AQUAS - A quality control tool at GeoSphere Austria

Rain gauge measurement network of the Austrian national weather service operated by GeoSphere Austria comprises about 270 weather stations equipped with weighing rain gauges and, at a smaller part, with tipping bucket rain gauges. Each gauge is additionally equipped with a precipitation monitor that detects the beginning and the end of precipitation events. Precipitation data are checked for plausibility and completeness in several steps within a framework of an automated quality control tool called AQUAS (short for Austria Quality Service). The software was developed in 2016 at ZAMG (now GeoSphere Austria) in Vienna as part of the quality management in the area of real-time processing of near-surface observation data.
The basis for quality control procedure is formed by standard methods for checking meteorological and climatological data in accordance with the WMO recommendation (e.g. plausibility check, temporal, spatial and internal consistency check, etc.); in addition, test procedures are developed that take into account the specific errors in the measuring devices.  The test methods are continuously improved and further developed within the framework of AQUAS. Individual system components are designed to test the incoming observation parameters in real time - at the time resolution of 10 minutes, for example, for wind or temperature data and down to 1 minute time resolution for precipitation data. In AQUAS, each parameter can be processed independently of the other measured variables of a weather station. In addition, data from other sources are implemented in AQUAS, such as radar and satellite data. Data from numerical weather prediction models and data from other measurement networks, such as hydrological network or another third-party network, can also be integrated.

Some examples for the operational use of AQUAS and the current state of research on quality control of precipitation data will be presented. As an example, a novel method for real-time quality control of 1-minute weighing gauge precipitation data is demonstrated, which detects missing gauge precipitation based on the observation of the precipitation monitor and the total weight changes of the rain gauge.