Nitrate, silicate, and phosphate are essential nutrients in diatom based food webs. Chlorophyll and nutrients are good indicators of phytoplankton abundance and diversity. Phytoplankton, being integral to the Possession Sound ecosystem, can be indicators of greater change in an ecosystem. By studying phytoplankton abundance and diversity along with chlorophyll and nutrient levels spatially and temporally, correlation can be determined and used to help understand the health of Possession Sound. It was predicted that nutrients and chlorophyll abundance are inversely proportional, while chlorophyll and phytoplankton abundance and diversity are directly related. Thus, higher nutrient levels indicate less chlorophyll and fewer nutrients indicate higher chlorophyll and plankton abundance and diversity. Increased levels of nutrients in the fall and winter were expected, with greater chlorophyll and plankton levels in the spring and summer. The chlorophyll and plankton abundance and diversity were anticipated to have gone down over the past four years, while nutrient levels will have gone up slightly. Students at the Ocean Research College Academy (ORCA) collect monthly samples as part of the longitudinal study: State of Possession Sound (SOPS). Results from three locations were utilized from 2016 to 2019. Chlorophyll is measured by a YSI EXO2 Sonde, while nutrient samples are taken using the Niskin bottle and sent to the University of Washington Marine Chemistry lab to test for results. By evaluating seasonal data, temporal trends of chlorophyll, nutrients, and plankton abundance and diversity were discovered. Changes in data can be linked to environmental and anthropogenic variations. It would be compelling to analyze ecosystem changes by exploring dissolved oxygen and pH levels.

Phytoplankton, one of the primary sources of dissolved oxygen in marine ecosystems, are dependent upon nutrients for growth. However, there is evidence that eutrophication, the overabundance of nutrients, can lead to hypoxia in marine ecosystems. Because they are a primary source of dissolved oxygen and are dependent upon nutrients, phytoplankton density can indicate how nutrients are affecting dissolved oxygen at depth. It was hypothesized that an increase in phosphates, nitrates, and nitrites would correlate with an increase in phytoplankton density. Furthermore, it was predicted that with greater phytoplankton density there would be a greater difference in dissolved oxygen at the halocline versus dissolved oxygen 30 meters below the halocline. This study used data collected from 2016 to 2019 by students at the Ocean Research College Academy (ORCA) at two sampling stations in Possession Sound, WA. Water was collected and sent to the University of Washington Marine Chemistry Lab for nutrient analysis. Phytoplankton density was calculated using samples collected during 3-minute horizontal tows at the halocline. Dissolved oxygen data was collected using a YSI Exo2 Sonde at different depths. Preliminary results suggest that greater levels of phosphates, nitrates, and nitrites may show a steeper oxycline due to an increase in density of phytoplankton from the nutrients. The potential for hypoxia is increasing because of anthropogenic nutrients, so understanding the influence humans have over nutrients in marine environments is critical. This study will help us to understand how humans are influencing Possession Sound and marine ecosystems as a whole as a result of the impact of nutrients on phytoplankton and dissolved oxygen.

The Port of Everett is tucked into the Snohomish River Estuary in the small city of Everett, Washington. The Snohomish River feeds into Possession Sound, an inlet of Puget Sound, creating a salt wedge estuary that is host to a lively ecosystem, which can be greatly harmed by the presence of plastic debris. Students of Everett Community College’s Ocean Research College Academy (ORCA) program began collecting data from a Seabin located on the Port of Everett’s K Dock in November 2019. The Seabin is a device with a mesh net set inside a pump designed to draw surface water and floating plastic marine debris into it. However, observations of the Seabin located in Everett, Washington show it also collects natural and other anthropogenic debris. It was hypothesized that river discharge would have a positive correlation with mass collected by the Seabin. Data were collected weekly and cataloged by mass and qualitative observations. Each catch was weighted both wet and dry, then sorted when dry to determine the composition of each catch. Anthropogenic debris was separated and cataloged into one of the following eight categories: plastics, metals, textile fibers, cement, oils, papers, paints, and fiberglass, while natural debris was recorded similarly as plant matter, algae, dirt or mud, hair, fish, birds (including feathers), and live organisms such as bugs, fish and plankton. The wet and dry mass of each catch were recorded as well and compared to river discharge data recorded by the United States Geological Survey from a probe set in the Snohomish River. Data from this project will be part of an awareness campaign for educating marina goers.

During September and October, endangered Olive Ridley sea turtles have been observed swimming up onto the beaches of Salinas Grandes, Nicaragua to nest. Literature has identified different environmental factors that affect sea turtle nesting activity. This study compared moon phase, lunar illumination, and tide height to the number of nests observed. The data for these factors were collected by tides4fishing, a company that collects data on tides, solunar activity, moon phases, lunar illumination, and fishing sites in North and Central America. Data for the number of nests, along with the time of night they were counted, were provided by Turtle Tribe, a sea turtle conservation project run by a non-profit called Water and Light International in Salinas Grandes, Nicaragua. I partnered with Turtle Tribe to use their data for conducting research that could aid in future conservation efforts; I even collected some nest data myself when I traveled to Nicaragua. It was hypothesized that the greatest number of nests would occur when there is the least amount of lunar illumination and at a high tide level. The limited light could act as protection for the sea turtles from predators and the high tide would allow them to walk farther up the beach where the nests are not in danger of being drowned by the tide. Least-squares regression analysis was performed to check for correlation between these factors. The hypothesis was not supported by these data. More data are needed to conclusively determine whether there is a correlation between these environmental factors and the number of Olive Ridley sea turtle nests. Additional years of data and environmental factors such as the steepness of the beach would be useful.

