EAPS research scientist Bryan Riel is a geophysicist and engineer working with the Minchew group. He's interested in utilizing large-scale geodetic data to extract meaningful insights into the behaviors and properties of physical systems. Generally, the spatiotemporal signals of interest encoded in these data are confounded by noise, external forcing effects, data corruption, etc. Bryan’s research is focused on combining the power of modern machine learning algorithms and physical modeling to disentangle the primary factors of variation in geodetic data and associate those factors with physically-relevant processes. This form of theory-guided learning can potentially improve the generalizability of predictions and forecasts while providing a framework to probe the physical properties of complex systems using data. Bryan is currently developing and applying these tools to velocity time series for outlet glaciers in Greenland and Antarctica in order to study basal mechanics and ice-ocean interactions.