Towards Evaluation of the National Mosaic and Quantitative Precipitation Estimate (NMQ) Product by an Experimental Rain Gauge Network

Quantitative Precipitation Estimation (QPE) products from the National Mosaic and QPE System (NMQ) are compared with an independent rain gauge network in the Southern Delmarva Peninsula. The gauge network consists of 14 sites, where each site has three collocated tipping bucket rain gauges. The observational period extends from May 2009 to April 2013; however, this study focuses on the Hurricane Sandy event. A continuous record of gauge observations was constructed at each gauge site after careful quality control of the data. Events with frozen precipitation were manually deleted and the collocated gauge observations showed excellent agreement. The gauge record consists of the time of tip and the data (to the nearest second) and is integrated over 5-minutes to match the NMQ temporal resolution. The NMQ products employed here have three different versions: basic Hybrid Scan Reflectivity (HSR), Radar Quality Index-filtered (RQI) and operational gauge-adjusted plus RQI-filtered (HCF). Preliminary results show that the rain/no-rain agreement between the gauges and HSR NMQ is 80.3%, and increases to 85.2% and 83.4% respectively when compared with the RQI and HCF NMQ. Given that NMQ will be used as a validation product for the NASA's Global Precipitation Measurement mission satellite precipitation products and is also used by many other groups for rainfall estimation, it is important to carefully evaluate the NMQ and to provide feedback to its developers and our GPM partners at NOAA.