Exploration of Using Simulated Reflectivity to Objectively Identify Convective Mode in Convection-Allowing Models

Roberts and  Lean  (2008)  and Schwartz  et  al.  (2010) describe a process for calculating “neighborhood” fractional probabilities from a single model.  By applying their formula to a convection-allowing model (CAM) at each grid point, the number of grid boxes with 1-km AGL simulated reflectivity ≥ 40 dBZ within a  40-km  radius of influence (ROI) can provide an estimate of  convective storm coverage   within  that “neighborhood”.   The resultant field is hypothesized to provide some information on convective mode (e.g., given the presence of updraft helicity, low fractional reflectivity coverage may correspond better with  isolated supercells than mesoscale convective systems).  For the current work, a few severe weather days would be subjectively evaluated to determine  whether neighborhood fractional coverage of simulated reflectivity can serve as a reliable proxy for identifying convective mode.  Further, we hope to establish some initial probability thresholds which can be used routinely to distinguish meteorological features.   Based on encouraging results from this investigation, the output will be examined in real-time during the next Hazardous Weather Testbed Spring Forecasting Experiment.