Monitoring marine mammals can help researchers observe the adverse effects of pollution and climate change in the ocean. Passive acoustic monitoring is one of the many ways researchers observe mammals, as most aquatic mammals communicate with sound. I utilize new object detection methods and vision-based neural networks to automatedly detect and observe marine life. First, I use annotated fin-whale acoustic data to produce spectrograms needed to train and test the neural network. The neural network is composed of two main parts: the feature extractor CNN and the object detector. A pretrained Convolutional Neural Network (CNN) is a neural network trained on over a billion images from Image.net, thus it “understands” how to look for features. Here, I use a pretrained CNN Resnet18 to extract important visual features from the spectrograms. I then change the last layers of the pretrained neural network to include You Only Look Once v2 (YOLO) model, which is an object detection model that classifies parts of an image into different categories. The resulting network should be able to take a spectrogram as input and identify which part of the image contains the fin-whale call (if any). The findings from this study offer a new way to detect fin-whale calls using underwater acoustic data.

In the study of plankton, it is common to count and identify them manually with the use of a microscope and sampling containers, which can be a tedious process. To address this problem, a Python-based neural network will be created to automatically identify common phytoplankton genera in Possession Sound. Since the most abundant phytoplankton in Possession Sound are diatoms, which include Thalassiosira, Coscinodiscus, and Chaetoceros, the network’s primary purpose will be to identify these genera. The neural network will be trained using approximately 1000 photos of each genus in varying orientations and lighting conditions, with the images being drawn from research trips aboard the Ocean Research College Academy vessel Phocoena beginning in 2007. After completing the training process, the network’s performance will be validated using samples taken at two sites around Possession sound, and it will be determined whether it meets a benchmark of 80% accuracy. It is expected that a number of challenges will be encountered with distinguishing between phytoplankton that are distorted or layered on top of one another, and these issues could be further addressed in the future. Despite these possible problems, the neural network shows promise as a low-cost alternative to current automated phytoplankton identification devices such as the FlowCAM, which can cost upwards of $100,000.

Nearly half of Americans live in earthquake prone areas. Many primary fault zones that host large earthquakes, such as the Cascadia, Alaska, and San Andreas fault zones, extend into the offshore regime. These offshore fault systems have been historically difficult to study due to challenges in observational techniques. Through the creation of an algorithm that uses geospatial analytical tools, this study seeks to identify seafloor faulting structures from data collected by high frequency multichannel acoustic methods. In doing so, we improve our capabilities of characterizing offshore fault zones. In addition, we examine these geospatial analytical methods for accuracy and explore the impact of data collection and post-processing procedures on associated errors. Data utilized subsists of bathymetric data collected in the Cascadia and South African regions, which are active and passive margins respectively. Methods for surface fault identification include visual inspection, as well as geospatial analytical methods consisting of the Bathymetric Position Index, slope, and aspect of surface morphology. Faults identified from surface morphology are compared to those identified using a coherence-based detection method from seismic reflection data. Surface expressed faults indicate high-amplitude and/or recent geologic deformation and can give insight into tectonic stress regimes and associated faulting hazards. An improved understanding of faulting hazards through efficient surface fault identification would aid in preparation and planning for earthquakes. Through the creation of this algorithm, our capabilities to accurately identify surface expressed faults in bathymetric datasets will be enhanced and thus our understanding of global tectonic processes and earthquake risks to population centers like those in the Pacific Northwest will be improved.

 Nutrient presence is intrinsic to and required by living organisms in all marine environments. Among the most prominent of these nutrients are nitrogen and phosphorus, which are essential to aquatic flora, and iron, which is required by phytoplankton. Such nutrients are introduced to estuarine marine environments by a variety of means including but not limited to agricultural and urban runoff, river discharge, precipitation, marine sediments, and ocean upwelling. Patterns of nutrient presence therefore affect patterns of life in estuaries. In an effort to determine a relationship between the biology linked intrinsically to nutrient presence and bathymetry, we explored the question of how nutrient presence changes in areas of varying seafloor depth in the Snohomish River Estuary. We compiled raw data concerning ammonium, nitrate, nitrite, silicate and phosphate presence measured in µm in water samples (tested by the University of Washington Marine Chemistry Lab) as well as corresponding measurements of seafloor depth, as taken by a Niskin bottle and single-beam sonar respectively, provided and recorded by Ocean Research College Academy (ORCA) at field sites within the Snohomish River Estuary System. Our hypothesis stated that phosphate, silicate, nitrate, nitrite, and ammonium collected from shallow locations would have a higher and more homogenous presence at varying depths than those collected in deeper ones due to there being greater vertical mixing in shallow areas. Additionally, we expected more nutrient diversity in samples taken from shallower areas because tides and ocean upwellings can facilitate the mixing of nutrients of varying densities and functions better there than in deeper areas. Any findings of a relationship between seafloor depth and nutrient presence could aid in the development of the ability to predict potential nutrient dynamics and consequent biological implications in the Snohomish River Estuary System using bathymetric data.