We have investigated characteristics of the large-scale traveling atmospheric disturbances (LS-TADs) generated during geomagnetically quiet and disturbed periods using a whole atmosphere general circulation model (GCM). The GCM simulations show that various TADs appear in association with passages of regions with large temperature gradients near the solar terminator, midnight temperature anomaly, and auroral oval which move with the Earth's rotation. These TADs, which are superimposed on each other, appear even when a geomagnetically quiet period. The TADs generated during a geomagnetically quiet period show structures extending in the longitudinal direction at high-latitude and in the latitudinal direction at mid- and low-latitude. These structures disappear after their short-range propagations. The TADs generated during a geomagnetically disturbed period show structures extending widely in the longitudinal direction and propagate from high- to low-latitude. These simulation results suggest the different generation mechanisms and features between the TADs generated during geomagnetically quiet and disturbed periods.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 The large-scale traveling atmospheric disturbances (LS-TADs) are quite important for energy and momentum transfer in the thermosphere/ionosphere. The TADs cause changes in the ionospheric height and density observed as the traveling ionospheric disturbances (TIDs). Since Hines  established the theory of the atmospheric gravity waves (AGWs) in the upper atmosphere, many researchers have investigated the TADs and TIDs as manifestations of AGWs [e.g., Hunsucker, 1982; Prölss, 1995; Hocke and Schlegel, 1996]. Recently, numerical simulations have been performed with the thermosphere/ionosphere general circulation model (GCM) by e.g., Lu et al.  and Lee et al.  to understand the observed storm-time TIDs. In addition, the Global Pointing System (GPS) enables us to investigate TID propagation by monitoring the horizontal distribution of the total electron content (TEC) [e.g., Saito et al., 1998; Shiokawa et al., 2003].
 The basic properties of the TADs/TIDs are, however, as yet imperfectly understood. Many researchers have simulated only the LS-TADs/TIDs generated during a geomagnetic storm/substorm activity [e.g., Millward et al., 1993; Fuller-Rowell et al., 1994; Fujiwara et al., 1996; Balthazor and Moffett, 1999] although the LS-TIDs are sometimes observed during geomagnetically quiet periods [Tsugawa et al., 2004; Hawlitschka, 2006]. Tsugawa et al.  analyzed the TEC data obtained from the GPS observations to investigate statistical features of the LS-TIDs over Japan. They found that 28% of the LS-TIDs were observed during geomagnetically quiet periods (Kp < 3). Hawlitschka  also found the LS-TIDs occurring at daytime even when the magnetosphere was extremely quiet (K < 2) although the LS-TIDs could generally be related to magnetic storms. At nighttime, the LS-TIDs were found more often regardless of the K-index. Galushko et al.  observed TIDs which would be generated by the moving solar terminator. They reported characteristics of the observed TIDs: propagation velocity of 300–400 m/s corresponding to the solar terminator speed at the ionospheric height, period of 1.5–2 hours, and propagation direction perpendicular to the solar terminator.
 In this study, we investigate characteristics of the disturbances in the upper thermosphere with a whole atmosphere GCM which covers all the atmospheric regions from the ground to exobase [Miyoshi and Fujiwara, 2003, 2006]. This GCM describes the thermospheric global circulation with day-to-day variability originated from the lower atmosphere. We focus our attention on difference between the TADs generated during geomagnetically quiet and disturbed periods.
2. Description of the GCM and Numerical Simulations
 In this study, we use a whole atmosphere GCM developed by Miyoshi and Fujiwara  as an extension of the middle atmosphere GCM developed at Kyushu University [Miyahara et al., 1993; Miyoshi, 1999]. The whole atmosphere GCM solves the full nonlinear primitive equations for momentum, thermodynamics, continuity, and hydrostatics. Although Miyoshi and Fujiwara  used the empirical thermospheric composition, the continuity equation of mass mixing ratio for the major species, N2, O2, and O, is solved taking into account the photo-dissociation of O2 and oxygen chemistry in the present version of the GCM.
