| Summary: | The abundance and size distribution of marine organic particles are two major factors controlling biological carbon sequestration in the ocean. These quantities are the result of complex physical-biological interactions that are difficult to observe, and their spatial and temporal patterns remain uncertain. This dissertation describes our analysis of particle size distributions (PSD) and the resulting export, from a global compilation of \textit{in situ} Underwater Vision Profiler 5 (UVP5) optical measurements. In Chapter 2, we demostrate the ability to extrapolate sparse UVP5 observations to the global ocean from well-sampled oceanographic variables, using a machine learning algorithm. We reconstruct global maps of the biogenic PSD parameters (biovolume and slope) for particles at the base of the euphotic zone. These reconstructions reveal consistent global patterns, with high chlorophyll regions generally characterized by high particle biovolume and flatter PSD slope, i.e., a high relative abundance of large vs. small particles. The resulting negative correlations between particle biovolume and slope further suggest amplified effects on sinking particle fluxes. Our approach and estimates provide a baseline for understanding the export of organic matter from the surface ocean. Chapter 3 describes how applying a simple empirical relationship to our reconstructions of the PSD, we can calculate the total export. In this Chapter, we explore the seasonal and spatial patterns of carbon export. Taking advantage of the high vertical resolution of the UVP5, we quantify the export from the surface using two previously established depth horizons. We identify a larger export from the Southern Ocean than most other models of export. Similarly, we find the lower part of the euphotic zone to be dominated by heterotrophy, rather than autotrophy. Being able to reconstruct the PSD and particle flux at multiple depths allows for further exploration of the full 3-dimensional particle field. Chapter 4 describes a full 3-D model, ...
|
|---|