Towards the Habitable Worlds Observatory: Retrieval of Reflection Spectra from Evolving Earth Analogues using 1D CNNs
A one-dimensional Convolutional Neural Network (1D CNN) was trained on over a million synthetic planetary spectra to rapidly retrieve key atmospheric parameters like molecular abundances from low-resolution reflection data. The model shows that recovery depends on spectral structure and abundance, identifying specific wavelength regions crucial for detecting gases like O2 and CH4 across different Earth analogs.