Extreme precipitation (>95%) events have major societal impacts, therefore it is important to understand the ways in which these events may change in the future. Learning which processes are most important to creating extreme precipitation is integral to this understanding because a) the distribution of processes in model historical simulations can be assessed by comparison to observed distributions and b) the distributions of processes may be altered by climate change. Of particular interest are meteorological processes which produce upward motions because these are a necessity to create extreme precipitation. The processes we consider are represented by the forcing terms and lower boundary condition of the full omega equation, and include: vorticity advection, thermal advection, convection, local changes in time, and orographic ascent. We invert the omega equation to find the relative contributions to vertical motion of these processes for each event in the record of interest. Events often spring from a mixture of these processes, so the multidimensional distribution of contributions is analyzed for different areas and seasons in reanalysis data. These distributions in the current climate can be used to provide a baseline “truth” to compare with future simulations using different climate forcings. Individual events are aggregated over regions based on the similarity of the annual cycle of precipitation using an artificial neural network defined by Swenson and Grotjahn (2019). This allows us to discover the seasonality of the driving dynamical processes causing the extreme rainfall in a location. The analysis provides a natural link between the climate of a particular area and the types of weather that create extreme precipitation within the area.