A probabilistic Kp and Hp driven auroral boundary model using 28 years
of DMSP data
Abstract
This paper uses energy flux data from the DMSP F6 thru F19 satellites to
construct a new equatorward auroral boundary model. This method of
processing the data allows one to easily extend boundary detections to
different instrument generations because it is based on a standard
deviation value instead of hard thresholds. Using this method, the paper
provides statistics from just under 1 million auroral boundaries between
1986 and 2014. These statistics are based on normal distribution fits
for each Kp/Hp and MLT bin, which allows one to specify the boundary for
an arbitrary distribution percentile. The paper compares how the model
performs for driving by Hp30, Hp60, and Kp and finds that Hp30 is the
best representation of the data. Finally, the paper provides
specifications to recreate the model for the 50th and 95th percentiles
for Kp and Hp30.