[1] The authors welcome the comment from J. E. Mazur and T. P. O'Brien [Mazur and O'Brien, 2012] on our recently published study [Choi et al., 2011]. In our paper, we investigated the geostationary Earth orbit (GEO) satellite anomalies archived by Satellite News Digest (SND) during 1997-2009 in order to search for possible influences of space weather on the anomaly occurrences. There were good relationships between geomagnetic activity (as measured by the Kp index) and anomaly occurrences of the GEO satellites; the satellite anomalies occurred mainly in the midnight-to-morning sector, and the anomalies were found more frequently in spring and fall than in summer and winter. A comparison of the SND data with data from Los Alamos National Laboratory satellites showed that low-energy (<100 keV) electrons exhibit behavior similar to that of spacecraft anomalies and implied that the spacecraft charging may be a primary contributor to the GEO spacecraft anomalies reported on the SND Web site (http://www.sat-index.co.uk). [2] Mazur and O'Brien [2012] point out that some anomalies used in our analysis were not obviously caused by space weather effects. In fact, we intentionally used all of the GEO satellite anomalies listed on the SND Web site in order to exclude a subjective selection effect. Our event list, chosen from the SND database, represented major satellite anomalies that had significant financial impacts. There were numerous minor satellite anomalies reported from many agencies, and we want to emphasize that their tendency was similar to that of our event list. While the number of events may not be large enough to analyze local time dependence, when the anomalies reported by SND from 2010 to 2011 (indicated in Figure 1 by star symbols) are included, it is very clear that the anomaly occurrences are more frequent at nighttime than during the daytime. [3] Mazur and O'Brien [2012] also mention the relationship between GEO satellite anomalies and charging effects. Internal charging may be concerned with high-energy electrons and independent of its local time, while external charging is related to low-energy electrons and dependent on its local time. As we noted in Choi et al. [2011], the flux of 100 keV electrons on GEO orbit shows nonuniform distribution on the local time, yet these electrons don't have enough energy to penetrate satellite walls and charge internal components. At this moment, we don't fully understand the mechanism by which charged particles bring about the anomalies. [4] We support the proposal made by Mazur and O'Brien [2012] that an agency be established to maintain adequate and open anomaly and abnormality lists containing all information about events. We also anticipate that our paper will serve as encouragement to all the agencies concerned to make their anomaly data publicly available and to investigate an occurrence mechanism of spacecraft anomaly. References

[1] We analyze the dynamics of the high-latitude thermsopheric wind system below 170 km for negative IMF B z by using a fully nonlinear model with a realistic distribution of the forcing. A transition of the forcing patterns and their relative contribution to the high-latitude lower thermospheric wind system occurs around 123 km under various conditions, weak or strong IMF, summer or winter. Winds around and above 123 km are sustained by the gradient-wind balance among divergent/convergent pressure gradient, Coriolis, and horizontal momentum advection (mainly centrifugal) accelerations. Below 123 km winds are maintained by the approximate balance of divergent/convergent pressure gradient, Coriolis, and Hall ion drag accelerations through modified geostrophy. The dominant contribution to the wind tendency (time rate of change) is the rotational component of the ion drag acceleration. The wind tendency above 123 km tends to resemble rotational Pedersen ion drag acceleration well, which reflects a rotated pattern of the E Â B velocity. Near and below 123 km the wind tendency is also affected by the rotational component of the Hall ion drag acceleration whose pattern no longer closely resembles the pattern of the E Â B velocity, and the wind pattern can differ significantly from that well above 123 km. Simulations for different strengths of the IMF and different seasons indicate that largely divergent/convergent Coriolis and horizontal momentum advection accelerations tend approximately to balance with the horizontal pressure gradient (as well as with divergent/convergent ion drag at lower altitudes) under various conditions. As the forcing increases the radius of curvature of the strong winds also tends to increase, so that the centrifugal acceleration does not increase quadratically with the maximum wind speed, and the tendency for a rough balance between the Coriolis and horizontal momentum advection accelerations in the duskside vortex above 123 km is maintained.

[1] We analyze the forces acting on the high-latitude lower thermospheric wind system below 170 km for Southern Hemisphere summer conditions, as a function of the interplanetary magnetic field (IMF) direction, on the basis of numerical simulations. The pattern and magnitude of the forces and their relative contributions to the wind system vary strongly with respect to the direction of the IMF. At higher altitudes, above 130 km, for negative B y , strong anticyclonic winds are accelerated primarily by rotational Pedersen ion drag and are maintained by an approximate balance among the divergent/convergent Coriolis, horizontal advection, and relatively weak pressure-gradient accelerations. For positive B y , the pressure-gradient acceleration is increased, while the inertial forces are reduced. For negative B z , in comparison with negative and positive B y , the winds and forces extend to lower latitudes. The patterns of the accelerations for positive B z are similar to those for negative B z , but the magnitudes tend to be significantly smaller. At lower altitudes, below 120 km, the horizontal advection acceleration is less important but still contributes significantly to the maintenance of the neutral circulation in the polar cap region for positive B y . The difference of winds and forces above 130 km for negative and positive B y , with respect to winds and forces for zero IMF, show a simple structure with a strong anticyclonic or cyclonic vortex near the pole, respectively, centered differently for the two B y directions. The difference of winds and forces for negative and positive B z are more complex than those for negative and positive B y and extend to lower latitudes. Below 120 km, the difference of winds and forces for negative and positive B y are much stronger near the pole than for negative and positive B z , indicating that the IMF B y component tends to dominate effects on the neutral winds in the polar cap at low thermospheric altitudes. For all IMF conditions, at higher altitudes, the rotational ion-drag acceleration makes the dominant contribution to the neutral velocity tendency. This feature is most pronounced when the IMF B z is negative.