Not yet reviewed by Pith; the record is open.

The use of supervised methods in space science have demonstrated powerful capability in classification tasks, but unsupervised methods have been less utilized for the clustering of spacecraft observations. We use a combination of unsupervised methods, being principal component analysis, self-organizing maps, and hierarchical agglomerative clustering, to make predictions on if THEMIS and MMS observations occurred in the magnetosphere, magnetosheath, or the solar wind. The resulting predictions are validated visually by analyzing the distribution of predictions and studying individual time series. Particular nodes in the self organizing map are studied to see what data they represent. The capability of deeper hierarchical analysis using this model is briefly explored. Finally, the changes in region prediction can be used to infer magnetopause and bow shock crossings, which can act as an additional method of validation, and are saved for their utility in solar wind validation, understanding magnetopause processes, and the potential to develop a bow shock model.