Global observations of injection region evolution during a reconnection initiated substorm

Recent studies have shown that dispersionless injections can be identified in riometer data [Spanswick et al., 2007]. This raises the possibility of using a network of riometers to identify the time when and location (at least as projected in the ionosphere) where an injection began, and of following its subsequent spatio- temporal evolution. In this paper, we present a detailed picture of the evolution of a large substorm injection. The event occurred on August 27, 2001, and has been discussed in numerous publications. The expansion phase in question followed several smaller activations that would reasonably be described as pseudobreakups. Baker et al. [2002] used Cluster in situ and IMAGE FUV observations to argue that near- Earth reconnection preceded the auroral onset. Blake et al. [2005] used Cluster, Polar, and Chandra data to investigate the large-scale expansion of the injection region, subject to the caveat that the injection was well underway when it expanded over the satellites. In this paper, we present riometer and in situ observations of an injection seen at more than 20 separate locations (including 6 satellites: Cluster, Polar, Chandra, and 3 LANL spacecraft). The injection is observed to be dispersionless at 12 of these locations (three of them are observations previously reported by Blake). We are able to identify the time and location at which the injection begins. Combining these observations with information from the GOES-8 geosynchronous satellite we argue that the injection initiated near-geosynchronous orbit and expanded poleward (tailward) and equatorward (earthward) from its beginning. Further, the injection began several minutes after the reconnection identified in the Cluster data, providing concrete evidence that near-Earth reconnection, in at least some events, has little if any ionospheric signature.