Syphons in tipping bucket rain gauges: How do they affect rainfall intensity estimates derived from inter-tip times?

Abstract

Tipping-bucket rain gauges (TBRGs) are widely used globally to record rainfall amounts over nominated accumulation durations such as 15-minute rainfall, hourly rainfall, and daily rainfall. To reduce the under-estimation of rainfall amounts that results from high rates of inflow to the tipping buckets in intense rainfall, small syphons are commonly installed between the rain-collecting funnel and the bucket mechanism. The syphon is designed to regulate the inflow, reducing the systematic error inherent in the basic TBRG mechanism.

The inter-tip times (ITTs) of a TBRG are often used as a source of rainfall intensity data, since the ITTs provide the highest temporal resolution possible from such gauges. However, the influence of the syphon is generally neglected in this approach to intensity estimation. Experimental results are presented here from a high-quality TBRG fed with known rates of water inflow, simulating rain of various constant intensities. Tests were replicated with the syphon installed, and with it removed.

Results confirm that the syphon significantly perturbs the sequence of ITTs, causing apparent fluctuations in intensity that do not in fact exist. At least two factors contribute to this. The first is the lack of correspondence between syphon and bucket volumes, such that more than one syphon operation may be needed to fill one bucket. The second factor is the continued inflow to the syphon during its emptying cycle. This lengthens the time required for the syphon to cycle, as well as worsening any mismatch between the amount of water discharged through the syphon to the tipping-bucket mechanism and the bucket capacity. In fact, there can be no fixed relationship between the syphon discharge volume and the capacity of a TBRG bucket. Therefore, one syphon emptying cycle can deliver more than the nominal 0.2 mm needed to trigger a tip event, such that the varying excess (dependent on the rainfall intensity) drains to the second bucket, which in turn tips earlier than would be expected. Consequently, even at constant intensity, the ITTs display increased variability, with widely-varying apparent intensities among successive ITTs. The variance among ITTs is shown always to be greater in data from a syphon-equipped TBRG than from a “straight-through” (no-syphon) gauge, such that the reliability of intensity data obtained from the ITTs of a syphon-equipped TBRG are reduced by the presence of the syphon.