Our growing capacity to gather and process continental water coverage data based on remote sensing has opened new possibilities to understand the spatial and temporal dynamics of floods. Here we used a 30-year spanning dataset (Global Surface Water Extent) to elaborate a worldwide geographical characterization of floods (1-degree grid), and weighted the relative contribution of seasonal, interannual, and long-term fluctuations on overall variability, and quantified precipitation-flooding delays where the seasonal component was dominant. We explored the distribution of flooding timings in relation to climate, represented by five main Köppen-Geiger classes, and hydro-topography, represented by seven classes derived from modeled water table depths. Our results showed that, globally, the mean extent of floods averaged 0.48% of the land area and was predominantly associated with hydro-topography (> 2 times higher in cells dominated by flat bottomlands) and secondarily to climate (less than half in arid climate cells). Seasonal (interannual) variability decreased (increased) from boreal to tropical to temperate and arid climates. Predominantly positive, long-term trends dominated temporal variability in 14% of the grid cells. We hypothesize that climate change may have a deeper impact on the temporal variability of floods and its components while land use and water regulation infrastructure would be more important influencing their extent. Remote sensing will allow a continuous update of the geography of floods while the variables used to describe them here may prove sensitive to distinct dimensions of global change.