A Self-Evaluating Disdrometer for the Measurement of Raindrop Size and Charge at the Ground

Abstract

A disdrometer for the measurement of the co-spectra of raindrop size and charge is described which can evaluate to a significant extent spurious data caused by unavoidable drop overlap within the charge detector volume. The sizes of individual drops in the range 0.1?2.5 mm radius and charges within the limits of magnitude 0.1?10 pC could be determined. The use of two size detectors enabled the measurement of drop velocity and the detection of drop overlap within the charge-sensitive volume to be made. Non-ideal data which arise from natural conditions, measuring sites, and from fundamental or unavoidable deficiencies in disdrometer design, could be tested and monitored automatically using a central processor under software control: TIME OUT, when a drop failed to pass through both size detectors; COINCIDENCE, caused by simultaneous occupation of both size detector volumes; SIZE MISMATCH, when the separate size measurements did not agree within limits predetermined by software; HIGH or LOW VELOCITY, when the actual drop velocity was not close to the terminal value expected from size measurement. The self-evaluating disdrometer cannot be designed to provide the minimum error content possible but offers the advantage of assessing the proportion of spurious data present; it is argued that this may be preferable to the situation in which the error content is lower but unknown. The performance of the instrument was assessed using individual drops generated in the laboratory and by exposure to natural rain falling through an aperture into a chamber shielded from wind and associated turbulence. The latter test was made at a non-ideal site in order to demonstrate the ability of the disdrometer to provide information on invalid data so that raw co-spectra may be corrected. In the preliminary tests described a substantial proportion of the drops possessed fall speeds significantly below their expected terminal velocity, in some cases as much as 30% less, and not more than 20% of the drops detected satisfied all criteria for acceptance. Further, an examination of drop arrival rates showed that not all data could be fitted to a Poisson-type distribution, either because of rapid changes in the mean arrival rate or on account of the clustering of drops. The potential seriousness of the drop overlap problem, which is fundamental to all methods of measurements, is emphasized in the trial analysis: uncertainties in the exact form of the size distribution, particularly for drops in the radius range 0.1?0.5 mm, render the design of any instruments of fixed entrance aperture size dubious; the co-spectra must be expected to show appreciable distortion unless data associated with drop overlap, particularly within the charge-sensitive volume, are excluded. Some improvements in the current disdrometer design are suggested.