This dataset contains the results of the paper published in Tropical Oceanography titled "Influence of underwater hydrodynamics on oil and gas blowouts off Amazon River mouth" and the paper published in Dynamics of Atmospheres and Oceans with the title "Amazon River plume influence on Western Tropical Atlantic dynamic variability", in addition this dataset was also used in the paper titled" Amazon river plume influence on planktonic decapods in the tropical Atlantic" which is published in the Journal of Marine Systems. This dataset was obtained through a simulation using the ROMS model (Regional Ocean Model System) and was adjusted to reproduce the best possible hydro-thermodynamics in the area of the mouth of the Amazon and Para rivers, as well as in the area of the retroflexion of the North Brazil Current (NBC), showing how the low salinity waters of the Amazon and Para rivers are transported in the different months and weeks of the year. This dataset is located in an area belonging to Western Tropical North Atlantic (WTNA), geographically framed at 60.5°–24°W and 5°S–16°N. The simulation was performed with 0.25° horizontal resolution and 32 vertical levels. As the Amazon River has a well-branched delta that is defined by 4 channels, the total monthly mean discharge was divided among the 4 mouths according to their dimensions, that is, into 4 virtual rivers; The river Para does not have a delta, so it was considered as it is in reality. The temperature, salinity and marine currents distributions obtained numerically were compared with SODA (Simple Ocean Data Assimilation), with the in situ observations of the PIRATA project buoys (Prediction Research Moored Array in the Tropical Atlantic) located at (38°W, 8°N) and at (38°W, 12°N) and with the surface currents of SCUD model (Surface Currents from Diagnostic model), in all cases small differences were obtained between the numerical variables and those of the SODA, PIRATA and SCUD datasets, obtaining an optimal hydro-thermodynamic model of that area.