On the accuracy of optical disdrometer measurements

Disdrometers are widely used to measure the microphysical properties of precipitation, synthesised by the drop size and fall velocity distribution, and to infer integral properties like rainfall intensity, kinetic energy, visibility and radar reflectivity. Although often used to validate other methods of inferring precipitation products, including radar, satellite and opportunistic sensors, the accuracy and reliability of disdrometer measurements are rarely quantified. This study is the first to rigorously and systematically quantify the instrumental bias of optical disdrometer measurements in the laboratory. We tested two widely used optical transmission disdrometers, the OTT Parsivel 2 and the Thies LPM, using a raindrop generator that was specifically developed for calibrating non-catching precipitation measurement instruments. We also conducted tests using metallic spheres and demonstrated that the calibration obtained using artificial particles is not appropriate for optical disdrometers. The instruments investigated significantly underestimate drop size by up to 20%, with peaks of up to 40% for the Thies LPM when using the internal post-processing algorithm. The OTT Parsivel 2 overestimates fall velocity by about 5 to 10%, while the Thies LPM underestimates it by up to 15%. Slight linear trends of the percentage relative bias with drop size are observed for both drop diameter and fall velocity. However, adjustment of field measurements may be challenging, given the large dispersion observed. The mean bias and dispersion of drop size and velocity measurements vary along and across the laser beam, especially for the Thies LPM, with differences of up to 30% observed between drops falling at the opposite edges of the beam. Integral properties are largely underestimated, particularly at low rainfall intensity, although performance improves as intensity increases. These results highlight the importance of laboratory calibration to assess and quantify the instrumental bias of disdrometer measurements prior to their use in research or operational applications, especially when used to calibrate meteorological radars or to compare integral measurements with those of other instruments that measure precipitation in situ. While they provide relevant information about individual drop size and velocity, optical disdrometers need to be further improved before they can compete reliably with traditional rain gauges in measuring rainfall intensity. • Optical disdrometers were calibrated using generated water drops of various sizes. • Calibration with metallic spheres is inappropriate for testing optical disdrometers. • A relevant bias was observed in the measurement of drop size and fall velocity. • The measurement bias varies both along and across the instrument's sensing area. • Integral properties are largely underestimated especially at low rainfall intensity.