Cathy J. Kessinger, Edward A. Brandes and Jonathan W. Smith
Submitted to the 27th Conference on Radar Meteorology, Vail, CO, Amer. Meteor. Soc., Boston.
ABSTRACT
Aviation hazards associated with hailstorms include in-flight engine shutdown due to hail ingestion, damage to airfoils with subsequent loss of lift, and cracking or shattering of windshields. A hail detection algorithm that provides timely and accurate warnings could have substantial economic benefit. The Next Generation Radar (NEXRAD) Hail Detection Algorithm (HDA) and the National Severe Storms Laboratory (NSSL) HDA are evaluated. The NSSL HDA consists of the Probability of Severe Hail (POSH) and the Probability of Hail (POH). The NEXRAD HDA is currently in use at National Weather Service offices with a Weather Surveillance Radar-1988 Doppler (WSR-88D) installation. The NSSL HDAs are planned replacements of the NEXRAD HDA. All HDAs assess hail potential via a reflectivity-based determination of storm characteristics. The NEXRAD and the POH HDAs are designed to detect hail of any size (>0 mm) while the POSH HDA is designed to detect severe (>19 mm) hail.
The NCAR/Research Applications Program conducted a Hail Project in northeastern Colorado during the summer months of 1992 and 1993. Documentation of precipitation type and hailstone sizes comprises a "ground truth" verification data set for comparison with the predictions from the three HDAs. Statistical quantities are calculated to evaluate algorithm skill and performance at increasing hail size thresholds (>0, >6 mm, >13 mm, >19 mm).
The NEXRAD HDA performed best for hail of any size with a Critical Success Index (CSI) of 0.47. The NSSL POH had its best CSI performance at the smaller hail size categories with typical CSI values between 0.80 and 0.90. The POH HDA performed significantly better than the NEXRAD HDA by all measures except the False Alarm Ratio (FAR) where results from all HDAs were similar. For severe hail, the POSH HDA has the highest CSI score of 0.29 at the 50% probability threshold over the NEXRAD and the POH HDAs. The POSH HDA has a higher Probability of Detection (POD) than the NEXRAD HDA but the POH HDA attained the highest POD (0.98 at 50% probability threshold). We conclude that the NSSL POH HDA represents a significant improvement over the NEXRAD HDA for detecting hail of any size and that the NSSL POSH HDA exhibits some skill in detecting severe hail events.
Results from this study are compared to previous studies. Good agreement is found with a NEXRAD study done in Colorado. An NSSL study done with Oklahoma and Florida data produced better results than shown here and is believed to result from procedural differences in matching the precipitation observations to the algorithm predictions.