopic data. Aims. Our objective is to understand and predict the effective detection limits associated with spectro-imaging data after processing with molecular mapping. This involves analyzing the propagation of fundamental noise sources, such as photon and detector noise, and comparing these predictions to real instrument data to assess performance losses due to instrument-based factors. Our goal is to identify and propose potential mitigation strategies for these additional sources of noise. Another key aim is to compare the predictions made by our analytical approach with actual observational data to validate and refine the model’s accuracy where necessary. Methods. We analyzed JWST/MIRI/MRS data using the recently developed semi-analytical and numerical tool, FastCurves, and compared the results with outputs from the end-to-end MIRI simulator. This simulator allows one to examine nonideal instrumental effects in detail. Additionally, we applied principal component analysis (PCA), a statistical method that identifies correlated patterns in the data, to help isolate systematic effects, both with and without molecular mapping. Results. Our analysis involves investigating the systematic effects introduced by the instrument, identifying their origins, and evaluating their impact on both noise and signal. We show that valuable insights are gained regarding the effects of straylight, fringes, and aliasing artifacts, each linked to different residual systematic noise terms in the data. The results are further supported by principal component analysis, which also demonstrates its effectiveness in mitigating these effects. Additionally, we explore the similarities and discrepancies between observed and modeled companion spectra from an astronomical perspective. Conclusions. We modified FastCurves to account for systematic effects and improve its modeling of MIRI/MRS noise, with its signal-to-noise predictions validated against empirical data. In high-stellar-flux regimes, systematic noise imposes an ultimate contrast limit when using molecular mapping alone. Our methodology, demonstrated with MIRI/MRS data, could greatly benefit other instruments, aiding in the planning of observational programs. For future instruments like ELT/ANDES and ELT/PCS, it could also inform and guide their development.