On Improving Tornado Detection in the Northeastern U.S. Via an Objective Radar and Near-Stormenvironment Algorithm

Tornado detection via radar interrogation is one of the most challenging tasks of operational
forecasters. In the northeastern United States, this is especially the case given that a vast
majority of tornadoes are weak and short-lived. Early identification of the radar signatures that
are critical to improved detection of tornadoes remains an important need of operational
forecasters. This study investigates the utility of an objective algorithm using radar and near-
storm environmental parameters first developed at the National Weather Service Weather
Forecast Office (WFO) in Boston/Norton to discriminate tornado-producing storms from
nontornadic storms (a subset of which were tornado-warned) during the 2010 to 2019 period in
the WFO Mount Holly county warning area (CWA). Preliminary findings suggest that the
algorithm improves upon office tornado warning verification statistics during the ten-year time
frame and that little or no variation of the algorithm is required from that developed for the WFO
Norton CWA. Remaining challenges in the real-time implementation and automation of the
algorithm are discussed.
1 Rutgers University
2 NWS WFO Mount Holly, NJ