The need exists for measurements with high vertical resolution when observing the variety of atmospheric processes with extremely small vertical extent, such as microscale turbulence and scattering layers associated with inertia gravity waves. For example, recent in situ observations have shown that both humidity and temperature sheets, with thicknesses of the order of meters, exist throughout the lower atmosphere. Hampered by bandwidth constraints, however, standard pulsed radar systems have shown only limited usefulness in the detection of such phenomena. Frequency domain interferometry can be used to estimate the position and thickness of a single scattering layer within the resolution volume. Using two closely spaced frequencies, the method is derived under the restrictive assumption of a single, Gaussian-shaped layer. We will now introduce range imaging (RIM), which fully exploits the general advantages of frequency diversity. Using a set of closely spaced transmitter frequencies, a generalized method based on constrained optimization will be used to reconstruct high-resolution images of the average power density as a function of range. The technique will be studied using simulated radar data and will be shown to be capable of resolving complex structures similar to Kelvin-Helmholtz billows, which can be much smaller in vertical extent than the resolution volume. ¿ 1999 American Geophysical Union