Physics of Coronal Mass Ejections: Recent Theoretical and Observational Advances James Chen Plasma Physics Division Naval Research Laboratory Coronal mass ejections (CMEs) are now understood to be the solar driver of large geomagnetic storms and other geospace disturbances. Since their discovery in the early 1970's, theoretical and modeling ideas have risen and fallen. With new ideas and unprecedented observations by SOHO, especially LASCO and EIT instruments, several aspects of theoretical and observational understanding of the CME phenomenon are converging. In particular, the magnetic flux-rope configuration has emerged as a key element of the new understanding. In this talk, I will discuss some recent theoretical advances, focusing on the 3D magnetic geometry underlying CMEs, which determines the Lorentz force acting on the magnetic structure, and the near-Sun acceleration of CMEs and associated prominences. While coronal magnetic fields cannot be measured at the present time, acceleration can be obtained from data. Because acceleration is a direct manifestation of the driving force, much physics pertaining to the magnetic geometry and driving mechanism can be extracted from the data. A recent finding is that the main acceleration phase of CMEs and eruptive prominences (EPs) exhibit a scaling behavior characteristic of expanding flux ropes with stationary footpoints: the height where maximum acceleration is attained is in the range of Sf/2 to 1.5 Sf, where Sf is the separation distance between the two footpoints of a flux rope. This theoretical prediction has been tested using a set of EPs and CMEs and found to be in good agreement with data. The fact that the observed events exhibit this scaling property implies that pre-eruption structure is a flux rope or a flux rope is formed early enough in the eruption process. Calculations of flux-rope propagation to 1 AU and beyond show that erupting flux ropes consistent with CMEs observed near the Sun do evolve into structures closely resembling observed magnetic clouds. The talk will present a unified picture of CMEs and 1 AU consequences with important validations based on near-Sun and 1 AU solar wind data. Work supported by ONR and NASA.