A carregar...
Projeto de investigação
Sem título
Financiador
Autores
Publicações
Drivers of ocean heat content variability and predictability in the North Atlantic
Publication . Costa, Teresa Maria Grazina do Carmo; Machado, Ana Margarida Silva Pereira Teles; Dutra, Emanuel Nemésio de Sousa; Ortega Montilla, Pablo
This PhD thesis focuses on the variability and predictability of the upper ocean heat content (OHC) in the North Atlantic (NA) basin, investigating the drivers of decadal predictive skill of OHC in climate models and potential differences among them. The NA basin is characterized by prominent decadal variability driven by internal and external sources, which modulates the local warming rates and the Northern Hemisphere climate. However, the contributions of the internal and external drivers of variability in the region remain largely unknown. This thesis aims at investigating how internal and forced variability contribute to local trends, as well as their roles in the local prediction skill using both climate model simulations and ocean reanalysis, to account for the observational and model uncertainty.
A first analysis based on historical and decadal prediction simulations with the EC-Earth3 climate model shows that internal variability is essential to understand the spatial pattern of North Atlantic OHC trends, contributing decisively to the local trends, and providing high levels of predictive skill in the Eastern Subpolar NA, the Irminger-Iceland Sea, and to a lesser extent in the Labrador Sea (LS). Skill and trends in other NA areas were mostly externally forced. Large observational uncertainties affect the evaluation of trends, interannual variability and predictability in the Central Subpolar NA, the only region exhibiting a cooling during the study period, for which results should be taken cautiously.
The analysis is further expanded by including seven additional climate models, to understand if they provided similar skill, or if intrinsic model characteristics, like key mean state biases, could influence the relative role of external forcings and internal variability. Particular attention is given to the LS and its surroundings, since it is found to be a region with low observational uncertainties and high inter-model spread of OHC decadal prediction skill. Large incertainties in the representation of the forced signals and the occurrence of initialization shocks in some models prevent us from reaching definitive conclusions about the origins of the LS decadal prediction skill.
This PhD thesis highlights the necessity of using several climate models when assessing OHC predictability. Relying on a single model can lead to misleading conclusions about the true sources of predictive skill.
Unidades organizacionais
Descrição
Palavras-chave
Contribuidores
Financiadores
Entidade financiadora
Fundação para a Ciência e a Tecnologia
Programa de financiamento
Número da atribuição
COVID/BD/152668/2022