Evaluation of South Atlantic Thermohaline Properties from BESM-OA2.5 and Three Additional Global Climate Models

• Main Authors: Broggio, Micael Fernando, Garcia, Carlos Alberto Eiras, Silva, Renato Ramos da
• Format: Article in Journal/Newspaper
• Language: English
• Published: Instituto Oceanográfico - Universidade de São Paulo 2022
• Subjects: Climate model; Models evaluation; Thermohaline properties; Antarc*; Antarctic; NADW; South Atlantic Ocean; Southern Ocean
• Online Access: https://revistas.usp.br/ocr/article/view/199402

Important global oceanic processes, such as the meridional overturning circulation, are governed by the temperature andsalinity of the ocean. As such, it is essential that these properties be correctly represented in high-quality global climatemodels. This study aims to evaluate thermohaline properties both historically and under two simulations of the BrazilianEarth System Model BESM-OA2.5 in the South Atlantic Ocean (Representation Concentration Pathway (RCP) 4.5 and 8.5).Since error assessment in the global climate model (GCM) is fundamental to infer climate change projections, comparisonswere made for thermohaline properties among four GCMs (HadGEM2-ES, MIROC-ESM-CHEM, MIROC5, and BESM-OA2.5)against data from ocean monitoring programs and from ORAS5-ECMWF. The results show common surface spatial patternerrors in all models, commonly related to mesoscale processes. Specific to BESM-OA2.5 over the Southern Ocean, weobserved an increase in the temperature bias during autumn and summer, probably due to subsurface temperatureoverestimation linked to North Antarctic Deep Water (NADW) formation. With respect to salinity, the underestimationsin the Subtropical/Subantarctic Zones and in the north of the South Atlantic subtropical gyre were linked to simulationerrors in the Malvinas current. All models presented overestimated annual historical temperature rates, with BESM-OA2.5being the closest to ORAS5. In the subsurface, the BESM-OA2.5 did not easily simulate the South Atlantic Central Water(SACW) formation, though in deep water, the model was able to better simulate the Antarctic Intermediate Water andNADW patterns. Statistically, the multi-model means performed better, while the BESM-OA2.5 performed worst amongthe models in both methodologies applied. In terms of projected scenarios, the models demonstrated sensitivity tovariations in greenhouse gas emissions between the RCPs, with higher magnitude warming predicted in the equatorialzone, except for BESM-OA2.5