The Impact of Freshwater and Phytoplankton on Mixed Layer Heating

Thesis (Ph.D.)--University of Washington, 2022Mixed layer heat budgets are governed by a range of horizontal and vertical processes, including net heat flux into the mixed layer. Net heat flux into the mixed layer is determined not only by the values of the heat fluxes at the air-ocean interface, but also by the depth to which shortwave radiation penetrates and the depth of the mixed layer. This dissertation investigates a range of physical and biological processes influencing these latter two factors, using observational data from autonomous profiling floats, as well as machine learning and a one-dimensional model. In Chapter 2, I present a new method for identifying barrier layers, a phenomenon whereby the near-surface density stratification is determined by changes in salinity. Barrier layers are typically associated with shallower mixed layers, and therefore potentially of critical importance in influencing the mixed layer heat budget in regions where they cover a large percentage of the ocean. The method developed in this Chapter differs from previous methods in that it does not require choice of a temperature threshold, something that has hindered comparison of previous studies, and is therefore applicable without alteration throughout the tropical and subtropical oceans. In Chapter 3 I use this method to identify the prevalence of barrier layers in the Arabian Sea, and show that they are largely associate with decreased heating or increased cooling of the mixed layer because they allow more shortwave radiation to penetrate below the base of the mixed layer. In some cases, they allow enough radiation to penetrate that the net heat flux into the mixed layer shifts from positive to negative. In Chapters 4 and 5, I turn to the question of how the distribution of chlorophyll in the water column affects the absorption of shortwave radiation. Because no climatology of vertical chlorophyll in the global ocean exists, in Chapter 4 I use a machine learning clustering method to identify regional and seasonal patterns in the distribution of chlorophyll. I show that there are distinct patterns in the types of profiles that occur in different regions. I use the patterns developed in Chapter 4 as a tool for a one-dimensional modeling study in Chapter 5, where I quantify the expected impact of various chlorophyll-dependent parameterizations of shortwave radiation. I identify regions where the choice of parameterization is most sensitive to chlorophyll, and specifically those regions where the vertical distribution leads to a large departure from a surface-average based parameterization