RT Dissertation/Thesis
T1 Neural networks and the seawater CO2 system. From the global ocean to the Ría de Vigo
T2 Redes neuronais e o sistema do CO2 no medio mariño. Dende o océano global ata a Ría de Vigo
A1 Broullón Durán, Daniel
K1 2510 Oceanografía
K1 2502.9 Cambio climático
K1 2510.08 Interacciones Mar-Aire
K1 1203.04 Inteligencia Artificial
AB More than 98% of the carbon of the ocean-atmosphere system is stored in the oceans, dissolved in the form of inorganic carbon. During the 20th century, the ocean has absorbed approximately 30% of all man-made CO2 emissions from fossil fuel burning, cement production, and land use change. Consequently, the incessant increase of the concentration of this gas in the atmosphere causes a progressive decrease in the pH of the sea water. Many studies have shown that this acidification gives rise to different responses in marine organisms, many of them negative.The carbonate system regulates the pH of seawater. Knowing in depth the different processes that govern it at different spatiotemporal scales is necessary for a better understanding of future changes in the ocean. In the Iberian Upwelling Ecosystem, there are no models that adequately represent the carbon cycle. Therefore, this thesis aims to contribute to the increase of knowledge of this system.At an early stage, the objective is to design robust climatologies of the carbonate system variables: alkalinity (AT), dissolved inorganic carbon (TCO2), CO2 partial pressure (pCO2) and pH, as well as derived and relevant variables for assessments of impacts on organisms (saturation of aragonite and calcite). The potential of artificial intelligence techniques will be used in this section to improve the scarce climatologies available to date.The climatologies will be used in the second objective to model the carbonate system in the Iberian Upwelling Ecosystem. The modeling will be carried out with the Regional Ocean Modeling System (ROMS) for physics and the biogeochemical model PISCES for biogeochemistry. It is intended to begin with a low-resolution model that includes the Iberian basin and try to achieve a modeling of greater detail at the estuary level.
LK http://hdl.handle.net/11093/3995
UL http://hdl.handle.net/11093/3995
LA eng
DS Investigo
RD 15-oct-2024