Award: OCE-0936143

Carbon sedimentation (10 Pg C/year) via the ocean bio-carbon pump (Fig 1) is important to the regulation of atmospheric CO2. Due to limitations of current observational methodology (moorings/ships), carbon export (or sedimentation) is poorly observed in space and time and thus is poorly understood and parameterized in carbon cycle simulations. Technology capable of autonomous operation in the ocean will enable improved carbon-cycle prediction. This project addressed a major gap in upper ocean carbon flux observations through the substantial completion of the development of the Carbon Flux Explorer (CFE), a fully autonomous robotic system designed to measure (and relay in real time via Iridium satellite link) the hourly/diurnal variations of particulate organic carbon (POC) and particulate inorganic carbon (PIC) sedimentation at various depths in the upper kilometer(s) of the ocean for seasons to year-long time scales. CFEs are the successful integration of the Sounding Oceanographic Lagrangian Observer (SOLO) float (developed at Scripps) and LBNL/UC BerkeleyÆs imaging Optical Sediment Recorder (OSR).The SOLO dives to its target depth and the optical sedimentation recorder (OSR) begins operation. During time at depth sinking particles settle into a high aspect ratio funnel before reaching the sample stage. The particles are imaged under Dark Field, Transmitted, and Transmitted – cross-polarized illumination. The OSR executes a hydrodynamic cleaning cycle and collects a reference image set, and sleeps. At timed intervals the OSR wakes up and repeats image sets. Cleaning occurs after a set number of image sets. After the specified number of hours at depth, the OSR takes another image set, does stage cleaning, and takes a final reference image set. The SOLO surfaces to report position, engineering data, CTD data, and OSR data; it then dives to its next planned depth. At project start in October 1 2009, we began with a 14 point list of CFE development needs which included: (1) addressing the failure of the CFE to sink on its first deployment in 2007 due to adsorbed air, (2) A tilt of 5° while under water, and (3) a pressure induced optical interference with cross polarized light imaging. The project assessed (4) bio-fouling of the sample imaging stage, (5) fouling by gelatinous organisms and tentacles during up-profile, (6) and migrator invasion. The project further (7) simplified the CFE complexity by using a symmetric layout (Fig 2) and elements of the SOLO pressure case design; (8) replaced the PTP protocol Nikon camera (Bishop, 2009) with a faster low power Sumix imager; (9) remade the down light (Fig. 2) and below-sample polarizer with higher transmissivity polarizers to enable shorter image exposure times; (10) modified the firmware controlling polarizer rotation to allow only a single crossed and uncrossed position to minimize pixel offsets of images and thus simplify image processing; (11) used lower cost and more simplified Li battery packs; (12) replaced the limited Linux Arcom computer with a Linux Gumstix cpu; (13) migrated many OSR functions to a low-power microcontroller; and (14) completed engineering documentation for most assemblies to facilitate technology transfer. Utilizing 7 research vessel expeditions, the project was able to sytematically test and prove new system elements and achieve a robust CFE design capable of operation in high sea states. The project improved CFE mission capability from ~7 days to 8 months (at hourly resolution) and two CFEs completed open ocean missions lasting 41 days (2011 – California Current) and 10 days (2013 – subarctic N Pacific); bio-fouling of optics and instrument has been minor. Software algorithms for off-line data reduction have been developed. Multiple expeditions from 2011 through 2013 to the 1900 m deep the Santa Cruz Basin, CA have yielded scientific dividends. An unprecidented CFE image data set (Fig 3) has enabled the comparison of satelli...