REU 2023 Projects

Image
headshot of Verenice Becerril-Gonzalez

Verenice Becerril-Gonzalez

Geology major, Northland College
CEOAS mentors: Shan de Silva and Alejandro Cisneros de Leon

The age and origin of Cerro Volcán Quemado: An enigmatic young volcano in the Altiplano of Bolivia

In the Central Andes, volcanoes have had limited research due to their remoteness, one is Cerro Volcán Quemado located in Bolivia. The main objectives of the investigation are to 1) radiometrically eruption date the age of Cerro Quemado and 2) understand why Cerro Quemado contains rhyolitic magma instead of andesitic magma. Currently, researchers believe Cerro Quemado was formed during the Quaternary period, but there has been no published radiometric dating of this volcano. Cerro Quemado is a silicic phreatomagmatic volcano containing rhyolitic magma, making it unique since the volcanoes surrounding Cerro Quemado contain andesitic magma, not rhyolitic magma. Objective 1 will be understood through Uranium-Thorium-Helium dating of zircon crystals and objective 2 will be understood through whole-rock and mineral chemistry of major and trace elements as well as isotopic compositions.


Image
headshot of Kristin Cohrs

Kristin Cohrs

Physics and Biology major, Colby College
CEOAS mentor: Ken Hughes

Validating measurements of surface waves from turbulence profiling float using ocean wave model as baseline

Flippin’ χSOLO (FχS) is a specially designed Argo Float that measures turbulence, both on downward and upward profiles, and does so every 35 minutes. FχS also can measure surface waves. Flippin’ χSOLO's first long-deployment occurred in June 2023 and lasted 35 days in the west pacific. Here, we compare surface wave data from the FχS deployment against the WavewatchIII ocean wave model from Pacific Islands Ocean Observing System (PacIOOS). Comparisons of significant wave height showed FχS and WavewatchIII had similar wave height measurements with a percent error of -14% +- 26%; negative percent error means the model predicts larger values than the FχS with +- indicating standard deviation. However, WavewatchIII data dropped to zero unexpectedly on two different days, which is concerning since it is a simulated model. To further test WavewatchIII's reliability, we compared our results against results from two other datasets in a similar part of the world collected using a surface-following platform. These gave comparable results from a 2019 percent error of -8% +- 20% and in 2018 percent error of -15% +- 16%. In all cases, wavewatchIII over predicted by 10-15%. Although mean percent errors are reasonably good, the standard deviations are considerable, which limits the usefulness of wavewatchIII as a surface buoy proxy.


Image
headshot of Rachael Durr

Rachael Durr

Oceanography major, Oregon State University
CEOAS mentor: Mo Walczak

The final days of the southern Cordilleran Ice Sheet: geochemical and physical characterization of offshore sediments deposited by deglacial outburst floods from the Columbia River

Following the Last Glacial Maximum, the Cordilleran Ice Sheet (CIS) entered a deglacial period characterized by unstable episodes of growth and decay, including numerous flood events that have been recorded in both the terrestrial and marine record. The Columbia River was the main discharge point for these flood waters, which is how they were able to reach the Pacific Ocean and deposit offshore of Oregon. An improved understanding of how the CIS destabilized during the transition into the Holocene climate can help provide insight as to how the Greenland Ice Sheet could respond to the current climate change Earth is experiencing. OC1706B-11JC and OC2006A-27JC, two marine sediment cores from a sheltered offshore site on the upper continental slope south of Astoria Canyon were used to create a temporal framework of sedimentation across the early degalcial period. These two cores, collected about thirty meters apart, show many similarities but also distinct differences reflecting small-scale variability in sediment depositional processes. We here deconvolve stochastic variability in sedimentation from robust regional signals reflecting the flood record and ice sheet deglacial history. This stratigraphic framework was created based on density from Computed Tomography (CT) scans, grain size data, and X-ray Fluorescence (XRF) bulk geochemistry to uniquely characterize environmental variability recorded in each core. Through these analyses, we were able to create a unified record providing insights into the sources and the depositional energetics of the flood deposits and improving our understanding of deglacial dynamics of the Southern Cordilleran Ice Sheet.


