[6] The ionosphere investigation by the Kharkov IS radar is
based on the measurements of the signal correlation function (CF).
Measurement and data processing methods were described by
* Taran* [1979, 2001],
* Emel'yanov* [1999],
* Lysenko* [1999a, 1999b],
and
* Pulyaev* [1999].
From the measured CFs, electron
*T*_{e} and ion
*T*_{i} temperatures, ion composition, vertical component
*V*_{z} of
the plasma drift velocity and other ionospheric parameters are
derived. The electron density profiles
*N*_{e}(*h*) are obtained using the
power profile method from the following formula
[*Evans*, 1969]:

(1) |

[7] The errors in estimation of the signal CFs and ionospheric parameters depend on the signal-to-noise ratio, noise background, parameters of the equipment, and other factors. Table 1 shows the statistical errors of ionospheric parameters for typical daytime conditions, 15-min signal integration, and the given signal-to-noise ratios.

[8] The vertical plasma velocity
*V*_{z} is found from the Doppler
shift of the IS signal spectrum estimated based on the measured
quadrature components of the signal CF
[*Emel'yanov*, 1999].
To
increase the accuracy of the measurements, a trapezoidal
smoothing of CF over altitude is performed
[*Holt et al.*, 1992;
* Lysenko*, 1999a].
The RMS deviation
*s*_{Vz} of the measured velocity
depends on the signal-to-noise ratio
*q* and varies with altitude.
Usually,
*s*_{Vz} 5-20 m s^{-1}
for altitudes of the ionospheric
*F* region
at
*q* 0.2 and an integration time of 15 min.

[9] The method used to determine
*T*_{i} and
*T*_{e} temperatures
should be considered in more detail. These temperatures are
calculated taking into account the ion composition at altitudes
below the
*F*2 layer peak. In this case the
*T*_{e}/*T*_{i} and
*T*_{i}/*m*_{i} ratios,
where
*m*_{i} is ion mass, are found from the measured CFs of a
scattered signal
[*Farley*, 1969]
by comparing these functions with
the theoretical CFs using the least squares technique. Certain
conditions are imposed in order to eliminate the ambiguity in the
solution of the problem
[*Pavlov et al.*, 1999;
* Schlesier and Buonsanto*, 1999].
The average molecular weight of ions
(O_{2}^{+} and
NO^{+} ) was taken equal to 31. Gradual transition from the 100%
concentration of molecular ions at 120 km altitude (where it was
considered that
*T*_{e} *T*_{i} *T*_{n} 355 K) to the 100% concentration of
O^{+} ions at an altitude of 230-300 km
was assumed. This height was
selected depending on specific conditions: day-night, winter-summer.
A change in
*T*_{i} within an altitude interval of 10 km is
restricted additionally:
*D**T*_{i}( max) = 0.1 *T*_{i}. It should be noted that
the applied technique only approximately reflects changes in the
concentration of molecular ions, which are especially significant
during magnetic disturbances, and results in additional error in
determining
*T*_{i},
*T*_{e}, and
*N*_{e}. The problem of correcting measured
ionospheric parameters
*T*_{i},
*T*_{e}, and
*N*_{e} depending on the applied
model of ion composition was first discussed by
* Waldteufel* [1971].
It is known that this problem is solved in the modern models of the
ionosphere [see, e.g.,
* Mikhailov and Schlegel*, 1997;
* Schlesier and Buonsanto*, 1999].
A comparison of the data on electron density
obtained by the Kharkov radar using the power profile
technique and the Faraday rotation measurements (this technique is
described, e.g., by
* Grigorenko* [1979])
made it possible to estimate
the error in determining
*T*_{i} and
*T*_{e} below
*F*2 region peak. This error
was not higher than 15% under quiet conditions.

**Powered by TeXWeb (Win32, v.2.0).**