Multi-spectral remote sensing has already played an important role in mapping surface mineralogy. However, vegetation - even when relatively sparse - either covers the underlying substrate or modifies its spectral response, making it difficult to resolve diagnostic mineral spectral features. Here we take advantage of the petabyte-scale Landsat datasets covering the same areas for periods exceeding 30 years combined with a novel high-dimensional statistical technique to extract a noise-reduced, cloud-free, and robust estimate of the spectral response of the barest state (i.e. least vegetated) across the whole continent of Australia at 25 m2 resolution. Importantly, our method preserves the spectral relationships between different wavelengths of the spectra. This means that our freely available continental-scale product can be combined with machine learning for enhanced geological mapping, mineral exploration, digital soil mapping, and establishing environmental baselines for understanding and responding to food security, climate change, environmental degradation, water scarcity, and threatened biodiversity.