The tropical Atlantic Ocean (TAO) controls exchanges from one hemisphere to the other and is a place of strong interactions with the atmosphere. The TAO is home to five of the world's largest rivers as well as intense rainfall in the intertropical convergence zone (ITCZ). This induces large spatial variability of salinity and of air-sea CO2 flux. While the global ocean is a strong CO2 sink, the TAO is a strong source of CO2 to the atmosphere due to the deep waters rich in inorganic carbon upwelled to the surface at the equator. However, this source is mitigated by the low CO2 concentrations in the Amazon River plume whose freshwater is low in inorganic carbon and favours phytoplankton blooms. It is in this context that propagate the tropical instability waves (TIWs) and the North Brazil current (NBC) rings, the two dominant mesoscale forms in the TAO. The objective of this work is to describe and understand the variability of the surface salinity and CO2 fluxes associated with the mesoscale. In-situ observations collected during cruises and Argo floats are coupled to surface satellite salinity, temperature and chlorophyll-a. In the equatorial Atlantic the salinity gradient between the fresh water from rainfall under the ITCZ and the salty water of the equatorial upwelling is very strong in May-June. The TIWs strongly distort this gradient, and are therefore particularly well observed in surface salinity. The observation of TIWs in salinity provides complementary information to their observation in surface temperature on their seasonal and interannual variability. Furthermore, salinity does not only play the role of a passive tracer, as together with temperature, it determines the seawater surface density. This affects the energy that allows TIWs to develop and propagate. One of the energy sources is the potential energy generated by the deformation of the density gradient. The effect of salinity on this energy is as strong as that of temperature, which means that by adding the contribution of salinity, the potential energy is doubled. TIWs modify and are modified by the salinity in the equatorial Atlantic. On the western edge of the basin, the Amazon plume results in even more salinity variability than in the equatorial Atlantic. The NBC rings, eddies that are 200 km in diameter, are highly contrasted structures. They trap the salty, CO2-rich waters of the NBC, but their rotation stirs water from the Amazon plume. The fresh water of the plume enhances the exchanges of CO2 and heat with the atmosphere. The northwestern TA in February 2020 was found to be a CO2 sink 10 times stronger than expected, and more than 40% of this flux is due to the effect of eddies. Their role is twofold, on the one hand they stir the plume which becomes a strong carbon sink, but also, they do not retain the CO2-rich surface signature of the waters they trap, and instead often stir freshwater filaments. The situation in summer is very different from the one in winter. The NBC changes its orientation by 90° and instead of following the South American coastline, it flows towards Africa. The NBC passes the mouth of the Amazon that is close to its maximum discharge and advects the plume eastwards. However, the NBC rings and the winds change this classical pattern. The formation and propagation of eddies make the plume discontinuous, and the winds favour a northwestward transport of fresh water. Thus, in August -September, whereas part of the plume indeed flows eastwards, another part is advected towards the Lesser Antilles. Particularly strong examples of this were observed in late summer 2021. This work shows the importance of the oceanic mesoscale for understanding key phenomena, such as the propagation of the TIWs and of the Amazon plume and the TAO carbon budget.