Authors: Chip Manchester (University of Michigan), Igor Sokolov (University of Michigan), Gergely Koban (University of Michigan), Nishtha Sachdeva (University of Michigan), Lulu Zhao (University of Michigan)

We present results of a continuously evolving AWSoM-R model of the corona, in which the inner boundary data is updated at a one hour cadence by interpolating in time near-real-time GONG synoptic magnetic maps. The maps are provided by NSO with a cadence of 1 hour and data latency (from time of observation) of only 1-2 hours. The time-accurate AWSoM model is the central component of our recently developed SOlar‐wind with FIeld‐lines and Energetic‐particles (SOFIE) model, which simulates the acceleration and transport of solar energetic particles (SEPs) making continuous forecasts of energetic proton fluxes at Earth and throughout the heliosphere. The time-evolving corona improves the forecast accuracy compared to our previously used steady state coronal model, which employed a local time stepping technique to accelerate convergence. In addition, the model spontaneously produces dynamical features and reconnection events that do not appear in the time-stationary model. In this talk, we discuss the improvements offered by the model as well as the nature of evolving coronal structures.