Morphological changes of the spectral trace during our observations, as determined from a principal-component analysis (PCA) by following a similar method to L.-P. Coulombe et al. (2023). We use the scikit-learn (F. Pedregosa et al. 2011) routine IncrementalPCA, applying it to the 3D data cube after normalization using scikit-learn’s StandardScaler. We find that two PCA terms are sufficient to describe the systematic trends, and our identified signals are similar to those presented in L.-P. Coulombe et al. (2023): a trend associated with changes to the y-position of the trace (upper panels) and a beat pattern associated with changes to the FWHM of the trace (lower panels). This latter effect has been attributed (e.g., C. Piaulet-Ghorayeb et al. 2024) to thermal oscillations in the telescope (M. W. McElwain et al. 2023).