Revised timing and onset location of two isolated substorms observed by Time History of Events and Macroscale Interactions During Substorms (THEMIS)

We report timing analysis on two previously published substorm events captured by the Time History of Events and Macroscale Interactions During Substorms (THEMIS) spacecraft. During the 29 January 2008 0713 UT substorm, the solar wind velocity had a strong southward component corresponding to a 6.5 degrees southward tilt of the magnetotail. Viewed in a rotated system, the magnetic field at the most distant probe, P1 (X-GSM = -29.5 R-E), shows a bipolar magnetic signature interpreted herein as a tailward moving plasmoid; P2 (X-GSM = -18.5 R-E) also observed magnetic signatures indicating tailward motion at onset. P3 (X-GSM = -10.8 R-E) and P4 (X-GSM = -10.6 R-E) captured dipolarization fronts and earthward flows at the same time. After allowing for the more general case of different magnetosonic speeds on the two sides of the reconnection site, timing of the first signatures in space and ground reveals that tail reconnection initiated at similar to 18 R-E down tail, similar to 2 min prior to auroral intensification. Allowing for different magnetosonic speeds on either side of the reconnection site is warranted by the large separation between the inner (10-12 R-E) and outer (25-30 R-E) probe locations and differing ion temperatures and equatorial magnetic fields expected at those locations. The same technique was applied for the 2 February 2008 0740 UT substorm event during which midtail data from P2 were unavailable. A previous study obtained a reconnection site location of X-GSM = -11 similar to-17 R-E assuming the same propagation speed on both sides of the reconnection site. Relaxation of the constant-speed condition results in a reconnection location of similar to 22 R-E and an inferred reconnection time of similar to 3-4 min before the auroral intensification. Our results are consistent with other THEMIS event studies that are unaffected by large solar wind deflections or incomplete probe coverage, suggesting that reconnection triggering of substorm onset is a common occurrence.