A technique to measure the magnetotail reconnection rate from the grou
nd is described and applied to 71 hours of measurements from 20 nights
. The reconnection rate is obtained from the ionospheric flow across t
he polar cap boundary in the frame of reference of the boundary, measu
red by the Sondrestrom incoherent scatter radar. For our measurements,
the polar cap boundary is located using 6300 Angstrom auroral emissio
ns and E region electron density. The average experimental uncertainty
of the reconnection rate measurement is 11.6 mV m(-1) in the ionosphe
ric electric field. By using a large data set, we obtain the dependenc
e of the reconnection rate on magnetic local time, the interplanetary
magnetic field, and substorm activity, with much higher accuracy. We f
ind that two thirds of the average polar cap potential drop occurs ove
r the 4-hour segment of the separatrix centered on 2330 MLT, that the
linear correlation between the reconnection electric field and the hal
f-wave rectified dawn-dusk solar wind electric field VB, peaks between
1.0 and 1.5 hours, with a maximum linear correlation coefficient of 0
.46 at 70 min; and that following substorm expansion phase onset, the
reconnection electric field becomes larger than the experimental uncer
tainty, with an average delay of 23 min. The 70-min delay of the recon
nection rate with respect to VB, is a typical convection time for a fl
ux tube across the polar cap. This result indicates that reconnection
in the magnetotail is influenced by the solar wind electric field VB,
on the field line being reconnected.