A Joss-Waldvogel disdrometer derived rainfall estimation study by collocated tipping bucket and rapid response rain gauges

This article studies the rainfall estimates derived from a JossWaldvogel disdrometer, using an extensive dataset of raindrop spectra for the period of 2003-2010. Four rain gauges (one tipping bucket and three rapid response drop counting devices) are employed for the appraisal with the disdrometer observations. The appraisal has been carried out in view of hourly rainfall accumulations, time series accumulative rainfall, and rain rate observations. From the yearly timescale statistics, the correlation between the disdrometer derived hourly rain accumulations to those measured by the rain gauges are in the range of 0.890.99 (mean absolute error, 0.100.45 mm, and normalized mean bias -1.03% to -50.28%). Especially, the estimated rainfall by the tipping bucket rain gauge is in sound agreement with the disdrometer observations, which is also reflected in time series accumulative rainfall comparisons, showing no more than 20% differences roughly. On the other hand, the results reveal that regardless of any influence of the integration period, the agreement between the disdrometer and the three rapid response rain gauges are quite consistent. Nevertheless, the association of the tipping bucket rain gauge is sensitive to the integration periods. In fact, increasing the integration period improves the rain rates agreement. Further to the appraisal for various rain classes, there is an underestimation to overestimation trend of disdrometer estimated rain rates with the increase of rain classes. Copyright (C) 2012 Royal Meteorological Society