 This GCM is a global spectral model (triangular truncation of T21) with 75 vertical pressure levels (vertical resolution of 0.4 scale height above the tropopause) and contains all the atmospheric regions from the ground to exobase (∼500 km in the solar minimum and geomagnetically quiet condition). The time step for integrating the equations is 100 s. The displacement of the geomagnetic poles relative to the geodetic poles is considered in the GCM. The effects of auroral particle precipitation on heating the neutral gases are evaluated by using an analytical prescription into the auroral oval [Roble and Ridley, 1987]. The magnetospheric convection electric field modeled by Volland  and empirical ionosphere are used for calculating Joule heating and ion-drag force. The more details of the GCM are described by Miyoshi and Fujiwara .
 In order to investigate the TADs generated during geomagnetically quiet and disturbed periods, we perform two numerical simulations: (1) solar minimum (F10.7 = 70) and geomagnetically quiet (the cross polar cap potential is 30 kV) condition on November 5 (summer in the southern hemisphere and winter in the northern hemisphere), (2) the same as (1) except for geomagnetic condition (GCM simulation starts from the geomagnetically quiet condition with the cross polar cap potential of 30 kV. During the first one hour, the cross polar cap potential is enhanced to 60 kV and is set to 30 kV again after that). Hereafter, we call the two simulations (1) and (2) as case-1 (quiet case) and case-2 (disturbed case), respectively. The present approach for TAD generation is similar to that by Balthazor and Moffett . These approaches would not represent specific TADs of real observed conditions because the empirical high-latitude energy inputs, e.g., the electric field or Joule heating, used in GCM simulations do not include local and rapid fluctuations of the observed ones. As mentioned by Balthazor and Moffett , we also intend to apply known control inputs to generate TADs in order to study their general generations and propagations and have general conclusions, rather than to simulate specific ones.
Figures 1a and 1b show the global distributions of the horizontal neutral wind and temperature on an isobaric surface of about 306 km height at (a) UT = 0100 and (b) UT = 0400 calculated in case-1. The ordinate and abscissa in each panel are the geodetic latitude and longitude (local solar time). The maximum arrows denote the wind velocity of 355 and 309 m/s in Figures 1a and 1b, respectively. In the dayside, high temperature and strong poleward winds blowing toward the nightside are remarkable. Strong equatorward winds are seen in the nightside high-latitude region. In addition to large-scale temperature distribution (day-night temperature difference), there are localized small-scale temperature structures. The temperature differences between the temperatures at UT = 0100, 0400 and the 40-minute (UT = 0040–0120, 0340–0420) averaged ones are shown in Figures 1c and 1d, respectively. Figures 1c and 1d are the same as Figures 1a and 1b, respectively, except for the temperature differences. Small-scale structures are clearly seen both in Figures 1c and 1d. These temperature structures seem to extend mainly in the longitudinal direction at high-latitude and in the latitudinal direction at mid- and low-latitude.
Figures 2a–2d are the same as Figures 1a–1d, respectively, except for calculated results in case-2. The cross polar cap potential is set to be 60 kV in the period of UT = 0000–0100 in this case. The maximum arrows denote the wind velocity of 561 and 362 m/s in Figures 2a and 2b, respectively. The electromagnetic energy input due to the 1-hour enhancement of the cross polar cap potential (or Joule heating) causes strong winds and high temperatures in the high-latitude regions as seen in Figure 2a. The high temperature structures extend widely in the longitudinal direction. These structures propagate to the low-latitude region as the TADs. As a result, high temperature region appears in the low-latitude in Figure 2b. In Figure 2b, high temperature structures are also seen in the polar regions because of the other TADs passing over the polar cap. Figures 2c and 2d show longitudinal extents of temperature structures or TADs superimposed on the small-scale temperature structures as shown in case-1. It is found that several packets of the TADs are produced (Figure 2d) when the TADs generated in the high-latitude region (Figure 2c) propagate to the low-latitude region. The TADs are also affected by the Coriolis force when they propagate.
 The TAD generations and propagations are clearly seen in case-2 as shown above. On the other hand, the TADs in case-1 are not so clear in comparison to those in case-2 because the TADs, which have smaller spatial scales and amplitudes than those in case-2, disappear after their short-range propagations. In order to clarify the TAD generations and propagations during a geomagnetically quiet period, we give an additional example of temperature variation in case-1. Figures 3a and 3b present global temperature distributions which are the same as Figures 1a and 1b except for temperatures at UT = 1000 and 1100, respectively. There are small-scale temperature structures superimposed on the large-scale day-night temperature distribution. In particular, wave-like temperature structures are remarkable in the region from the evening to night and in the high temperature region in the southern high- to mid-latitude. These temperature structures seem to appear in association with the temperature gradients near the solar terminator, midnight temperature anomaly, and auroral oval.
