1. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 Chris Melick 1,2, Patrick Marsh 1, Andy Dean 1, Israel L. Jirak 1, and Steven J. Weiss 1 1 NOAA/NWS/NCEP/Storm Prediction Center 2 CIMMS, University of Oklahoma Norman, OK Lightning Characteristics and Relationship to Preliminary Local Storm Reports

2. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 INTRODUCTION The National Weather Service (NWS) Storm Prediction Center (SPC) is responsible for forecasting lightning and severe thunderstorms, as well as for other high impact hazardous weather (including fire and winter precipitation) For SPC, observations serve crucial role given short-term nature for many of the products issued –Mesoscale Discussions (MD)/Watches focus on few hours after issuance –Provide Situational Awareness –Evaluation and Verification

3. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 MOTIVATION NWS defines a severe thunderstorm to contain wind gusts ≥ 50 kt, hail ≥ 1” in diameter, and/or any tornado –However, all thunderstorms (regardless of severity) still also require presence of lightning. –While relatively rare, winter thunderstorms can occur where lightning is coincident with frozen precipitation (e.g., thundersnow) More detailed information about convective cells can be provided by combining observational networks –Total [Cloud-to-Ground (CG) and Intra-Cloud (IC)] Lightning –Preliminary Local Storm Reports (LSRs) –Others not examined here (e.g., Radar and Satellite, ASOS/AWOS)

4. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 The relationship between preliminary LSRs and total lightning flash data is explored using time-matched, high- resolution gridded data from 2014. The purpose is to provide some initial statistical characteristics of the lightning data when stratified by geographic region, time of year, and type of flash (IC or CG). Details on the location and frequency of active lightning grid points relative to severe and winter type reports are also explored. MOTIVATION

5. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 Lightning Data at SPC SPC has access to real-time data feeds from two vendors: A.Vaisala’s National Lightning Detection Network (NLDN) B.Earth Networks Total Lightning Network (ENTLN) SPC also gets quality-controlled (QC) NLDN CG flashes for prior year. Transition from CG  Total Lightning For 30 years, CG lightning flash data used at SPC to monitor on-going convective systems (Bothwell 2014). CG part is small in Total Lightning when compared with IC. Further, IC flashes often precedes CG flashes in convective initiation (MacGorman and Rust 1998). Concern addressed as SPC acquired Total Lightning data in late 2012. For current work, characteristics from the QC CG NLDN data for 2014 will be contrasted to the CG and IC flash counts from ENTLN

6. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 Local Storm Reports (LSRs) Local Storm Reports (LSRs) – Routinely used at SPC for decades Preliminary but quick way to summarize recent/current weather events. Traditionally, focus placed on Severe (tornado, wind, hail) LSRs Recently, SPC developed LSR decoder to include Winter weather related type reports (Sullivan et al. 2014). NEW: Snow, Heavy Snow, Freezing Rain, Sleet, Blizzard, Ice Storm All LSR types examined here for 2014 except for Blizzard and Ice Storm as they comprise a very small proportion of the Winter sample size

7. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 DATA PROCESSING For all of 2014, Hourly Bins created for: 1.CG/IC/TOTAL ENTLN and QC CG NLDN Flash counts 2.Separate tallies of Total Severe LSRs and Total Winter LSRs Placed on nearest 0.04deg x 0.04deg Lat-Lon grid point Analysis domain restricted to continental United States QC of problem data reduced sample size to 8712 hours (363 days).

8. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 QC CG NLDN > CG ENTLN SUMMER: IC SLOPE INCREASES

9. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 89% of ENTLN Detections are IC NLDN (QC CG) has 1.22 times more detections than ENTLN CG

10. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 CG ENTLNQC CG NLDN DIFFERENCE FIELD: QC CG NLDN – CG ENTLN More areas in US where NLDN > ENTLN (RED)

11. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 CG ENTLNIC ENTLN Extensive, much higher IC compared to CG IC > 1000 in FL/Gulf Coast/Missouri Valley

12. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 CG ENTLNIC ENTLNTOTAL ENTLN

13. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 DATA PROCESSING For an effective comparison, binary (1/0) event grids in both observation datasets were constructed by specifying thresholds. For the LSRs, a severe object was recorded if ≥ 1 reports occurred within the hour. Similarly, the case of ≥ 1 Flashes indicated a grid point with an active thunderstorm. Grid-point Matching of Event Grids Results in 2x2 Table: LSR with LightningLSR with NO Lightning Lightning with NO LSRNeither Lightning nor LSR

14. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 LIGHTNING AND LSRs: MATCHING ILLUSTRATION Nearest Grid Point versus Neighborhood

15. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 Nearest Grid Point Matching LTNG (Blue dots) with LSRs (Red Stars) LSR but NO LTG LSR with LTG

16. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 Neighborhood Matching LTNG (Blue dots) with LSRs (Red Stars) ROI: 40-KM

17. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 Lightning Dataset QC CG NLDN CG ENTLN IC ENTLN TOTAL ENTLN Percentage of ALL Severe LSRs with Lightning [%] (n = 21,194 grid points) 97.46297.21197.93398.014 Lightning Dataset QC CG NLDN CG ENTLN IC ENTLN TOTAL ENTLN Percentage of ALL Winter LSRs with Lightning [%] (n = 64,332 grid points) 0.5550.6220.6390.759 2014 Domain Percentages: Neighborhood Lightning with Severe LSRs (Top) and Winter LSRs (Bottom) Better Relationship with Total Lightning: Higher percentages!

18. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 Severe LSRs with NO Lightning Counts of ≥ 1 Objects within 40-km ROI QC CG NLDN (Left) Versus Total ENTLN (Right) Less Severe LSR Events without Total Lightning compared to just CG!

19. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 Winter LSRs with Lightning Counts of ≥ 1 Objects within 40-km ROI QC CG NLDN (Left) Versus Total ENTLN (Right) More Winter LSR Events with Total Lightning compared to just CG!

20. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 SUMMARY 2014 Bulk Lightning Datasets: –Hourly, High-resolution CG/IC Lightning Flash Grid –Observations linked to LSRs by defining “Active” Objects The use of NLDN/ENTLN permitted: 1.A systematic evaluation and comparison of annual lightning properties from different detection networks 2.Identifying contrast in thunderstorms by incorporating IC in Total lightning versus just using CG flashes

21. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 CONCLUSIONS Flash Counts of ENTLN and NLDN show: –Similar CG pattern but NLDN detection 1.22 x more than ENTLN –IC more extensive/numerous than CG across US as IC portion of Total Lightning was large [~0.89] –FL/Gulf Coast highest frequency TSTM location with secondary max more evident with IC in Missouri Valley Relationship between Lightning and LSRs: –Severe (Winter) LSRs consistently (rarely) coincide with CG/IC Flashes as greater (less) than 97% (1%) of grid points meet criteria. –Still, better relationship found by incorporating IC in Total Flashes –Density of Severe LSRs occurring without lightning greatest in Mid-Atlantic States. Principal locations of Winter LSRs occurring with lightning in Central/Southern Plains, near Great Lakes, and Rocky Mountains

22. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 FUTURE WORK Some Items to Explore: 1.Diurnal: Which hour has most lightning and strongest correlation? 2.Seasonal: Highest frequency of lightning with winter LSRs? 3.Methodology: Does relationship increase with time-window? 4.Case study: What distinguishes a success mode of LSRs in close proximity to lightning flashes from a null scenario? ACKNOWLEDGMENTS This work was prepared with funding provided by NOAA/Office of Oceanic and Atmospheric Research under NOAA-University of Oklahoma Cooperative Agreement #NA11OAR4320072, U.S. Department of Commerce. The statements, findings, conclusions, and recommendations are those of the author(s) and do not necessarily reflect the views of NOAA or the U.S. Department of Commerce.

23. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 WINTER SEVERE

24. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 Lightning with Severe LSRs Counts of ≥ 1 Hours within 40-km ROI

25. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 Winter LSRs with NO Lightning Counts of ≥ 1 Hours within 40-km ROI

26. 40 th National Weather Association Annual Meeting Oklahoma City, OK October 20 th, 2015 WINTER SEVERE