Image
headshot of Josie Garcia-Stalker

Josie Garcia-Stalker

Physics (Astronomy) major, Brigham Young University - Idaho
CEOAS mentor: Erica Fleishman

Coincidence of social vulnerability and potentially hazardous levels of ultraviolet irradiance, ozone, and PM2.5 in Alaska, Idaho, Oregon, and Washington from 2005-2020

The Centers for Disease Control and Prevention--Agency for Toxic Substances and Disease Registry’s Social Vulnerability Index (SVI) uses sixteen factors from the US Census to identify communities that would need extra assistance before, during, and after hazardous events. Ultraviolet irradiance (UV) is associated with skin aging, skin cancer and cataracts, and can harm marine and terrestrial animals and plants, including crops. Ground level ozone can result in severe respiratory reactions in people and animals and contribute to foliar injury in plants. Fine particulate matter (PM2.5) can inflame and damage human and animal respiratory systems and adversely affect ecosystems including plants, soil, and water. Using data from the Environmental Protection Agency, we compared the overall SVI score for each county in Alaska, Idaho, Oregon, and Washington from 2005-2020 with their exposure to coincidental dangerous levels of UV, ozone, and fine particulate matter monthly and multi-annually. We also used the function “dataframe.corr” from “pandas” in python to determine the correlation coefficient (R value) among monthly average values of PM2.5, ozone, and UV irradiance. We found that exposure to high levels of UV, ground-level ozone pollution, and PM2.5 pollution occurs in the same month in several counties in Idaho, Oregon, and Washington but not Alaska. Almost all of these counties have at least a medium SVI score, indicating that they may need specialized preparedness, response resources, and support surrounding hazardous events.


Image
headshot of Anne Hansen

Anne Hansen

Marine Science major, California State University, Monterey Bay
CEOAS mentors: Brodie Pearson and Amrapalli Garanaik

Mapping the influence of wave-induced ocean surface mixing in the Arctic through non-dimensional parameters

The turbulence that drives mixing in the ocean's surface layer is driven by three main sources: convection, sheer and waves. While the latter source was formally understood as less significant, recent studies in open-ocean conditions have demonstrated that wave-driven mixing plays a key role in the transfer of energy in the upper layers of the ocean. In an effort to better characterize the relationship between wave-driven mixing and sea ice, we sourced model data depicting both wave and sea ice characteristics in the arctic, combining different parameters to calculate the Wave-Shear Mixing Ratio: a non-dimensional parameter that describes how much mixing energy can be attributed to waves or shear. The Proximal Wave-Shear Mixing Ratio, which accounts for mixing adjacent to sea ice, and not just under, was also calculated. We observed that not only are there high-levels of wave-dominant mixing adjacent to the edge of the arctic ice sheet in the model, but there are also areas well within the Marginal Ice Zone, a key area of ice melting and formation, that are adjacent to high levels of wave-dominant mixing.


Image
headshot of Anders Johnson

Anders Johnson

Biochemistry major, Spokane Falls Community College
CEOAS mentors: Rene Boiteau and Karl Romanowicz

Understanding the metabolic influence of iron on Trichodesmium colonies in the Red Sea

Iron is an essential trace nutrient for phytoplankton growth and cellular function that is often limited in low nutrient, oligotrophic regions of the ocean. To overcome iron limitation, phytoplankton such as the colony-forming cyanobacterium Trichodesmium have developed symbiotic associations with colony-dwelling epibiont bacteria that help solubilize iron-rich dust through the production of secondary metabolites. While previous work has illustrated the impact of iron deposition on Trichodesmium colony growth and iron bioavailability, the exact metabolites associated with the dissolution and uptake of dust-bound iron have yet to be defined clearly. By examining the suite of metabolites involved in iron dissolution and uptake, Trichodesmium-epibiont interactions can be better understood. Here, we incubated natural Trichodesmium colonies collected from the Red Sea in dust-rich and dust-poor conditions and performed metabolite extraction coupled with liquid chromatography mass spectrometry (LCMS) for untargeted metabolomic analysis. Our results indicate a clear distinction in metabolite production and abundance between iron-mediated growth conditions after 24-hour incubation. Specifically, we recovered over 14,000 unique metabolites of which 1,132 differed significantly in abundance between iron-rich and iron-poor growth conditions. We further found that the abundance of 648 metabolites was two-fold greater under iron-rich conditions relative to ironpoor conditions, with 227 metabolites greater than two-fold abundance under iron-poor conditions. These differences in metabolite abundance between iron-rich and iron-poor conditions following 24-hour incubation account for 73% of the variation in metabolite abundance between iron-mediated conditions. Together, our metabolomic analysis reveals important metabolic pathways and molecules associated with iron-rich dust utilization and uptake in Trichodesmium colonies and provides key insights into the molecular mechanisms of Trichodesmium and their epibionts that thrive in nutrient-poor waters.