Figures 4a and 4b present the temperature differences which are the same as those shown in Figures 1c and 1d except for UT = 1000 and 1100, respectively. The white and black lines in Figures 4a and 4b denote temperature structures near the solar terminator. These structures move with the Earth's rotation. There are also some temperature structures aligned along and moving with the solar terminator. Figures 1a, 1b, 2a, and 2b are the snapshots of animations that are available through the auxiliary material associated with this paper (see 2006GL027103-ms1.avi and 2006GL027103-ms2.avi). The generations and propagations of the TADs can be viewed in these animations.
 The TAD features mentioned above are also seen in the meridional wind and composition distributions. Figure 5 is the same as Figure 4b except for differences in (Figure 5a) meridional wind and (Figure 5b) O/N2 ratio at UT = 1100. There are some structures similar to temperature difference: namely, structures extending in the longitudinal direction at high-latitude and in the latitudinal direction at mid- and low-latitude. The white and black lines denote structures near and aligned along the solar terminator.
4. Summary and Discussion
 The generations and propagations of the TADs during geomagnetically quiet and disturbed periods are simulated with a GCM which covers all the atmospheric regions from the ground to exobase. This GCM enables us to consider effects of the lower atmosphere on the thermosphere without any boundary conditions in the atmospheric regions. The rather small-scale temperature structures appeared in the upper thermosphere should be affected by the lower atmospheric variability although the lower atmospheric effects are largely damped by molecular diffusion, thermal conductivity and ion drag. The more smoothed temperature distributions are calculated when we artificially suppress the lower atmospheric effects by setting winds of 0 m/s and the time-independent globally uniform temperature profile below about 70 km altitude in the GCM (not shown in this study). This suggests the importance of coupling between the lower and upper atmospheres for generating disturbances in the thermosphere during geomagnetically quiet periods.
 In this study, the TAD generations and propagations are seen near the regions with the localized temperature structures (temperature gradients) near the solar terminator, midnight temperature anomaly, and auroral oval. These TADs have something in common with the TIDs presented by Galushko et al. : spatial scales, generations near the solar terminator, propagation directions almost perpendicular to the solar terminator, and propagations with the Earth's rotation.
 The simulated TADs during a geomagnetically disturbed period show good agreements with those presented by previous studies, e.g., Balthazor and Moffett : TAD propagations from high- to low-latitude with propagation velocity of about 700 m/s, TADs passing over the polar cap, and TAD propagations affected by the Coriolis force.
 The TADs simulated in a geomagnetically quiet condition have smaller spatial scales and amplitudes than those simulated in a geomagnetically disturbed condition. In addition, the TADs in a geomagnetically quiet condition disappear after their short-range propagations, while the TADs generated in a geomagnetically disturbed condition propagate from high- to low-latitude. These features of the TADs simulated in both the conditions would explain qualitatively the observational probabilities of the TIDs shown by Tsugawa et al.  and Hawlitschka .
 We have reported the first results from GCM simulations which show generations of the TADs even during a geomagnetically quiet period. In order to make comparisons between the observed TADs/TIDs and simulated ones quantitatively, we should have more GCM simulations in specific cases as the next step. Many of previous studies [e.g., Prölss, 1993; Bauske and Prölss, 1997; Lu et al., 2001; Lee et al., 2004] have shown changes in the ionospheric height, meridional wind, and composition in association with TAD/TID propagations. We will also investigate details of changes in various parameters to understand TAD/TID effects on the thermosphere/ionosphere. In addition, the lower atmospheric effects on generation of disturbances in the thermosphere will be described in future works.
 This work was supported in part by Grant-in-Aid for Scientific Research and the 21st Century COE program “Advanced Science and Technology Center for the Dynamic Earth” by the Ministry of Education, Science, Sports and Culture, Japan, and the joint research program of the Solar-Terrestrial Environment Laboratory, Nagoya University.