Comparisons of Properties and Detection Differences Between the National Lightning Detection Network and the Global Lightning Mapper Using an MRMS-based Convective Feature Database

Lightning detection between the Global Lightning Mapper (GLM) and the National Lightning Detection Network (NLDN) elicits respective bias, with the GLM having better detection of total lightning, and NLDN having better detection with cloud to ground lightning. In this study, a three-year Multi-Radar Multi-Sensor (MRMS) Convective Feature (CF) dataset from 2017 to 2019 is used to analyze the known variables associated with lightning production across the MRMS domain. CFs are defined by grouping the contiguous areas of composite reflectivity greater than 40 dBZ and seamless hybrid scan height less than two kilometers. This methodology is applied to MRMS data with 6-minute intervals and results in 220 million CFs over the years analyzed. Many lightning relevant variables including areas of various reflectivity thresholds at 0°C, -10°C, and -20°C, 18 dBZ echo top height, Vertically Integrated Liquid (VIL), and Maximum Estimate Size of Hail (MESH), are analyzed with respect to NLDN and GLM lightning rates. Geographic, seasonal, and diurnal relationships of these variables are shown. To understand the detection differences between the GLM and the ground NLDN network, the properties of CFs are further examined along with the differences between the GLM and NLDN lightning rates in CFs.