Geophysical fluid dynamics (GFD) is the study of natural large-scale fluid flows, such as oceans, the atmosphere, and rivers. Recent years have seen the use of autonomous underwater and surface vehicles (AUVs and ASVs) to study various physical phenomena in geophysical fluid environments. These include mapping of ocean temperature and salinity profiles,understanding the distribution of plankton assemblages, and tracking harmful algal blooms. However, operating AUVs and ASVs in geophysical fluid environments poses significant challenges since GFD flows are naturally stochastic and aperiodic. Nevertheless, GFD flows exhibit dynamical coherent structures which are important for the estimation of the underlying geophysical fluid dynamics and, thus, the prediction of the various physical, chemical, and biological processes in them.
In this talk, I will present a strategy for distributed autonomous sensing and tracking of a class of coherent structures called Lagrangian coherent structures (LCS). LCS are important for quantifying transport phenomena and coincide with minimum energy and time trajectories in the ocean for underwater robots. Since these coherent structures delineate boundaries between regions with distinct flow dynamics, they also denote regions where more escape events occur. I will show how knowledge of these structures can be used to improve the design of ocean monitoring strategies for teams of mobile sensing resources. I will also show how new techniques developed for studying stochastic hybrid systems can be employed to analyze collectives of autonomous robots operating in time-varying and uncertain environments like the ocean. I will conclude with a brief
presentation of our current efforts in experimentally validating some of these strategies in a laboratory setting.