Any suggestions, links, or research on how TDA or compressed sensing can be used to make sense of spatiotemporal datasets (gridded climatology time series data, change/anomaly detection in time-lapsed satellite imagery etc.)?