1. Reanalysis of Southern New England Tornadoes To Improve Warning Verification Daniel Brook, Lyndon State College* Joseph DelliCarpini, NOAA/NWS Taunton, MA David Vallee, NOAA/NWS NERFC, Taunton, MA *Present affiliation: Meridian Environmental Technology 1953 Worcester Tornado (Stanley H. Smith/Library of Congress Photo)

2. Background Verification data showed a need to improve tornado detection in BOX CWA –Low frequency, high impact events 14 cases included, dating back to mid 1990s (WSR-88D implementation) July 3, 1997 was most significant, with 4 tornadoes in one afternoon near southern New Hampshire

3. Purpose and Methodology Help operational forecasters improve warning performance –Identify days with tornadic potential Improve forecaster situational awareness –Identify critical signatures on radar Compile days with tornadoes to determine underlying causes. –Synoptic and mesoscale composite analyses –WSR-88D data from KBOX, KENX, and KOKX –Re-classify intensities to EF scale for consistency

4. Classification System A.Southern Ontario Closed Low (7) B. Open Great Lakes Trough (3) C. Deep Great Lakes Trough (1) D. Tropical Remnant (1) E. Cold Pool Aloft (2)

5. A: Southern Ontario Closed Low High CAPE –Over 1500 J/kg Strong Shear –More than 160 m 2 /s 2 High PW –Greater than 2 inches

6. Example – July 3, 1997 EF2 – Greenfield, NH Discrete cells –Storm that produced the EF2 had a history of producing three EF1 tornadoes Subtle indication in reflectivity data Velocity data shows gate-to-gate rotation (50 kt) as the event takes place –Little indication beforehand

7. 2357z | 03-Jul-1997 Base Reflectivity Greenfield, NH EF2 Storm-Rel. Velocity

8. B: Open Great Lakes Trough Moderate CAPE –1000-1500 J/kg Moderate shear –Near 150 m 2 /s 2 Moderate PW –Near 1.5 inches

9. Example – June 2, 2000 EF1 Leeds/Northampton, MA Tornado developed on a squall line Subtle hook echo visible in reflectivity Broad rotation –Sampling at mid levels

10. 2102z | 02-Jun-2000 Base Reflectivity Leeds, MA EF1 Storm-Rel. Velocity

11. C: Deep Great Lakes Trough Low CAPE –200-400 J/kg Moderate shear –Up to 150 m 2 /s 2 Moderate PW –Near 1.5 inches

12. Example – June 23, 2001 EF0 – East Hartland, CT Good example of how surrounding radars can be very important to examine. –Nothing significant seen from KBOX –KENX much more useful Discrete low topped cells –Subtle indication in reflectivity data –Broad rotation in velocity data

13. 1911z | 23-Jun-2001 Base Reflectivity Data from KENX East Hartland, CT EF0 Storm-Rel. Velocity

14. D: Tropical Remnant Low CAPE –400-600 J/kg Moderate shear –120 m 2 /s 2 High PW –Over 2 inches

15. Example – June 17, 2001 EF2 – Princeton, MA Remnants of Tropical Storm Allison –HP supercell (bookend?) embedded within line of convection Best radar signature in the study –Strong Mesocyclone status –Rotational velocity of 40 kt One of few cases where rotational couplet takes up several pixels instead of a few.

16. 1536z | 17-Jun-2001 Base Reflectivity Princeton, MA EF2 Storm-Rel. Velocity

17. E: Cold Pool Aloft Low CAPE –200-400 J/kg Little or no shear –Light wind profile Low PW –Less than 1 inch

18. Example – May 20, 2006 EF0 - Portsmouth, RI Formed along line of low-topped convection Very subtle reflectivity signature SRM shows no sign of rotation

19. 2242z | 20-May-2006 Base Reflectivity Portsmouth, RI EF0 Storm-Rel. Velocity

20. Summary: Synoptic/Mesoscale Findings Key points found –Usually a strong upper level jet in vicinity –“Tropical” airmass – Dewpoints of 70° or higher Low LCLs –Moderate to highly sheared environment Helicity – at least 120 m²/s² –Over 1000 J/kg of CAPE Found to be less consistent than shear –Sharp gradients as important as actual values –Low level boundary present for “spin up”

21. Stronger Tornadoes High-end EF1 and EF2 usually require –Moderate/High Shear(130 m 2 /s 2 ) –High CAPE (1500 J/kg) –Moderate PW(1.5 inches) A limited amount of “trading” can occur between each criteria above. Without a combination of these, weaker tornadoes form (EF0, low-end EF1)

22. Summary: Radar Findings Radar signatures can be found with a trained eye … –But many are comprised of very few pixels and some don’t show up at all (most EF0s) –Few have very prevalent signature

23. Summary: Warning Strategy Use multiple radars Issue SVRs for “short lived” EF0 tornadoes –Difficult to detect with subtle signatures Issue TORs for EF1 and stronger tornadoes –25 kt rotational velocity with depth and persistence

24. Integrating Research into Operations: July 23, 2008 EF1 Tornado –Warren RI and Swansea MA Forecasters recognized synoptic pattern and were aware of tornadic potential –Type A: Southern Ontario Closed Low As event unfolded, utilized warning strategy (SMW  SVR  TOR)

25. A: Southern Ontario Closed Low High CAPE –1800 J/kg Strong Shear –150 m 2 /s 2 Moderate PW –1.8 inches

26. Example – July 23, 2009 EF1 - Warren, RI / Swansea, MA HP supercell formed along line of convection moving NE from Long Island Subtle hook signature on reflectivity data prior to tornado formation Velocity data showed subtle gate-to-gate (45 kt) signature

27. 1946z | 23-July-2008 Base Reflectivity Warren, RI / Swansea, MA EF1 0.5 Velocity

28. GR2 Analyst Cross Section of Storm just West of Bristol, RI just West of Bristol, RI 2000 UTC

29. Barrington Beach, RI Credit: Chris Legro, NWS Davenport, IA Former NWS Taunton student intern

30. Verification SMW issued at 335 PM –Mentioned waterspout SVR issued at 350 PM TOR issued at 404 PM –Touched down 405 PM First tornado warning issued and verified with tornado report since July 1997 –Cheshire County, NH

31. NWS Taunton Storm Survey Photos

33. Reanalysis of Southern New England Tornadoes To Improve Warning Verification Questions ??? 1953 Worcester Tornado (Stanley H. Smith/Library of Congress Photo)