Image
headshot of Annalise Khandelwal

Annalise Khandelwal

Environmental Science major, The Ohio State University
CEOAS mentors: Pam Sullivan and Sidney Bush

Relationship between rooting depth and hillslope position in two montane forests with contrasting rain-snow precipitation regimes

Tree roots play a crucial role in maintaining the forest ecosystem by supplying trees and surrounding soil with water, nutrients, and stability. Different hydrological and physical properties develop along a hillslope, impacting the size, distribution, and overall depth of roots. Soil development and plant-water availability are directly affected by precipitation patterns which leads plant rooting depths to vary depending on climate and moisture. Establishing the relationship between rooting depth, hillslope position, and soil moisture broadens our understanding of rooting strategies under a changing climate and variable water accessibility. Here we assess how rooting patterns differ between a hillslope’s toeslope and backslope within and among two coniferous forests with xeric climates and contrasting precipitation regimes. Our study was conducted at the rainfall-dominant H. J. Andrews Experimental Forest (HJA) in the central Cascade Mountain of Oregon, and the snowfall-dominant Sagehen Creek Field Station (SC) in the northern Sierra Nevada of California. We analyzed soil and rooting patterns from the toeslope and backslope at both sites by compiling root photographs from the soil pits in Fuji’s ImageJ. Using this software, we quantified coarse and fine roots embedded within each soil horizon by depth. We found that on average, the root abundance within the toeslopes at both H.J. Andrews and Sagehen Creek Field Station decreased by 25% and 26%. Within each backslope position at H.J. Andrews and Sagehen Creek Field Station, average rooting depth increased by 5% and 8%. The relationship between root abundance and depth was statistically significant with p values < 0.05 between the toeslopes at each site between 0-20cm in depth, the backslopes between each site at 20-40cm in depth, and both the toeslope and backslope positions between each site at a depth range of 40-80cm. The average fine root abundance decrease within the toeslopes at each site coupled with the statistical significance of both the backslopes and toeslopes at greater depths might reflect the prediction that rooting depth changes depending on hillslope position and groundwater accessibility.


Image
headshot of Alana Kihn

Alana Kihn

Computer Science major, Oregon State University
CEOAS mentors: Brodie Pearson and Amrapalli Garanaik

Accelerating Ocean Models with GPUs

This research investigates the performance of two computational software packages, Oceananigans.jl and GeophysicalFlows.jl, in simulating oceanic turbulence on different hardware configurations. Traditionally, Central Processing Units (CPUs) have been used for such simulations, but with increasing complexity, the need for higher computational power has led to a shift towards using Graphical Processing Units (GPUs). This study conducts a comprehensive performance benchmark analysis of the software packages across CPUs and two generations of NVIDIA GPUs (V100 and A100). The benchmark analysis evaluates aspects such as execution time and energy efficiency. Additionally, the benchmark examines the effect of different physical and numerical factors on simulation performance. Results show that GPUs are faster and use less energy per simulation when compared to CPU simulations, and that these GPU benefits become greater as the simulations increase in complexity (I.e., have more grid points). For the biggest simulations conducted here, the A100 GPU is approximately 1000 times faster and consumes 100 times less energy than the equivalent CPU simulations. Comparing GPU hardware, the new A100 GPU is about twice as fast and consumes half the energy per simulation when compared to the older V100 GPU.


Image
headshot of Claire Millwater

Claire Millwater

Environmental Sciences major, Oregon State University
CEOAS mentors: Kristen Buck and Léo Mahieu

