Because of this concern, MDL analyzed the quality of archived MRMS data for the period January 2012 to March 2015, as these data were used to develop the LAMP products. This analysis revealed the random presence of false precipitation echoes, on rare occasions, throughout the MRMS grid domain and more frequent systematic false echoes across parts of the Central and Northern Plains, parts of the inter-mountain western US, and for all Canadian radars just north of the US border. In response, MDL developed and applied a comprehensive set of seasonally- and geographically-stratified dynamic consistency checks between MRMS composite reflectivity ≥ 40 dBZ in 5-km gridboxes and coincident TL measurements, other MRMS products, and fine scale and large scale numerical weather prediction model output parameters to identify and screen the random and systematic false echoes. In addition, static seasonal grid masks were developed and applied to mask MRMS products where reflectivity climatology maps revealed residual false precipitation echoes and where reflectivity is underreported due to rugged mountainous terrain or sparse network radar coverage.
Where MRMS products are masked by this MDL supplemental quality control (QC) process, corresponding simulated reflectivity products from the High Resolution Rapid Refresh (HRRR) model, developed and recently implemented by the NOAA Earth System Research Laboratory, are substituted to accommodate the LAMP applications. The impact of these supplemental MRMS QC processes on forecast performance of LAMP convection probabilities was assessed through skill score sensitivity testing and case study examination. Findings will be discussed in the conference presentation.