Modeling of raindrop size distributions from multiwavelength rain attenuation measurements

Abstract

Abstract. Techniques for modeling the distribution of raindrop sizes, in terms of the lognormal and modified gamma distribution, from multiwavelength rain attenuation measurements at millimeter and infrared wave bands have been demonstrated. In order to obtain the three-parameter lognormal distribution, three experimental measurements are required, which can be attenuations at two frequencies and rain rate. Three measurements are used to form three equations, which are treated as nonlinear and are solved using a numerical technique to obtain the distribution parameters. The gamma distribution is treated as a two-parameter distribution, taking a fixed value of 2 for the index. The two parameters of the gamma distribution are obtained from measurements of rain rate and infrared attenuation following the technique described by Maitra and Gibbins [1995]. From an analysis of two rain events it is found that the measured attenuations at millimeter wavelengths show somewhat better agreement with the calculated attenuations obtained from the lognormal distribution than from the modified gamma distribution. The gamma model gives higher number densities for small drops than those given by the lognormal model. IntroductionThe distribution of raindrop sizes has been a subject of continued interest with the availability of more data from different locations. The suitability of a technique for DSD measurement,depends on the type of application. Direct measurements using a distrometer have limitations, given that it gives a localized measurement and that the impact distrometer has a relatively large time constant (---1 min). Radar measurements of DSD are suitable for obtaining a picture throughout the vertical extent of rain. The average DSD over a certain 657

A comparison of rain attenuation and drop size distributions measured in Chilbolton and Singapore

“…Fitting of measured drop size distributions to gamma, shifted lognormal, and Weibull distributions has shown that the shifted lognormal distribution performs most effectively in modeling the measured distributions, at most rain rate and frequency combinations. This is in agreement with the results reported by Maitra and Gibbins [1999]. The Weibull distribution performs almost as well in many cases, while the gamma distribution performs least effectively.…”

Modeling of fog droplet size distributions from infrared light attenuation measurements

“…The fine tuning of the methods to predict optical attenuations over free-space optical links, working under certain fog environments, requires the optimum scaling of the fog DSD parameters that lead to the reliable behaviour of these links under harsh environments. To compute EDSD parameters ƒ and N 0 against the fog events recorded at Prague, Maitra and Gibbins procedure is adopted (see, for example, [11]). This procedure considers three nonlinear equations to determine three parameters of the lognormal distribution in case of raindrops.…”

A novel method for predicting site dependent specific rain attenuation of millimeter radio waves

“…In order to obtain stability over some time, a time series rain event is frequently modeled as a process where raindrops are distributed as for instance gamma or log-normal functions of (slowly changing) time varying parameters [12], [13]. By assuming that the parameters vary slowly and not significantly within a single rain event, one can categorize events by these parameters.…”