Determination of Labile nickel in the Gulf of Mexico

Nickel (Ni) plays a very important role in marine ecosystems, especially in oligotrophic environments like the Gulf of Mexico because it is an essential micronutrient phytoplankton; however, its chemistry is not fully understood. Harmful algae blooms plague the Gulf of Mexico, being fueled by the nitrogen that Trichodesmium produces. Trichodesmium is a species of phytoplankton that utilizes Ni to increase its metabolic and population growth rates, since Ni is a cofactor for the enzymes implicated in nitrogen uptake. The minimum concentration of Ni in the Gulf of Mexico is ~1.8 nM; therefore, if there is excess Ni, why aren’t the Trichodesmium cells taking it up? A possibility is that the Ni that is left is not accessible to the phytoplankton. In order to better understand the relationship between Ni and harmful algae blooms, the amount of labile (accessible) Ni was determined in a surface transect and a depth profile off the Florida shelf; this was done using an approach based on cathodic stripping voltammetry on a mercury drop electrode and standard addition. It was found that above 100 m and below 750 m, all of the total dissolved Ni (dNi) is labile. However, this fraction changes between inshore and offshore at the surface, where half of the dNi is labile inshore and all of the dNi is labile offshore. The intermediate depths, between 160 and 560 m, showed that about half of the dNi is labile. More research is needed in the intermediate waters to determine whether the smaller fraction of LNi is due to this water coming in with more inert (non-labile) Ni or if there is Ni uptake occurring at these depths.


Image
headshot of Cassidy Ochoa

Cassidy Ochoa

Climate Science major, Oregon State University
CEOAS mentor: Andreas Schmittner

Variability of ocean δ13C from the Last Glacial Maximum to the Holocene

The Ocean Circulation and Carbon Cycling working group (OC3) has recently compiled a database of 287 sediment cores with data that represent the carbon isotopic composition of Earth’s oceans from the Last Glacial Maximum (roughly 23,000 to 19,000 years ago) to the Late Holocene (less than 10,000 years ago) (Muglia et al., 2023). Here we examine the spatio-temporal variability in this dataset by conducting an Empirical Orthogonal Function (EOF) analysis (also known as a Principal Component Analysis (PCA)). The original data were filtered by removing cores that don’t cover the full deglaciation and interpolated on a regular time axis with a time step size of 500 years. We find that the first two principal components (PC)account for 97.6% of the spatial and temporal variability in the dataset.  PC 1, which explains 90.6% of the total variability, is associated with a smooth global increase in d13C except for the upper North Atlantic where d13C values decrease. PC 2, explaining 6.9% of the total variability, shows a high loading mostly restricted to the upper North Atlantic, and a steep decline during the early phase of the deglaciation (Heinrich Stadial 1) followed by a more gradual increase afterward. Sensitivity tests exploring different time steps with sizes ranging from 50 to 1500 years and site exclusions (where 10% of sites were removed from the dataset) indicate robust results. Processes explaining the patterns are discussed, such as changes in terrestrial biomass, alterations in the Atlantic meridional overturning circulation, and the relationship between ocean temperature and δ13C. Our results may be useful for future investigations, including modeling studies, into how carbon and its isotopes within the world’s oceans have changed from the LGM to the Late Holocene.


Image
headshot of Bryce Pendleton

Bryce Pendleton

Ecosystem Science and Policy and Chemistry double major, University of Miami
CEOAS mentors: Kristen Buck and Léo Mahieu

Nickel Bioavailability to Phytoplankton in High Latitude HNLC Region Waters of the North Pacific and Southern Ocean

Trace metals in ocean environments play a critical role in the biological processes of phytoplankton. Nickel (Ni) is unlike any other trace metals, as it is found in large concentrations in the ocean, leading to it being under-researched, despite being essential for phytoplankton growth. Phytoplankton from the Pacific Northwest have shown variable Ni uptake by phytoplankton in incubation experiments, potentially related to Ni lability. It is thought that inorganic Ni compounds, free nickel, and some organic Ni compounds are labile nickel, but this will be the first time measurements of labile Ni are used to determine which form of Ni phytoplankton grows on. Additionally, more dissolved Ni was seen taken up in longer incubations, pointing to Ni being a nutrient which could be uptaken later on in phytoplankton growth. Labile Ni was determined using adsorptive cathodic stripping voltammetry on a mercury drop electrode and standard additions of Ni and compared to total dissolved Ni concentrations to assess whether the dissolved Ni taken up by phytoplankton in the experiments was labile Ni. Furthermore, Labile Ni measurements represent the bioavailable pool of dissolved Ni. The results prove that if enough time is allowed, phytoplankton gets to assimilate large amounts of dissolved Ni. Additionally, the proportion of the dissolved Ni that was taken up was accounted for by the labile Ni. These results provide better context for future research into defining labile Ni and its bioactivity to phytoplankton.


