NASA Goddard Space Flight Center, Laboratory for Extraterrestrial Physics, Greenbelt, Maryland 20771

Results from a correlation analysis of Wind and IMP 8 magnetometer data for a total of 543 two hour periods covering Jan-Jul 1995 were analyzed to determine that: (1) the timing accuracy D t for advecting solar wind magnetic field structures goes as (d_per/d_par) x t _con where d_par is the spacecraft separation along the Sun-Earth line, d_per is the transverse separation, and t _con is the predicted convection lag time, (2) "good" correlation time periods (peak correlation coefficient > 0.80) are about twice as likely at solar maximum than at solar minimum, (3) geometry affects timing accuracy more than any propagation of features with respect to the solar wind, and (4) there is a "high" probability (greater than predicted by a Gaussian distribution) of "very bad" (|d_par D t /d_pert _con| >3) timing agreement due to the long tail on the probability distribution.

Geo. Res. Lett., Vol. 25, No. 14, 2509, 1998