Speaker
Description
Recent spectral observatories stand to revolutionise our ability to study exoplanets on a larger population scale than ever before. New space-based instruments such as Ariel, alongside existing ground-based spectrographs at the ELT are set to generate vast quantities of atmospheric spectral data over the next ten years. Analysing this data requires extracting information about the planetary atmosphere from the spectra. For anything outside a small number of targets this is very computationally resource intensive, which is a large barrier to entry as we move into these larger scale planetary surveys.
The use of ML has been proven as a powerful tool in tackling this, reducing computational resources required. However, the scale of these models means this advantage comes at the cost of understanding.
We go beyond existing approaches, presenting a novel method of interpretability based on physically motivated forward modelling, bridging the gap between ML and traditional approaches.
We trained a convolutional network architecture to predict the atmospheric abundances of 5 molecules across 40,000 simulated Ariel spectra, then compare a selection of existing techniques for interpreting predictions. Based on this analysis we propose a novel application of the perturbation sensitivity technique for interpreting ML predictions.
This method has potential for use outside of Ariel data, and we believe the opportunity to share it here would help unlock barriers to entry in the use of ML for planetary spectral analysis across ground and space based observations.