Image
headshot of Akira Regotti

Akira Regotti

Geobiology major, Pennsylvania State University - University Park
CEOAS mentor: Mo Walczak

The final days of the southern Cordilleran Ice Sheet: characterizing the palaeoceanographic conditions associated with by deglacial outburst floods from the Columbia River

The timing and mechanisms of retreat of the southern Cordilleran Ice Sheet remain poorly constrained, in part due to challenges in dating terrestrial geomorphic features. Two marine sediment cores–OC1706B-11JC (45° 51.949’ N, 124° 51.278’ W, 828 m water depth) and OC2006A-27JC (45° 51.953’ N, 124° 51.3008’ W, 828 m water depth)–collected from just south of the Astoria Canyon document recurrent late-Pleistocene deglacial flood events routed through the Columbia River. Characterizing the paleoceanographic conditions associated with deglacial flooding offers insight into climate feedbacks and impacts relevant to future climate change. Here we utilize radiocarbon dating of benthic and planktonic foraminifera to establish a temporal framework for the episodic late-Pleistocene floods. We present preliminary analyses of foraminiferal δ18O, δ13C (N. pachyderma, G. bulloides, U. peregrina, C. wuellerstorfi), and Mg/Ca ratios (N. pachyderma, G. bulloides) to reconstruct oceanographic temperature, salinity, and nutrient availability through the deglaciation, including conditions leading up to and immediately following episodes of outburst flooding. The faunal abundance and diversity of foraminifera throughout the cores reflect ecological shifts driven by environmental changes, further contextualizing oceanographic conditions. In concert, these observations offer insight into the feedbacks between Cordilleran Ice Sheet behavior and the adjacent ocean through the end of the last ice age.


Image
headshot of Genasea Riegard

Genasea Riegard

Microbiology major, Oregon State University
CEOAS mentor: Byron Crump

Are estuarine microbes adapted to respire DOC from their native estuary?

Estuarine microbes rely heavily on the native DOC that is available to them in their environment to go through respiration and make biomass. If the DOC they are accustomed to was swapped with DOC from a nonnative source (coming from a different estuary), their respiration rates could change. We hypothesize that the respiration rate will be lower when the microbes use nonnative DOC to respire because estuarine microbes should be well-adapted to respiring their native DOC. To test this hypothesis we will be taking microbes from Coos Bay estuary in Oregon and two estuaries in Alaska, Jago Lagoon and Elson lagoon, swapping their DOC, and then measuring their respiration rates. We will compare the respiration rates to see if it's lower or higher than their respiration rates with native DOC. This will tell us the importance of DOC lability and microbial diversity/adaptability for microbial growth.


Image
headshot of Genevieve Sauret

Genevieve Sauret

Geology major, Oregon State University
CEOAS mentors: Erin Pettit, Celia Trunz, Christian Wild, Meghan Sharp

A first look at the geometry and internal structure of an ice shelf basal melt channel on the Thwaites Eastern Ice Shelf

Basal channels, which form on the undersides of floating ice shelves connected to larger ice sheets, are a relatively new area of study necessitating further research to understand their impacts on both the cryosphere and humanity. We aim to examine one such basal channel on the Thwaites Ice Shelf to better understand its geometry and how it varies temporally and spatially. These channels may become weak zones within ice shelves due to local ice thinning, which in turn impacts ice shelf stability and the rate of ice sheet mass loss and future global sea level rise. To better understand these channels, we will examine ice-penetrating radar data collected from the Thwaites Ice Shelf to visualize both the channel and englacial layers within the ice at various points along the channel. Using this data, we can gain a better grasp on channel geometry and how channels may evolve through time and space and impact the shelf.


Image
headshot of Sabrina Uribe

Sabrina Uribe

Meteorology major, Florida State University
CEOAS mentor: Erica Fleishman

Relations between the El Niño-Southern Oscillation and river flow in Oregon and northern California

The El Niño-Southern Oscillation (ENSO) is one of the most prominent climate phenomena as it is responsible for many changes in weather patterns around the world. ENSO generally is believed to have different effects on the southwestern and northwestern United States. Effects of ENSO on Oregon and northern California, which lies in a transitional area between the southwestern and northwestern regions, are more ambiguous. We explored possible quadratic or linear relations between the strength of ENSO and precipitation and streamflow in eight basins: Upper Klamath Lake, Upper and Lower Rogue, North and South Umpqua, Sprague, and Williamson in southern Oregon and Upper Sacramento in California. Linear and quadratic correlations between ENSO and streamflow were statistically significant, but no correlation coefficient exceeded -0.062. Extreme El Niño values tended to be associated with a slight increase in precipitation values.


Return to Summer Research Internships for Undergraduates