Automated real-time operational rain gauge quality control tools in NWS hydrologic operations

Data and data quality are critical to the National Weather Service (NWS) mission. Rain gauge data are central to the hydrologic operational mission of the NWS. They are used not only to obtain the rain gauge-only analysis fields, but to correct for biases and errors in remotely sensed precipitation estimates from radar and satellites, and as ground truth for verification.

Even observations from well-sited rain gauges are subject to random errors due to mechanical problems, data transmission errors, and accidental collection stoppages caused by foreign objects or frozen precipitation. Currently, quality control of rain gauge data is performed mainly through a manual process at NWS River Forecast Centers and Weather Forecast Offices, through interfaces in the Advanced Weather Interactive Processing System (AWIPS). The manual interface is now being enhanced by the incorporation of features from the Mountain Mapper Daily QC process. While this manual process is necessary and adds value to precipitation analysis products, it is time consuming and subjective. Moreover, for flash flood forecasting needs, this manual quality controlling procedure is not practical because of timeliness demands. In order to provide quality controlled data on a timely basis, this quality controlling procedure needs to be automated as much as possible.

We have been developing semi-automated rain gauge data quality control assistance tools to support the NWS hydrologic operational mission. We will present a conceptual model for real time operational rain gauge data quality controlling, which includes both automated and manual aspects of the process. Then we will discuss some automated/semi-automated rain gauge data quality controlling techniques, some of which are operational currently (spatial consistency and multi-sensor checks), and others which have been proposed for implementation (temporal consistency and gauge history checks). These checks will be discussed in detail, and statistics showing the degree of agreement between manual and automated quality control will be presented.