Performance of a weighing rain gauge under laboratory simulated time-varying reference rainfall rates

Abstract The available calibration experiences about rain intensity gauges relying on the weighing measuring principle are based on laboratory tests performed under constant reference flow rate conditions. Although the Weighing Gauges (WG) do provide better performance than more traditional Tipping Bucket Rain Gauges (TBR) under constant reference flow rates, dynamic effects do impact on the accuracy of WG measurements under real-world/time-varying rainfall conditions. The most relevant biases are due to the response time of the measurement system and the derived systematic delay in assessing the exact weight of the volume of cumulated precipitation collected in the container. This delay assumes a relevant role in case high resolution rainfall intensity (RI) time series are sought from the instrument, as is the case of many hydrologic and meteo-climatic applications (the one-minute time resolution recommended by the WMO for rainfall intensity measurements is here assumed). A significant sampling error is also attributable to some kind of weighing gauge, which affects the low intensity range as well. A laboratory investigation of the accuracy and precision of a modern weighing gauge manufactured by OTT (Pluvio 2 ) under unsteady-state reference RI conditions is here addressed. Three different laboratory test conditions are applied: single and double step variations of the reference flow rate and a simulated real-world event. The preliminary development and validation of a suitable rainfall simulator for the generation of time-variable reference intensities is presented. The generator is demonstrated to have a sufficiently short time response with respect to the expected instrument behavior in order to ensure effective comparison of the measured vs. reference intensities. The measurements obtained from the WG are compared with those derived from a traditional TBR (manufactured by Casella) under the same laboratory conditions. The TBR measurements have been corrected to account for systematic mechanical errors and comparison is also proposed after applying further algorithms to reduce the sampling errors. Results indicate that the performance of the investigated WG under unsteady (real world) conditions in the laboratory is comparable or even lower than what can be obtained from more traditional TBRs, even in case corrections for sampling errors are not applied.