Floodplain plays an important role in modulating the magnitude and movements of river flow through inundation and flood events. However, floodplain dynamics is under-represented in most current large-scale river routing models and the spatiotemporal characteristics of flood inundation events are not well understood. In this study we applied a physically based inundation model coupled with a river routing model (Model for Scale Adaptive River Transport, MOSART) within the Energy Exascale Earth System Model (E3SM) framework to investigate the global spatiotemporal characteristics of flood inundation dynamics. The model features a DEM-based floodplain storage connecting to the river channel and a diffusion wave routing scheme to represent the back water effect. After calibrating against the satellite-based flood inundation extent data, river stage data and observed river discharge, the model reasonably captures the historical flood events in terms of the frequency and the spatial extent in 15 global river basins. We applied the model globally for simulations of 15 years long (1993-2007) and analyzed the spatial patterns and the temporal characteristics of the flood events in major global basins. We also explored the linkages between the flood inundation events and the climate as well as local water conditions such as snow storage and soil moisture, with the goals of improving prediction of future flood risk, providing guidance on flood control, and understanding the impacts of climate change.