The idea of expedient energy transformation by magnetic reconnection (MR) has generated much enthusiasm in the space plasma community. The early concept of MR, which was envisioned for the solar flare phenomenon in a simple two-dimensional (2D) steady-state situation, is in dire need for extension to encompass three-dimensional (3D) non-steady-state phenomena prevalent in space plasmas in nature like in the magnetosphere. A workshop was organized to address this and related critical issues on MR. The essential outcome of this workshop is summarized in this review. After a brief evaluation on the pros and cons of existing definitions of MR, we propose essentially a working definition that can be used to identify MR in transient and spatially localized phenomena. The word "essentially" reflects a slight diversity in the opinion on how transient and localized 3D MR process might be defined. MR is defined here as a process with the following characteristics: (1) there is a plasma bulk flow across a boundary separating regions with topologically different magnetic field lines if projected on the plane of MR, thereby converting magnetic energy into kinetic particle energy, (2) there can be an out-of-the-plane magnetic field component (the so-called guide field) present such that the reconnected magnetic flux tubes are twisted to form flux ropes, and (3) the region exhibiting non-ideal MHD conditions should be localized to a scale comparable to the ion inertial length in the direction of the plasma inflow velocity. This definition captures the most important 3D aspects and preserves many essential characteristics of the 2D case. It may be considered as the first step in the generalization of the traditional 2D concept. As a demonstration on the utility of this definition, we apply it to identify MR associated with plasma phenomena in the dayside magnetopause and nightside magnetotail of the Earth's magnetosphere. How MR may be distinguished from other competing mechanisms for these magnetospheric phenomena are then discussed.
Bibliographical noteFunding Information:
This work is supported in part by National Science Foundation Grant ATM-0135667 to the Johns Hopkins University Applied Physics Laboratory, in part by the NSFC Project 40390150 and 973 Project G200000784 to the Peking University, in part by the Belgian Federal Public Planning Service through ESA (PRODEX/Cluster), and in part by NASA Grant NAG5-11868 and NSF Grant ATM-0201703 to the University of Minnesota. We thank W. J. Heikkila for his initial participation on this project, G. Paschmann, M. Scholer, C. Jaroschek, R, Nakamura, A. Runov, and R. L. Lysak for valuable discussion.