MONTE CARLO SIMULATIONS OF RANDOMLY DISTRIBUTED FALLING RAINDROPS: PROPERTIES OF DROP SIZE SPECTRA AS MEASURED BY A VERTICALLY POINTING FM-CW DOPPLER RADAR

摘要

Simple Monte Carlo simulations of the backscattered E-Field of randomly distributed, falling rain drops have been used to study the effects of spectral leakage and the small lines (marked by vertical arrows in fig. 3) when measuring a drop collective with an FM-CW Doppler radar (Rayleigh scattering assumed). These small lines have been neglected in the original literature (Strauch, 1976). In the simulations, all other sources of error (turbulence, echos from side lobes, intermodulation by the mixer, etc.) have been disregarded. Note that turbulence could be easily included via the probability function of turbulence (PFT) since every single drop can be assigned a distinct v_r. The "calibration" of the constant C in (1) was done in a way that the rain rate R in the second height bin equaled the rain rate given by the predefinded drop spectrum. It has been shown that the number of small drops may be considerably overestimated by an FM-CW Doppler radar in heavy rain. But this problem could be minimized through the use of specific window functions (section 5). A predescribed secondary mode at D ≈ 2 mm could be resolved reasonably well. Echo statistics as described in Marshall and Hitschfeld (1953) are also applicable to FM-CW-power spectra, which is not presented in this paper but has been derived from the simulations. These results do not depend on a specific radar frequency or a specific range resolution. The effect of the increasing underestimation of drop concentrations with height has to be checked by further simulations.