Advancement of a Real-Time Automated Gauge Quality Control Process for Multi-Radar Multi-Sensor Precipitation Estimation

Automated gauge networks provide direct measurements of precipitation at the surface and have been utilized in numerous hydrometeorological applications. While gauge observations are widely considered as “ground truth,” they have an inherent set of limitations that could adversely impact gauges values used for the calibration of remotely sensed quantitative precipitation estimations (QPEs), hydrological model predictions, etc. The Multi-Radar Multi-Sensor (MRMS) system uses hourly gauge observations in real-time to generate gauge-based QPE products as well as generate a local gauge-corrected radar QPE; therefore, a careful quality control (QC) procedure is necessary. The QC scheme presented here leverages the multiple data sources within MRMS to best retain quality hourly gauge observations while removing suspect observations. These data sources include the MRMS radar-only QPE (Q3RAD), a Radar Quality Index (RQI) product, and model-derived surface wet-bulb temperatures (Twb). The MRMS gauge QC algorithm flag gauges that are missing, outside of a time window from the top of the hour, impacted by winter precipitation, or reporting a false or outlier value. This presentation will discuss the evolution of the MRMS gauge QC logic to its current form with the operational implementation of MRMS and the performance of the QC logic. Results of this study will show the average distribution of gauges that are retained or removed through the QC algorithm, as well as how the gauge QC scheme impacts the generation of a local gauge-corrected radar QPE.