A Study of InCloud and CloudtoGround Lightning in TornadoBearing Supercells in the Midwest

1
A Study of In-Cloud and Cloud-to-Ground
Lightning in Tornado-Bearing Supercells in the
Midwest

• Ben Herzog and Patrick S. Market
• Dept. of Soil, Environmental Atmospheric
  Sciences
• University of Missouri
• Columbia, MO

2
Introduction

• Knapp (1994) indicated a correlation between
  cloud to ground (CG) lightning flash frequency
  and the time of tornado touchdown

3
Objectives

• To verify Knapps (1994) work
• To determine if identifying CG as well as in
  cloud (IC) lightning flash frequency trends could
  be used as an effective forecasting tool to
  determine tornado touchdown.

4
Methodology

• Find tornadic thunderstorms
• Storm Prediction Center storm reports
• Verify tornadoes from the National Climatic Data
  Center
• Obtain Data
• March 2007 June 2007
• Rocky - Appalachian Mountains
• Radar data from National Climatic Data Center
• Lightning data from Vaisala, Inc

5
Methodology (cont.)

• Break each storm into 5 minute periods to
  identify flash trends
• Start 60 minutes before first touchdown
• Create spread sheets on each storm containing
• Total number of flashes per five minutes
• Number of cloud to ground (CG) flashes per five
  minutes
• Number of in cloud (IC) flashes per five minutes
• Number of negative CG (CGN) flashes per five
  minutes
• Number of positive CG (CGP) flashes per five
  minutes
• Create a spread sheet containing all data from
  every storm

6
Analysis

• 30 total storms analyzed
• 26092 total flashes analyzed
• Some five minute spans had 0 flashes
• Some five minute spans had over 300 flashes
• 53.4 of flashes were CG
• 89.7 of CG flashes were negative
• 10.3 of CG flashes were positive
• 46.6 of flashes were IC

7
Analysis

• Total Flashes

8
Analysis

• CG Flashes

9
Analysis

• IC Flashes

10
Analysis

• Negative CG Flashes

11
Analysis

• Positive CG Flashes

12
Analysis

• CGN Flashes

CGP Flashes
13
Results

• As suggested in the study by Knapp, there is an
  identifiable pattern before tornado touchdown
• Approximately 30 minutes before touchdown, there
  is a maxima in flash frequency
• Approximately 20 minutes before touchdown, there
  is a minima in flash frequency
• Approximately 10 minutes before touchdown, there
  is another maxima in flash frequency
• Approximately 5 minutes before touchdown, there
  is another minima in flash frequency

14
Results (cont.)

• The cloud to ground flashes show this pattern
  especially well
• The negative CG flashes also show this pattern
• There are very few PCG flashes, so finding a
  pattern in the flash trend may be of little
  utility. However, at T-30 minutes before
  touchdown, we found 0 total PCG flashes. At that
  same time, the maximum number of NCG flashes
  occurred.

15
Results (cont.)

• There is a pattern associated with the in cloud
  flashes
• There is a pattern of maxima and minima of
  flashes in IC storms, but it is not nearly as
  pronounced of a pattern as the CG flashes

16
Conclusions

• If a reliable method of identifying lightning
  flashes and determining if they are CG or IC is
  available to the forecaster, I believe that that
  tool can be used as a method for forecasting
  tornado touchdown can come form this research

17
Future Work

• Obtain data on storms from different years and
  see is the flash frequency pattern matches 2007
• Classify the storms into different categories
• High precipitation supercells
• Low precipitation supercells
• Classic supercells
• Squall lines
• Mesoscale convective complexes

18
Acknowledgements

• Knapp, David I., 1994 Using Cloud-to-Ground
  Lightning Data to Identify Tornadic Thunderstorm
  Signatures and Nowcast Severe Weather. National
  Weather Digest, 19(2), 35-42