Date of Award

Spring 5-2020

Level of Access Assigned by Author

Open-Access Thesis

Degree Name

Doctor of Philosophy (PhD)

Department

Oceanography

Advisor

Emmanuel Boss

Second Committee Member

Lee Karp-Boss

Third Committee Member

Larry Mayer

Additional Committee Members

Collin Roesler

Heidi Sosik

Abstract

Phytoplankton are microscopic photoautotrophs living in the surface ocean waters and help support all life on earth via photosynthetic production of oxygen. Thousands of species make up the bulk phytoplankton community, and the spatial and temporal distribution of different types of phytoplankton has relevance for many ocean ecosystem questions including marine food web dynamics, and carbon flux and sequestration. Methods to detect phytoplankton community composition (PCC) on the vast scale of the global ocean require estimates of PCC from remote platforms, namely earth-observing satellites. The use of satellite data to observe and interpret PCC in the surface ocean requires significant effort to develop and evaluate algorithms based on measurements made in situ; the work of this thesis contributes to that effort.

Information from both global and regional (North Atlantic Ocean) datasets is applied to develop methods to estimate phytoplankton pigment concentrations, phytoplankton size classes, and diatom carbon concentrations. Optical spectra, specifically hyperspectral remote-sensing reflectance, are used in the algorithm for estimating phytoplankton pigments, which resolves the concentrations of three pigments and one pigment group (chlorophylls a, b, c, and photoprotective carotenoids). This result has implications for use with hyperspectral ocean color data measured by satellite. A novel dataset of open-ocean image-in-flow cytometry is used to evaluate and improve a commonly applied phytoplankton size class algorithm, as well as to calculate diatom carbon and develop a model to map diatom carbon using environmental parameters as model input. Biases and uncertainties in the size class algorithm are reduced by our method relative to previously published work for all three size classes (pico-, nano-, and microplankton). Diatom carbon measurements from quantitative cell imagery elucidate the variability of diatom biomass as function of chlorophyll a concentration, and this novel information enables improved methods to detect diatoms from space.

The findings of this thesis are relevant to large-scale studies of ocean ecosystems and are critical for algorithm development using both current and upcoming earth-observing satellite data. Additionally, the results presented here provide tools that will benefit oceanographic research on spatial scales relevant to a changing ocean climate.

Download

Files over 10MB may be slow to open. For best results, right-click and select "save as..."

Share