OCEAN BIOGEOCHEMICAL RESPONSES TO FORCINGS ON WEEKLY TO CENTENNIAL TIMESCALES

Abstract

Anthropogenic carbon emissions are altering the Earth’s climate and oceans. Sea-water is a major sink for carbon dioxide (CO2), and the strength of this sink - about one third of human emissions - depends largely on the oceanic carbon cycle. This cycle is controlled by complex interactions between oceanic currents, mixing, temperature, primary productivity and carbonate chemistry. The intricacy of the carbon cycle poses a challenge for determining the present and future partitioning of carbon between the atmospheric, oceanic and terrestrial reservoirs. Observational, theoretical and numerical modeling efforts have been undergoing to elucidate the present-day processes that increase/decrease the carbon dioxide content in the ocean. These efforts have made it possible to estimate the Earth’s carbon budget and have informed policy makers about the urgency of reducing carbon emissions and protecting marine ecosystems. Yet, the future impacts of climate change on the carbon cycle need to be further refined to improve the accuracy of climate and marine ecosystem projections. In this context, this thesis contributes to the understanding of two components of the carbon cycle at different spatiotemporal scales. In Part I, the mechanisms of the future changes in seasonal and interannual variability of the global ocean’s partial pressure of CO2 (pCO2) are examined. In Part II, an offline high-resolution advection-diffusion model is developed and coupled to a marine ecosystem model based on the Marine Biogeochemistry Library (MARBL). This model is used to study the local effects of tropical cyclones on the biological activity of the Kuroshio region.In Part I, several fully coupled atmosphere-ocean-biogeochemistry models from the Coupled Model Intercomparison Project 5 (CMIP5), under the high emission scenario pathway 8.5 (RCP8.5), are used to answer the following questions: (1) what changes will the future temporal variability of the oceanic partial pressure of CO2 (pCO2) experience?, and (2) what are the mechanisms driving those changes? These questions are answered for two different time-scales: the annual seasonal cycle (Chapter 2) and interannual scales (Chapter 3). The global simulations reveal that the seasonal amplitude (climatological maximum-minus- minimum) of upper ocean pCO2 will increase by a factor of 1.5 to 3 times by 2080-2100 compared to 2006-2026 for the RCP8.5 emission scenario. The global interannual variability of the sea-surface pCO2 (calculated as 1σ) will increase by ∼ 64 ± 20% by 2045-2095 relative to the beginning of the industrial revolution. To unravel the mechanisms behind the amplification we use a complete analytical Taylor expansion of pCO2 variability in terms of its four drivers: dissolved inorganic carbon (DIC), total alkalinity (TA), temperature (T) and salinity (S). The linear approximation allows a separation of the effect of the buffering capacity of the ocean, from the drivers induced by physical and biological phenomena. This study shows that a decrease in buffering capacity is the main cause of pCO2 variability amplification, but not the only one. In regions dominated by T, the amplification is a consequence of mean CO2 build up. Further, a decrease in DIsC variability counteracts the seasonal and interannual amplifications. The intra-model differences in pCO2 variability are also characterized. For example, in the equatorial Pacific, at interannual scales, some models feature an amplification of the pCO2 variability dominated by T, while others show a decrease in pCO2 variability dominated by DIC. In Part II of this thesis, an offline marine ecosystem model is described (Chapter 3). The model is forced with ocean currents from the high-resolution Community Earth System Model 1.2.2., which features several nutrients, phytoplankton groups, zooplankton, dissolved organic matter and particulate matter. In Chapter 4, the following questions are studied: (1) How do tropical cyclones (TCs) affect nutrients and primary productivity on the Kuroshio region? and (2) what are some factors that control the magnitude of the TC-induced anomalies? Surface chlorophyll blooms due to TCs have been widely detected with satellite observations. But the phytoplankton’s response at depth is still not well understood. The results of this chapter show that a TC encountering a cyclonic-eddy largely promotes upwelling of nutrients and subsequent increase in subsurface primary productivity. However, the impact at the surface is negligible. Initially, the TC induces a negative chlorophyll anomaly, due to vertical advection of water from below the euphotic zone. About a week after the TC, an increase in new primary productivity is observed. This suggests that satellites may be missing these TC-induced subsurface anomalies, which could be a mechanism to sustain life in regions with deep nutriclines.