Whale migrations are poorly understood. Two competing hypotheses dominate the literature: 1. moving between feeding and breeding grounds increases population fitness, 2. migration is driven by dynamic environmental gradients, without consideration of fitness. Other hypotheses invoke communication and learned behaviors. In this article, their migration was investigated with a minimal individual-based model at the scale of the Global Ocean. Our aim is to test if global migration patterns can emerge from only the local, individual perception of environmental change. The humpback whale (Megaptera novaeangliae) meta-population is used as a case study. This species reproduces in 14 zones spread across tropical latitudes. From these breeding areas, humpback whales are observed to move to higher latitudes seasonally, where they feed, storing energy in their blubber, before returning to lower latitudes. For the model, we developed a simplified ethogram that conditions the individual activity. Then trajectories of 420 whales (30 per DPS) were simulated in two oceanic configurations. The first is a homogeneous ocean basin without landmasses and a constant depth of −1000 m. The second configuration used the actual Earth topography and coastlines. Results show that a global migration pattern can emerge from the movements of a set of individuals which perceive their environment only locally and without a pre-determined destination. This emerging property is the conjunction of individual behaviors and the bathymetric configuration of the Earth’s oceanic basins. Topographic constraints also maintain a limited connectivity between the 14 DPSs. An important consequence of invoking a local perception of environmental change is that the predicted routes are loxodromic and not orthodromic. In an ocean without landmasses, ecophysiological processes tended to over-estimate individual weights. With the actual ocean configuration, the excess weight gain was mitigated and also produced increased heterogeneity among the individuals. Developing a model of individual whale dynamics has also highlighted where the understanding of whales’ individual behaviors and population dynamic processes is incomplete. Our new simulation framework is a step toward being able to anticipate migration events and trajectories to minimize negative interactions and could facilitate improved data collection on these movements.