The purpose of this report is to document the Applied Meteorology Unit’s (AMU) evaluation of the Cell Trends display as a tool for radar operators to use in their evaluation of storm cell strength. The objective of the evaluation is to assess the utility of the WSR-88D graphical Cell Trends display for local radar cell interpretation in support of the 45th Weather Squadron (45 WS), Spaceflight Meteorology Group (SMG), and National Weather Service (NWS) Melbourne (MLB) operational requirements.
The evaluation guidelines, along with the selected case days, were determined by consensus among the 45 WS, SMG, NWS MLB, and AMU. Four case days for a total of 52 cells were selected for evaluation: 29 March and 23 April 1997 (cold season cases) and 11 July 1995 and 13 August 1996 (warm season cases). The WSR-88D (Level II) base data from the NWS MLB was used along with WATADS (WSR-88D Algorithm Testing and Display) Build 9.0 from the National Severe Storm Laboratory (NSSL) as the analysis tool. WATADS allows for the replay of historical data and has all the algorithms used in the WSR-88D Build 9.0 operational system so it closely matches the WSR-88D Principal User Processor (PUP) product output.
The analysis procedure was to identify each cell and track the maximum reflectivity, height of maximum reflectivity, storm top, storm base, hail and severe hail probability, cell-based Vertically Integrated Liquid (VIL) and core aspect ratio using WATADS cell trends information. One problem noted in the analysis phase was that the Storm Cell Identification and Tracking (SCIT) algorithm had a difficult time tracking the small cells associated with the Florida weather regimes. The analysis indicated numerous occasions when a cell track would end or an existing cell would be given a new ID in the middle of its life cycle.
Analysis of the four case days has shown that most cells associated with significant weather do show a trend in certain cell attributes (a signature). Hail was classified as an event if there was observed or reported .50 inch or greater in the vicinity of a cell. Wind cases were classified as an event if there was reported wind damage or if a microburst signature (gust GT 35 kt) was observed on CCAS wind tower network. Cells that exhibit a rapid increase or decrease in their height of maximum reflectivity and cell-based VIL over the same time period can be associated with significant weather. Those cells that do not show these changes are typically not associated with significant weather events. Other Cell Trends attributes did not show any useful signatures. As an example, the Hail Probability 100% threshold was indicated on 90% (47/52) of all storms.
This investigation has found that most cells, which produce large hail or damaging microburst events, have discernable Cell Trends signatures. Forecasters should monitor the PUP’s Cell Trends display for cells that show rapid (1 scan) changes in both the height of maximum reflectivity and cell-based VIL. For microburst development, there is generally
• A 8000 ft or greater decrease in the height of maximum reflectivity (initially at 18000 ft or greater) and
• A decrease in cell-based VIL by 10 kg/m2 or greater prior to the wind event.
For the hail events there is generally a rapid (1 volume scan)
• Increase in the height of maximum reflectivity by 8000 ft or greater to 18000 ft or greater and
• The cell-based VIL increased by 10 kg/m2 or greater prior to the hail events.
It is important to note that this is a very limited data set (four case days). Fifty-two (52) storm cells were analyzed during those four days. The above mentioned trends increase in the two cell attributes for hail events (reported ½ inch or greater) and decrease in the two cell attributes for wind events (observed or reported wind damage) were noted in most of the cells. Lead times ranged from 5 to 20 minutes. The probability of detection was 88% for both event types. The False Alarm Rate (FAR) was 36% for hail events and a respectable 25% for microburst events. In addition the Heidke Skill Score (HSS) is 0.65 for hail events and 0.67 for microburst events. For a random forecast, the HSS is 0 and for a perfect forecast, the score is 1.
Radar operators need to monitor storm cells and watch for trends that change quickly. Any quick change in a cell’s structure is usually a precursor that a cell characteristic has changed or is changing. These changes can be associated with the development or decay of a severe storm. Using the WSR-88D PUP Cell Trends display can help the forecaster in highlighting trends in a cell’s attributes of: maximum reflectivity, height of maximum reflectivity, storm top, storm base, hail and severe hail probability, and cell-based VIL. The AMU found that two of the important attributes to monitor are height of the maximum reflectivity and cell-based VIL. By monitoring these two attributes a forecaster can have a better advantage for determining if a storm cell is becoming severe and may produce a microburst or large hail
The 8th Conference on Aviation, Range, and Aerospace Meteorology