1. Introduction

[2]  This paper presents the measurements of the temperature and density of the neutral atmosphere within the height interval 100-113 km. Measurements of the temperature and density at these heights present a hard task. This is due to the limitations of the methods existing measurement in the lower thermosphere. The most adequate by the authors' opinion are rocket measurements, in particular, the falling spheres method which makes it possible to measure the temperature and density in situ [Lübken, 1999; Lübken et al., 1990]. This method was used to conduct measurements in high-latitude regions. During the recent decades several large cycles of lidar measurements were carried out [Collins et al., 1996; Jenkins et al., 1987; Kurzawa and von Zahn, 1990; Senft and Gardner, 1991; She et al., 2000]. Using lidars, a vast set of experimental data on the temperature, density, and wind velocity of the neutral atmosphere in the mesopause region has been obtained. However, lidar measurements are possible only in the dark period of a day and are strongly depend on meteorological conditions. The accuracy of measurements by this method decreases above 85-90 km.

[3]  The incoherent scatter method also was used for measurements of atmospheric parameters in the lower thermosphere [Kirkwood, 1986]. It was found that for the height interval 90-120 km this method is less effective than for the upper atmosphere because the measurement error increases sharply at T_e = T_i [Buonosanto, 1989].

[4]  The conducted measurements gave the following results. First, in the vicinity of the mesopause and especially above it, there exists a scatter of the values of the atmospheric parameters caused by wave activity: by propagation of internal and tidal waves. Second, many data (in particular, the neutral temperature profiles obtained by rockets and incoherent scatter method) in the above indicated height interval provide lower values than the commonly accepted atmospheric models (e.g., CIRA 86). Moreover, some studies demonstrate a two-level structure of the temperature at the mesopause [Mertens et al., 2004; She et al., 2000]. Also, finally, there is a deficit of the experimental data on atmospheric parameters at altitudes h 100 km.

[5]  The method of measurements of the temperature T and density r of the neutral atmosphere using the Bragg backscattering of radio waves at artificial periodic irregularities (API) was proposed by Benediktov et al. [1993]. It belongs to the group of methods of the ionospheric plasma and atmosphere diagnostics developed while studying the impact of a powerful high-frequency radio wave on the ionosphere [Belikovich et al., 1997, 2002b]. As a result of the interference of the radio waves falling on the ionosphere and reflected from it, a standing wave is formed at such impact. A periodic structure of the electron concentration perturbations is formed in the field of the latter wave. API are formed in the height interval from ~60 km up to the point of the powerful radio wave reflection from the ionosphere. When an API is sounded by short radio pulses, there occurs a Bragg backscattering from the altitude region where the lengths of the perturbing and testing waves are equal to each other. Registering the amplitudes and phases of the signals scattered at API, one obtains information on various parameters of the ionospheric plasma and atmosphere.

[6]  In the height interval 90-130 km, API are formed because of the diffusion redistribution of the heated plasma, and their relaxation occurs in the process of ambipolar diffusion. The relaxation time of API after switching off the powerful radioemission at these altitudes depends on the wave number of the standing wave and the ambipolar diffusion coefficient. The latter in turn is governed by the electron T_e and ion T_i temperatures and the collision frequency of ions with neutrals n_im, the frequency of the ion-molecule collisions being proportional to the atmospheric density. It is natural to assume that at 90-120 km there is a local isothermy, i.e., T_e = T_i = T. At the above indicated assumptions that are correct for the height interval in question, the temperature and density of the atmosphere can be determined from the vertical profile of the API relaxation time. It should be noted that the latter statement is not always correct for high latitudes where an increase of T_e often takes place due to energetic electrons precipitation.

[7]  A series of measurements of T and r was conducted in 1990-1991. A series of publications has been dedicated to the main results of these measurements. They were presented most completely by Belikovich et al. [2002a]. The dependence of the neutral atmosphere temperature on height was determined in the height interval 99-110 km in the morning hours for the fall season of 1990. A temperature minimum of ~140 K at a height of ~104 km, an increase up to 230-250 K below 100 km, and T 160 K at a height of 108 km were detected. A strong variability of the atmospheric parameters in individual measurements at altitudes of 95-110 km was detected, this fact being confirmed by the results of the temperature measurements in this height interval by other methods [Jenkins et al., 1987; Kirkwood, 1986; Kurzawa and von Zahn, 1990]. The time variations in the temperature with D T/T (0.05-0.25) are often registered, the latter value exceeding random errors of the measurements. The periods of the above mentioned variations lie in the interval from 15-30 min to a few hours. The common explanation of these variations is variations in atmospheric parameters due to propagation of internal gravity waves (IGW) and tidal waves through the observed region. The detailed description of the IGW characteristics obtained on the basis of the measurements of the vertical motion velocities using API is given by Belikovich et al. [2002b].

[8]  The joint analysis of the vertical motion velocity, temperature, and density of the neutral atmosphere at heights of 80-110 km based on the results of 1990-1991 measurements showed that these parameters simultaneously present oscillations with the same periods from 5-10 min to a few hours. A numerical simulation of acoustic gravity wave characteristics was performed using the linear theory of their free propagation in a wall-less isothermal undisturbed atmosphere [Bakhmet'eva et al., 2002]. It was shown that the relative amplitudes of the variations in the temperature and density calculated on the basis of the polarization relations for low-frequency waves using the measured amplitudes of the vertical velocity agree satisfactorily to the measurements only for the waves with periods of 15-30 min. Considering more long-period waves, one has evidently to take into account the medium motion and nonisothermy of the undisturbed atmosphere.

[9]  Using the data of the fall-winter seasons of 1990-1991, the influence of geomagnetic activity on the temperature and density of the neutral atmosphere at heights of 90-110 km was also considered. According to the results obtained by the incoherent scatter method at the EISCAT installation and analyzed by Kirkwood [1986], the dependence of the atmospheric parameters in the height interval 90-120 km on solar and geomagnetic activity is weak. The measurements near Nizhny Novgorod using API also found no significant dependence of T and r on the value of the 3-hour planetary Kp index. However, there exists an influence of magnetic activity on the atmosphere, though it has an indirect character [Belikovich et al., 2002a]. It is manifested in an intensification of the atmospheric wave activity in the given height interval in the periods of geomagnetic disturbances.

[10]  In this paper, new results of measurements of the neutral atmosphere parameters obtained during the summer months of 1999-2002 are presented. The main attention was paid to the measurements in the sunrise and sunset periods and also during the solar eclipse on 11 August 1999.