Development of an Autonomous Sampling Network for Plankton, Hydrography and Currents. Phase I. Incorporation of Plankton Imaging Capability into Autonomous Underwater Vehicles

Woods Hole Oceanographic Institution, Woods Hole MA

Investigators

Abstract

0000580 Davis This research project is conducted under the auspices of the National Oceanographic Partnership Program (NOPP). It addresses a major challenge in biological oceanography, which is the development of predictive capability. Pressing issues such as fishery yields, pollutant transfer, carbon cycling, and sound and light transmission in the ocean all require accurate predictions. The vast majority of marine species are planktonic during at least a portion of their lives and therefore have populations that grow and die while being transported by ocean currents. Episodic factors such as storms, water mass interactions, and predator-prey patchiness are a major source of variability in marine populations. At present, our predictive capability is limited by the sparseness of data and the inability to sample with adequate time and space resolution to capture episodic features that control population size. Recent technological advancements in automated plankton identification (via optical imaging) and in design of autonomous underwater vehicles (AUVs) make it plausible to sample plankton in association with other environmental variables (e.g. temperature, salinity, light, nutrients, fluorescence, currents) autonomously with high resolution in both time and space. Ultimately, networks of AUVs equipped with plankton imaging systems deployed in specific regions of the world ocean can be envisioned to provide quasi-synoptic high-resolution data in time and space. This project will merge plankton-imaging technology with AUV technology by incorporating the Video Plankton Recorder (VPR) into the REMUS AUV. The VPR is an underwater video microscope that images plankton and other particulate matter in the size range of 100 microns to 2 centimeters. The REMUS AUV is designed to carry a payload of scientific instrumentation on pre-programmed missions using a combination of long and ultra-short baseline navigation. With up to 6 hours of endurance using rechargeable lead acid batteries or 18 hours from primary lithium cells, REMUS is well suited for repetitive space/time data sampling. The VPR design will be made modular such that generic use in other AUVs will be possible. This initial design will allow for on-board real time compression and storage of digitized video using wavelet compression hardware. The compressed video then will be uploaded for analysis with an automated image processing system. The new VPR-REMUS system will be tested in a variety of habitats using Eulerian and Lagrangian sampling methods to automatically quantify distribution patterns of plankton taxa, fluorescence, hydrography, and currents.