[1] Daytime longitudinal structures of the electron density (N e ) and temperature (T e ) in the topside ionosphere observed by Hinotori and DEMETER are examined under various conditions of solar flux, local time, and seasons. Results from both satellites show a similar longitudinal N e structure in the morning from July to October, although the value of N e observed by Hinotori is higher than that of DEMETER owing to higher solar flux. This result implies that the longitudinal structure of N e may appear in any solar cycle. Further, a negative correlation between N e and T e in the longitudinal structures appears in the morning when N e is low, while a positive correlation appears around the magnetic equator when N e is sufficiently enhanced during noontime in the high solar flux. A spectrum analysis performed on the DEMETER data reveals that wave numbers 1-2 for N e and T e are dominant and nondominant. The observed wave numbers 3-4 for N e are dominant during November-May and June-October, while they are dominant for T e during October-June and July-September. Both N e and T e show the largest power of wave number 3 in December and wave number 4 in September. Further, observed annual variations of wave numbers 3-4 for N e and T e also differ from wave numbers 3-4 generated by waves in the lower thermosphere. It can be interpreted as discrepancies between the longitudinal distributions of N e and T e caused by difference in the condition of zonal winds driving E region dynamo and meridional winds modulating the ionospheric plasma structures.

Measurements of strain accumulation Hollister and Parkfield lies on a small circle

On 19 December 2009, the eastern part of Taiwan at approximate depth of 45 km was struck by a strong earthquake (Mw = 6.4). The epicenter was located about 20 km away from the Hualien (HLN) station in Taiwan. By analyzing data from the HLN station and Amami Oshima (AMA) in Japan as a remote reference, the geomagnetic east-west (D) and vertical (Z) components recorded at the HLN station are showing no correlation with those at the AMA station in December 2009. Anomalous variations of ∼10-15 nT started about one week before the occurrence of the earthquake and lasted for about two weeks. Also, an enhanced ULF signal in the range of Pc 3 (10-40 s) was observed a few days before the onset of the earthquake. Moreover, the polarization ratio (Z/H (north-south component)) of the Pc 3 amplitude at the HLN station decreased a few days preceding the earthquake. The mechanism behind these anomalous variations is not fully understood. However, we expect that crustal stress perturbations and underground conductivity changes associated with the earthquake played an important role for generating such observed geomagnetic variations.