Reanalysis of the eclipses of LHS 1140 c: No evidence of an atmosphere and implications for the internal structure of the planet

TL;DR

This study reassesses three 15 μm JWST/MIRI secondary eclipses of the warm super-Earth LHS 1140 c using two independent photometric pipelines (Eureka! and MIRIAM) to robustly constrain the planet's dayside emission. By applying multiple detector-detrending models and performing both per-eclipse and joint fits, the authors obtain consistent eclipse depths of about 262–271 ppm and brightness temperatures around 587–595 K, supporting a bare-rock scenario with little to no atmospheric heat redistribution. The timing offsets indicate a nearly circular orbit, and interior-structure analyses reveal a CMF near zero when constrained by mass and radius, yet stellar refractory abundances imply a higher CMF (~0.34), suggesting interior volatiles such as water may be present. The results have implications for the interior composition and for planning future JWST observations, particularly of LHS 1140 b, to probe atmospheric reten­tion and interior structure in this nearby M-dwarf system.

Abstract

We present the reanalysis of three 15 micron JWST/MIRI secondary eclipses of LHS 1140 c, a warm super-Earth (R$_{\rm{p}}$ = 1.272 R$_{\oplus}$) in a 3.78-day orbit around an M4.5 dwarf. We present a novel method for data reduction that leverages spatial derivatives of the point-spread function and compare it to widely used aperture photometry. Both methods yield eclipse depth consistent within 1 sigma of the values reported in the literature. We measure an eclipse depth of 271$^{+31}_{-30}$ ppm corresponding to a brightness temperature of $T_B=595^{+33}_{-34}$ K, consistent with a bare rock. The secondary eclipse occurs 4.1$\pm$0.8 minutes before the circular-orbit predicted time. We explore the implications of our results on the internal structure of LHS 1140 c, the orbital architecture of the system and the possibility of future observations with JWST. We find a core-mass fraction (CMF) informed by the stellar abundances of refractory elements of 0.34$\pm$0.11, inflated compared to the CMF from radius and mass measurements, suggesting the possible presence of bulk volatiles in the interior.

Figures (7)

• Figure 1: Comparison of the best-fit models for the three eclipses (UT2023-11-27, left; UT2024-07-07, middle; UT2024-07-19, right) with the two light curve extraction methods (top and bottom). The raw photometry is plotted in light colour in the top panels with error. The error is calculated at the extraction step. Three-minute bins are shown in darker colour for clarity. The vertical black dashed lines indicate the expected time of mid-eclipse with the shaded grey region showing its duration. The corrected photometry with the best-fit detrending model for each eclipse is shown in the middle panels. The error is fitted with the MCMC. The dark, solid line corresponds to the detrending model and the dashed line to the jointly fitted astrophysical signal. The shaded light vertical band is the derived time-of-mid eclipse $t_s$ with uncertainty. The last panels shows residuals when the astrophysical model is subtracted from the corrected flux. Three-minute bins are overlain in darker color for clarity. The best fit systematics models are E, DE, L, for the three eclipses respectively, and are conserved for both extraction methods.
• Figure 2: Left: Atmospheric models compared to the eclipse depth derived in this work and the values from the literature. The solid lines correspond to bare rock surface models and the dashed lines to atmospheric models. Right: The eclipse depth with errors for each eclipse derived in this work and from the literature. The data points are slightly offset horizontally for clarity. This figure was adapted from Figure 9 in connors2025.
• Figure A.1: Point-spread-function and the associated terms in the MIRIAM framework for Eclipse 1. The data products are virtually identical for other eclipses. Color scaling is linear with flux.
• Figure A.2: Time series of Eclipse 1 for detrending parameters from MIRIAM. Morphological changes in the PSF structure at the start of the sequence are readily seen in $c_\perp$, and to a lesser extent, in position.
• Figure B.1: Model comparison for the individual fits of each eclipse. The eclipse depth $F_p$ is expressed in ppm. The time of mid-eclipse, $t_s$, is expressed in minutes and negative values signify that the eclipse is occurring before the predicted time. The opacity of the box indicates the fit preference (darker is better). The preferred model is boxed in dark blue for each eclipse. The detector models as a function of time are shown as the columns, including the multiplied models. The rows indicate whether or not a detector model as a function of the centroid positions, $x_0$ and $y_0$ was used.