skip to main content


Search for: All records

Creators/Authors contains: "Roberts, Paul L. D."

Note: When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher. Some full text articles may not yet be available without a charge during the embargo (administrative interval).
What is a DOI Number?

Some links on this page may take you to non-federal websites. Their policies may differ from this site.

  1. Abstract

    The large data sets provided byin situoptical microscopes are allowing us to answer longstanding questions about the dynamics of planktonic ecosystems. To deal with the influx of information, while facilitating ecological insights, the design of these instruments increasingly must consider the data: storage standards, human annotation, and automated classification. In that context, we detail the design of the Scripps Plankton Camera (SPC) system, anin situmicroscopic imaging system. Broadly speaking, the SPC consists of three units: (1) an underwater, free‐space, dark‐field imaging microscope; (2) a server‐based management system for data storage and analysis; and (3) a web‐based user interface for real‐time data browsing and annotation. Combined, these components facilitate observations and insights into the diverse planktonic ecosystem. Here, we detail the basic design of the SPC and briefly present several preliminary, machine‐learning‐enabled studies illustrating its utility and efficacy.

     
    more » « less
  2. Abstract

    Modern in situ digital imaging systems collect vast numbers of images of marine organisms and suspended particles. Automated methods to classify objects in these images – largely supervised machine learning techniques – are now used to deal with this onslaught of biological data. Though such techniques can minimize the human cost of analyzing the data, they also have important limitations. In training automated classifiers, we implicitly program them with an inflexible understanding of the environment they are observing. When the relationship between the classifier and the population changes, the computer's performance degrades, potentially decreasing the accuracy of the estimate of community composition. This limitation of automated classifiers is known as “dataset shift.” Here, we describe techniques for addressing dataset shift. We then apply them to the output of a binary deep neural network searching for diatom chains in data generated by the Scripps Plankton Camera System (SPCS) on the Scripps Pier. In particular, we describe a supervised quantification approach to adjust a classifier's output using a small number of human corrected images to estimate the system error in a time frame of interest. This method yielded an 80% improvement in mean absolute error over the raw classifier output on a set of 41 independent samples from the SPCS. The technique can be extended to adjust the output of multi‐category classifiers and other in situ observing systems.

     
    more » « less
  3. Abstract

    Coastal physical processes are essential for the cross‐shore transport of meroplanktonic larvae to their benthic adult habitats. To investigate these processes, we released a swarm of novel, trackable, subsurface vehicles, the Mini‐Autonomous Underwater Explorers (M‐AUEs), which we programmed to mimic larval depth‐keeping behavior. The M‐AUE swarm measured a sudden net onshore transport of 30–70 m over 15–20 min, which we investigated in detail. Here, we describe a novel transport mechanism of depth‐keeping plankton revealed by these observations. In situ measurements and models showed that, as a weakly nonlinear internal wave propagated through the swarm, it deformed surface‐intensified, along‐isopycnal background velocities downward, accelerating depth‐keeping organisms onshore. These higher velocities increased both the depth‐keepers' residence time in the wave and total cross‐shore displacement, leading to wave‐induced transports twice those of fully Lagrangian organisms and four times those associated with the unperturbed background currents. Our analyses also show that integrating velocity time series from virtual larvae or mimics moving with the flow yields both larger and more accurate transport estimates than integrating velocity time series obtained at a point (Eulerian). The increased cross‐shore transport of organisms capable of vertical swimming in this wave/background‐current system is mathematically analogous to the increase in onshore transport associated with horizontal swimming in highly nonlinear internal waves. However, the mechanism described here requires much weaker swimming speeds (mm s−1vs. cm s−1) to achieve significant onshore transports, and meroplanktonic larvae only need to orient themselves vertically, not horizontally.

     
    more » « less
  4. Abstract

    Cross‐shore velocities in the coastal ocean typically vary with depth. The direction and magnitude of transport experienced by meroplanktonic larvae will therefore be influenced by their vertical position. To quantify how swimming behavior and vertical position in internal waves influence larval cross‐shore transport in the shallow (~ 20 m), stratified coastal waters off Southern California, we deployed swarms of novel, subsurface larval mimics, the Mini‐Autonomous Underwater Explorers (M‐AUEs). The M‐AUEs were programmed to maintain a specified depth, and were deployed near a mooring. Transport of the M‐AUEs was predominantly onshore, with average velocities up to 14 cm s−1. To put the M‐AUE deployments into a broader context, we simulated > 500 individual high‐frequency internal waves observed at the mooring over a 14‐d deployment; in each internal wave, we released both depth‐keeping and passive virtual larvae every meter in the vertical. After the waves' passage, depth‐keeping virtual larvae were usually found closer to shore than passive larvae released at the same depth. Near the top of the water column (3–5‐m depth), ~ 20% of internal waves enhanced onshore transport of depth‐keeping virtual larvae by ≥ 50 m, whereas only 1% of waves gave similar enhancements to passive larvae. Our observations and simulations showed that depth‐keeping behavior in high‐frequency internal waves resulted in enhanced onshore transport at the top of the water column, and reduced offshore dispersal at the bottom, compared to being passive. Thus, even weak depth‐keeping may allow larvae to reach nearshore adult habitats more reliably than drifting passively.

     
    more » « less