A Panchromatic JWST Spectrum of a Giant Starspot on the Fully Convective M-dwarf TOI-3884

Abstract

TOI-3884 b is a rare super-Neptune transiting a fully convective M dwarf that hosts a persistent giant polar spot. Because the planet occults this active region during every transit, the system offers a unique laboratory to directly probe the stellar surface and spot properties. We present seven James Webb Space Telescope (JWST) transits of TOI-3884 b observed with NIRISS and NIRSpec, spanning 0.5--5.3$μ$m. While all visits show a recurring spot-crossing signature, each transit exhibits a distinct spot-crossing morphology, enabling us to infer a stellar rotation period of $P$=11.102$\pm$0.003d and tightly constrain the pole-on stellar orientation ($i_{*}$=40.8$\pm$0.3$^{\circ}$, $λ_{*}$=148.9$\pm$0.4$^{\circ}$) and spot properties ($R_{\rm{spot}}=0.576^{+0.006}_{-0.005}$R$_{*}$, $φ_{\rm{spot}}$=84.69$\pm$0.12$^{\circ}$). We leverage this orbital configuration to measure the first empirical panchromatic spectrum of an M-dwarf starspot with JWST, establishing a direct observational benchmark for stellar atmosphere models in the fully convective regime. Comparison with 1D NewEra and SPHINX atmosphere models indicates that the spot is 185$\pm$2K cooler than the photosphere, consistent with previous ground-based measurements and expectations for mid-M-dwarf spot contrasts. While the models reproduce the observed contrasts at wavelengths longer than 1$μ$m, they significantly underpredict the contrasts at shorter wavelengths. These results demonstrate that M-dwarf stellar atmosphere models alone may not fully capture the wavelength dependence of stellar contamination in transmission spectra and highlight the importance of empirical spot spectra for robust interpretation of planetary atmospheres, particularly in the optical.

Figures (7)

• Figure 1: Spectroscopic time-series of the transit light curves of TOI-3884 b for all seven JWST visits, reduced with ExoTEDRF (Section \ref{['ss:datareduction']}). NIRISS/SOSS observations are presented in the top row, and NIRSpec/BOTS in the bottom. Each panel shows normalized flux as a function of orbital phase and wavelength, and the color scale is shared across all panels. In all visits, the planetary transit appears as a broad vertical band centered at phase zero, and the spot-crossing event is apparent in the first half of the transit. Several visits also exhibit continuum and line emission attributable to stellar flares. The horizontal black lines (0.85 $\mu$m) on the top panels indicate approximately where NIRISS orders 1 and 2 overlap (0.85--1.10 $\mu$m). The white horizontal bar in Visit 5 indicates the gap between the NRS1 and NRS2 detectors for the G395H mode (3.72--3.82 $\mu$m); in G395M mode only NRS1 is illuminated.
• Figure 2: White-light transit light curves of TOI-3884 b for all seven JWST visits using the ExoTEDRF reduction described in Section \ref{['ss:datareduction']}. NIRISS/SOSS observations are presented in the top row, and NIRSpec/BOTS in the bottom. For each visit, the upper panel shows the normalized flux as a function of transit phase as well as the best-fitting white-light models for each visit independently (Section \ref{['sec:indiv']}, dashed lines), and for the six fits jointly (Section \ref{['sec:joint']}, solid lines). We plot how the respective individual and joint spot fits for each visit would look on TOI-3884 in each panel. The lower plots show the corresponding flux residuals for both models. All visits exhibit a pronounced spot-crossing feature during the first half of the transit, though its shape clearly varies between visits. Stellar flares are evident in most light curves to some degree. Visits 1 and 3, in particular, show strong stellar flares that overlap the transit window. The NIRSpec observations, which probe longer wavelengths, show comparatively quiescent behavior, though flares are still present in all NIRSpec datasets. All flares that are masked during light curve fitting are highlighted in black (with corresponding gray shaded regions shown in the residuals).
• Figure 3: The best-fit TOI-3884 spot centers for the individual white light curve fits (solid circles with contours indicating the 1$\sigma$ and 2$\sigma$ regions from the posteriors) and the joint white light curve fit (stars). We plot ellipses indicating the extent of each spot from the self-consistent joint model. We find the best-fit solutions for Visits 5 and 6 have a spot above the transit chord (shaded grey), likely due to the geometric symmetry about the transit chord and the known radius--latitude degeneracy. Therefore, for those two visits we also include the lines of projected "longitude" along which we might expect to find degenerate spots (dotted colored lines). The black cross marks the stellar rotation pole ($i_{*}$=40.8$\degree$, $\lambda_{*}$=148.9$\degree$) derived in the self-consistent joint model and the dashed black line indicates how the center of the spot rotates around the pole. The stellar rotation phase, $\Psi$, for each visit is shown in the legend (assuming $\Psi$=0 at the first visit). This plot is inspired by Figure 3 from chakraborty_changing_2025.
• Figure 4: Best-fit TOI-3884 spot radii (top), instrument-dependent contrasts (upper middle) and planet-to-star radius ratios (lower middle) from each individual white light curve fit. The best-fit spot radius, contrasts, and radius ratios from the self-consistent joint model are indicated with horizontal lines. Visits 1 and 3 are significantly impacted by stellar flares, therefore are heavily masked, inflating the uncertainties of the derived parameters for these visits. The best-fit spot temperature for Visits 2, 4, 5, 6, and 7 are shown on the bottom plot.
• Figure 5: Top: The empirical spot contrast spectra for JWST Visits 2, 4, 5, 6, and 7. The dashed black line shows the best-fit NewEra model ($T_{\rm{base}}=$$3187 \pm 15$ K, $T_{\rm{spot}}=$$3002 \pm 15$ K) and dotted line shows the best-fit SPHINX model ($T_{\rm{base}}=3143\pm33$ K, $T_{\rm{spot}}=2966\pm34$ K). We overplot broadband contrasts derived in previous works by almenara_toi-3884_2022libby-roberts_-depth_2023tamburo_spot-crossing_2025mori_multiband_2025sagynbayeva_polka-dotted_2025chakraborty_changing_2025 (compiled in Table \ref{['tab:contrast_lit']}). Bottom: Residuals between the broadband and spectroscopic contrasts and the best-fit NewEra model. A histogram of the residuals for each visit is shown to the right of the residual plot.