Rain Gauge Inter-Comparison Quantifies Deficiencies in Precipitation Monitoring

Abstract

Efforts to correct precipitation measurements have been ongoing for decades, but are scarce for tropical highlands. Four tipping-bucket (TB) rain gauges with different resolution that are commonly used in the Andean mountain region were compared—one DAVIS-RC-II, one HOBO-RG3-M, and two TE525MM TB gauges (with and without an Alter-type wind screen). The relative performance of these rain gauges, installed side-by-side in the Zhurucay Ecohydrological Observatory, south Ecuador, at 3780 m a.s.l., was assessed using the TB with the highest resolution (0.1 mm) as reference, i.e. the TE525MM. The effect of rain intensity and wind conditions on gauge performance was estimated as well, using 2 years of data. Results reveal that (i) precipitation amount for the reference TB is on average 5.6 to 7.2% higher than rain gauges having a resolution ≥0.2 mm; (ii) relative underestimation of precipitation from the gauges with coarser resolution is higher during low-intensity rainfall—a maximum deviation of 11% was observed for rain intensities ≤1 mm h-1; (iii) precipitation intensities of 2 mm h-1 or less that occur 75% of the time cannot be determined accurately for timescales shorter than 30 minutes because of the gauges’ resolution, e.g. the absolute bias is >10%; and (iv) wind has a similar effect on all sensors. This analysis contributes to increased accuracy and homogeneity of precipitation measurements throughout the Andean highlands, by quantifying the key role of rain-gauge resolution.