These data contain the annual, seasonal and monthly biome partitioning of the global surface ocean on a 1° latitude x 1° longitude resolution for the ocean mixed layer. The biomes were defined using self-organizing maps and hierarchical clustering, drawing on the biogeographic patterns of 536 phytoplankton species, projected based on global in situ presence observations and biogeographic extrapolation using statistical species distribution modeling (Righetti et al., 2019). This approach resulted in nine global clusters at the monthly scale. The clusters were used to define monthly biomes as spatially coherent units with consistent phytoplankton community composition that cover at least 0.5% of the surface ocean area. We define the seasonal and annual biomes as those biomes most frequently occurring during the months of each season and during the full year for each 1°-pixel, respectively.