Dynamic real-time volumetric correction for tipping-bucket rain gauges

摘要

Tipping bucket rain gauges are the most popular devices for determining rainfall intensity and precipitation depth. Even though application of a tipping bucket gauge presupposes calibration, they are susceptible to various random, mechanical, and systematic errors. In general, precipitation measurements may be underestimated by 5% to 40%. This paper presents a method of determining volumetric correction that compensates for systematic errors caused by variable rainfall intensity. The main goal of this study was to develop a suitable mathematical model that can be readily implemented in real time during field measurements, including the possibility of compensation of each individual tip. The developed algorithm is based on a "tip interval" recording method; this is in contrast to the standard measurement method, which is based on "tip count" and nominal tip volume. Such a solution may be applied to any kind of tipping-bucket rain gauge and can be effortlessly implemented in modern digital data loggers. In addition, smaller canisters may be used in a seesaw-like mechanism that provides accurate measurements over a considerably wider range of rainfall rates, even including extreme rainfall. During meticulous laboratory experiments, various tipping buckets with nominal volumes of 3 cm(3), 4 cm(3), 5 cm(3), 10 cm(3)and 200 cm(3) were tested over a wide range of simulated precipitation rates from 8 mm.h(-1) to 500 mm.h(-1). The extent of error reduction was 5.2%, 11.4%, 17.7%, 25.8% and 37.7% for rainfall intensities of 50 mm.h(-1),100 mm.h(-1) , 200 mm.h(-1), 300 mm.h(-1) and 500 mm.h(-)(l), respectively (assuming that the nominal volume of the tipping bucket was 4 cm(3) and the collection area of the rain gauge was 200 cm(2)). Besides scrupulous laboratory tests, the presented algorithm was verified during field measurements at 29 research sites located in a temperate climate region in two vegetation zones: foothills and lower montane. The results of this research will considerably enhance the accuracy of the precipitation data that are essential in various hydrological studies.