1. Introduction

[2]  Neutral winds and electric fields dominate the dynamics in the E region where backscatter from the E -region field-aligned irregularities is observed. In general, ionospheric parameter measuring techniques can be divided into two groups: remote and direct. Electric fields are measured in situ by instruments which are mounted on rockets, balloons or satellites using a double-probe technique [Kelley, 1989; Schunk and Nagy, 2000].

[3]  Neutral winds are measured both in situ and remotely. In the case of direct observations, a chemical release is used as a tracer for measuring neutral winds in the mesosphere and lower thermosphere. The chemical release technique provides wind data for the altitude range covering the entire E region with a high altitude resolution of about 100 m or so. A detailed review of the winds measured by the chemical release technique over more than 40 years is given by Larsen [2002].

[4]  Remote diagnostics of ionospheric winds are mainly based on Doppler measurements and have been accomplished with different types of radar: meteor, mesosphere-stratosphere-troposphere (MST) and MF and optical interferometers [see, e.g., Hocking, 1997; Hocking et al., 2001; Jacka, 1984; Roper, 1984; Rottger, 1984, and references therein]. Neutral winds were also measured by the wind imaging interferometer on board NASA's Upper Atmosphere Research Satellite [Shepherd et al., 1993].

[5]  All these techniques are characterized by their cost, availability at the location of interest, time and altitude resolution and altitudes covered. In situ observations require special efforts and funding. They are expensive and relatively rare. Nevertheless, several experimental campaigns expressly aimed at studying ionospheric processes in the E region (sporadic E layers, in particular) and involving rocket measurements of winds and electric fields were carried out. Among those recently organized are the SEEK I [see Fukao et al., 1998; Larsen et al., 1998; Pfaff et al., 1998, and references therein] and SEEK II campaigns in Japan.

[6]  The remote techniques for measuring winds suffer from a limitation on the altitude range (they work mainly below 100 km altitude). Also the altitude resolution is not as good as that of the chemical release technique. For meteor radars the altitude and time resolutions depend on the number of meteors; for example, in the case study presented below, they were about 2 km and 2 hours, respectively.

[7]  It should be mentioned that for cases in which backscatter from field-aligned irregularities is observed from altitudes either well below or well above 100 km, the measured Doppler velocities may be attributed correspondingly to either a neutral wind or electric field alone [Kagan, 2002]. On the basis of this, Murthy et al. [1998] extracted meridional winds from the backscatter data below 97 km altitude measured by the Gadanki radar in India. Patra [2002] applied the same idea to deriving zonal winds at 90-97 km from the backscatter observed by Tsunoda and Ecklund [1999] at Pohnpei.

[8]  In this paper we propose an experimental method and a data processing procedure which allow the altitude-time reconstruction of background and polarization electric fields and neutral winds in the backscatter regions from the altitude-time distribution of line-of-sight Doppler velocities observed with a coherent scatter radar (CSR). This method does not require any additional expense. The altitude and time resolution of our method are equal to twice the altitude resolution and equal to the temporal resolution of the radar measurements, respectively.

[9]  To this end we have to solve two basic problems. First, we have to reconsider the theories of 3-m irregularity generation (most coherent scatter radars observe at frequencies close to 50 MHz; for example, the MU radar frequency is 46.5 MHz) to find a more correct expression for the phase velocity which takes into account neutral gas motions and ion magnetization (section 2). Second, we have to develop an experimental scheme and a data processing procedure to extract the desired information from observations by making use of the above theory (section 3). We show the efficiency and limits of our method using the middle and upper (MU) radar as a basic instrument in section 4. We discuss the validity and further development of the method in section 5.