Analyzing Downburst Precursors Using Automated Storm Identification and Polarimetric Radar Data

Strong thunderstorm winds produced by downbursts pose a threat to life, property, and aviation, yet they remain a challenge to predict in advance. Current operational understanding of downburst precursors include a divergent or convergent velocity signature on radar at the surface or mid-levels, respectively, a descending radar reflectivity (Z) core, and environmental characteristics (e.g., downdraft convective available potential energy (DCAPE), lapse rates) that are favorable for downbursts. However, divergence signatures only occur once downbursts have reached the surface and may not be observed at distances far from the radar, and not every downburst is reliably associated with mid-level convergence or a descending Z core. Similarly, environmental parameters are useful when forecasting for a broad area, but not every thunderstorm in an environment with those conditions favorable for downbursts will produce one. To address this dearth of reliable radar-based precursors, research is underway to explore whether polarimetric radar offers additional downburst precursor signatures.

Previous studies using polarimetric radar to analyze downburst-producing thunderstorms have observed characteristics such as a descending specific differential phase (KDP) core and a trough of decreased differential reflectivity (ZDR) extending below the melting layer. However, these studies either manually analyzed individual signatures or focused solely on case studies. This research expands on those studies by using the Multi-Cell Identification and Tracking (MCIT) algorithm to automate storm detection and analyze a large dataset of cases which span the contiguous United States. For each case, polarimetric radar variables, signatures, and derived products hypothesized to potentially be relevant to downburst formation, including Z, KDP, vertically integrated liquid (VIL), ZDR columns, hydrometeor classification algorithm (HCA), and convergence/divergence, are analyzed to find any consistent patterns leading up to downburst events.

Results so far indicate that signatures observed in previous studies and those used in operational forecasting are often present, including descending Z and KDP cores, a trough of decreased ZDR below the melting layer, and a divergence signature at the surface. However, not all of these characteristics are observed for every downburst event. A further investigation into which polarimetric radar characteristics, as well as favorable environmental parameters, are consistently observed prior to downburst occurrence will be conducted and discussed. The influence of geographic location on which signatures are present will also be assessed.