Mobile autonomous process sampling within coastal ocean observing systems

Abstract

Predicting when and where key oceanic processes will be encountered is problematic in dynamic coastal waters where diverse physical, chemical, and biological factors interact in varied and rapidly changing combinations. Defining key processes often requires efficient sampling of specific water masses and prompt sample return for subsequent analyses. This compound challenge motivated our efforts to develop mobile autonomous process sampling (MAPS) for use with autonomous underwater vehicles (AUVs). With this system, features are recognized by artificial intelligence that integrates AUV sensor data to estimate probabilistic states for adaptive control of survey navigation and triggering of targeted water samplers. To demonstrate the utility of the MAPS/AUV system, we focused on intermediate nepheloid layers (INLs), episodic transport events that may play a role in zooplankton ecology. During multiple field tests in Monterey Bay, California, the MAPS/AUV system recognized, mapped, and sampled INLs. Invertebrate larvae contained in the water samples were subsequently characterized with molecular probes developed for high‐throughput screening. Preliminary results support the hypothesis that INLs function as vehicles for episodic larval transport. Applying MAPS within a greater coastal ocean observing system permitted description of regional oceanographic dynamics that influenced the patterns and scales of INL and larval transport.