Aishwarya Iyer

Stellar Atmospheres, Magnetic Activity, and Their Impact on Exoplanet Characterization

My research focuses on understanding the physical complexity of low-mass stars—particularly M dwarfs—and quantifying how their atmospheres and magnetic activity impact the interpretation of exoplanet observations. These stars are prime targets for exoplanet discovery and characterization, yet their heterogeneous surfaces, strong magnetic fields, and poorly understood atmospheric physics introduce fundamental uncertainties in derived planetary properties.

To address this, I develop physically motivated stellar atmosphere models and retrieval frameworks that bridge theory, observation, and statistical inference. I am the lead developer of the SPHINX model grid, a self-consistent radiative–convective–thermochemical framework designed to capture key processes in M-dwarf atmospheres, including cloud formation, convection efficiency, and chemical composition. My work extends these models to explore degeneracies between stellar parameters—such as effective temperature, metallicity, C/O ratio, and surface heterogeneity—and their imprints on observed spectra across optical and infrared wavelengths.

A central component of my research is disentangling stellar contamination in exoplanet transmission and emission spectra. I use a combination of grid-based retrievals, Bayesian inference (e.g., MCMC and nested sampling), Gaussian processes for correlated noise, and dimensionality reduction techniques (e.g., PCA) to quantify how stellar variability and surface inhomogeneities bias exoplanet atmospheric measurements. This work is particularly relevant for interpreting data from missions such as JWST and Pandora, where stellar effects are often the limiting factor in detecting and characterizing terrestrial exoplanets.

In parallel, I investigate stellar magnetic activity and its role in shaping both stellar atmospheres and exoplanet environments. Using high-resolution near-infrared spectroscopy (e.g., Keck/NIRSPEC, IRTF/iSHELL), I study Zeeman-sensitive lines (e.g., FeH, Ti I, Na I) to constrain magnetic field strengths and topologies. I am also developing forward models that connect magnetic field structure (dipolar vs. multipolar) to observable diagnostics such as line broadening, Stokes profiles, and wind-sensitive tracers (e.g., He I 1083 nm), with the goal of understanding how stellar magnetism drives stellar winds, surface heterogeneity, and atmospheric escape in orbiting planets.

More broadly, my work sits at the intersection of physics-based modeling and data-driven inference, with an emphasis on building scalable, interpretable frameworks that can be applied to large observational datasets. I am particularly interested in advancing next-generation approaches that combine high-resolution spectroscopy, time-domain observations, and machine learning to enable “star-aware” exoplanet characterization—where stellar uncertainties are explicitly modeled rather than treated as noise.

Looking ahead, my research aims to (1) improve the physical realism of stellar atmosphere models, including cloud microphysics, convection, and magnetic effects; (2) develop robust retrieval techniques that quantify model inadequacies and observational biases; and (3) integrate multi-wavelength and multi-epoch datasets to break degeneracies between stellar and planetary signals. Ultimately, this work will enable more precise and reliable characterization of terrestrial exoplanets and their potential habitability.