The industry's recognized authority for design, engineering and application of equipment and services in the global ocean community
Issue link: http://sea-technology.epubxp.com/i/82362
(Top) Water sampling ports in the Dorado AUV's hull. (Bottom) The long-range AUV Tethys. particle backscatter similar to INLs but also return high values for chlorophyll fluo- rescence. Conversely, sed- iment-derived backscatter from INLs coincides with low chlorophyll signals. Molecular analysis of INL water samples from mis- sions conducted in Janu- hicle through water and to overcome avoidance behavior exhibited by some target organisms, such as copepods, which feed on phytoplankton. Feature Detection and Sampling Originally, the Dorado was tasked with collecting water samples at pre- programmed geographic locations, or waypoints. While this method was useful for interpreting relationships between water sample biology and associated environmental conditions, sampling was essentially random with regard to environmental conditions immediately surrounding the AUV. Comparative analysis of the collect- ed biological and associated environ- mental data enabled the definition of requirements for subsequent software development, resulting in algorithms that allow the vehicle to interpret en- vironmental data independently in real time and use that information to make decisions about where and when to collect water samples. In a similar fashion, the AUV is also now capable of identifying features of biological interest, and traversing and sampling them with precision. Intermediate Nepheloid Layers Intermediate nepheloid layers (INLs) are episodic sediment transport events mediated by bottom boundary layer dynamics. They are thought to play a role in benthic invertebrate lar- val transport. Multiple algorithms have been successfully developed and ap- plied to identify and sample INLs with the Dorado. Information from the AUV's sensor suite is used to differentiate INL signa- tures from other signals present in the surrounding water column in order to sample them selectively. For example, aggregations of phytoplankton produce 52 st / SEPTEMBER 2012 www.sea-technology.com ary and November 2008 in Monterey Bay, California, demonstrated the pres- ence of invertebrate larvae (i.e., poly- chaete worms, barnacles and mussels) in these features. Such ecological data can help inform studies of population connectivity, relevant to planning and managing marine protected areas. Thin Phytoplankton Layers In addition to forming large-scale blooms in response to processes such as wind-forced upwelling of nutrients, phytoplankton can aggregate into lay- ers ranging in thickness from fractions of a meter to several meters. Precisely sampling these layers, which are de- tectable by high-chlorophyll signal, is now possible with the Dorado, thanks to the development of an algorithm that can adaptively identify and cap- ture chlorophyll peaks. Thin phytoplankton-rich layers are difficult to sample with a moving AUV because a delay in chlorophyll peak detection (unavoidable by any real-time peak detection algorithm) of even a few seconds will result in water-sample collection occurring past the physical chlorophyll peak target. To solve this problem, an AUV peak- capture algorithm learns from environ- mental data in real time. Within each vertical profile, the vehicle registers the maximum chlorophyll signal on its first pass through a thin layer. On its second pass, the AUV triggers a Gulper as soon as the measured chlorophyll reaches the chlorophyll peak signal re- corded on the first pass, thus accurate- ly acquiring a peak-chlorophyll water sample without delay. This approach has enabled high spatial- and temporal-resolution stud- ies of phytoplankton bloom dynamics and ecology through the consistent sampling of chlorophyll peak maxima.