The problem of lightning echo interpretation is investigated through the analysis of time and range variations of echoes received by S-, L-, and VHF-band radars. It is suggested that the rise time of the lightning echo amplitude is due to penetration by the ionized channel through the radar resolution volume. For the first time, by combining VHF mapping information on the location of sources of radiation with L-band echoes from lightning, the spatial extent and the speed of lightning propagation were independently confirmed. Combining an analysis of storm evolution (precipitation echoes) with lightning echo density evolution shows that the maximum number of identified flashes per kilometer per minute (lightning echo density) tends to stay close to the leading edge of a moving precipitation core and does not always coincide in range with maximum precipitation echoes as commonly thought.

A group of sequential echoes that are separated in range but closely spaced in time has now been designated as 'associated discharges'. The hypothesis that 'associated discharges' represent an interaction between lightning processes in the different parts of a storm and is tested against the alternative hypothesis that they are a random coincidence of independent events. The conclusion of the test shows that 'associated discharges' cannot be described as independent events at the 5% level of significance. Hence, they may represent interaction between lightning discharges.

The effectiveness of circular polarization for improving the detection of lightning in heavy precipitation is analyzed both theoretically and experimentally. It is shown that use of circular polarization increases the number of discharges detected by L-band radar by about 40 percent.

The model for the radar cross section (RCS) estimate of lightning as a finite length dielectric cylinder at an oblique angle of incidence is offered. The model is based on the exact solution of Maxwell's equations for a dielectric cylinder of infinite length and on assumptions made about dielectric losses and radiation and reflection losses in joints between neighboring lightning elements. The proposed model is valid for cylinders with radii that are a small fraction of wavelength. The RCS per unit volume is suggested as a basic parameter to compare theoretically and experimentally obtained estimates of RCS.