Understanding molecular ratios in the carbon- and oxygen-poor outer Milky Way with interpretable machine learning

Context. The outer Milky Way has a lower metallicity than our solar neighbourhood, but many molecules are still detected in the region. Molecular line ratios can serve as probes to understand the chemistry and physics in these regions better. Aims. We used interpretable machine learning to study nine different molecular ratios to help us understand the forward connection between the physics of these environments and the carbon and oxygen chemistries. Methods. Using a large grid of astrochemical models generated using UCLCHEM, we studied the properties of molecular clouds with a low initial oxygen and carbon abundance. We first tried to understand the line ratios using a classical analysis. We then proceeded to use interpretable machine learning, namely Shapley additive explanations (SHAP), to understand the higher-order dependences of the ratios over the entire parameter grid. Lastly, we used the uniform manifold approximation and projection technique (UMAP) as a reduction method to create intuitive groupings of models. Results. We find that the parameter space is well covered by the line ratios, which allowed us to investigate all input parameters. The SHAP analysis showed that the temperature and density are the most important features, but the carbon and oxygen abundances are important in parts of the parameter space. Lastly, we find that we can group different types of ratios using UMAP. Conclusions. We show that the chosen ratios are mostly sensitive to changes in the initial carbon abundance, together with the temperature and density. Especially the CN/HCN and HNC/HCN ratios are shown to be sensitive to the initial carbon abundance. This makes them excellent probes for this parameter. Only CS/SO is sensitive to the oxygen abundance.