Observations of Near-Surface Thermodynamic and Wind Shear Profiles on Significant Tornado Days Observations of Near-Surface Thermodynamic and Wind Shear Profiles on Significant Tornado Days Photo Credit: Ming Ying Wei NWS Duluth Minnesota Great Lakes Operational Meteorology Workshop – Toronto, Onrario 22 March 2010 Dan Miller Science and Operations Officer NWS/WFO Duluth, Minnesota Dan Miller Science and Operations Officer NWS/WFO Duluth, Minnesota
Some Preliminary Thoughts… 1)Compilation of case observations/discussions 2)There are more questions posed than conclusions drawn from this talk 3)Evidence warrants further investigation by researchers of these topics through modeling/field ops/etc. 4)The soundings/hodographs to be presented are in no way to be interpreted in a universal manner for forecasting significant tornado environments! 1)Compilation of case observations/discussions 2)There are more questions posed than conclusions drawn from this talk 3)Evidence warrants further investigation by researchers of these topics through modeling/field ops/etc. 4)The soundings/hodographs to be presented are in no way to be interpreted in a universal manner for forecasting significant tornado environments!
Which VWP/Hodo is “Better” for Tornadoes? Lots of 2-3” Hail Limited Wind no Tornadoes Lots of 2-3” Hail Limited Wind no Tornadoes Multiple Cyclic Tornadic Supercells F2-F3 tornadoes Multiple Cyclic Tornadic Supercells F2-F3 tornadoes
Which Sounding is “Better” for Tornadoes? Multiple long-tracked F3-F5 tornadoes Classic Supercells Multiple long-tracked F3-F5 tornadoes Classic Supercells Lots of Hail/Wind 2 short-lived weak Tornadoes HP Supercells (strong cold pools) Lots of Hail/Wind 2 short-lived weak Tornadoes HP Supercells (strong cold pools)
Oklahoma: 3 May UTC m agl 350 m agl 350 m agl wind 1000 m agl Observed Storm Motion SFC Wind
Missouri: 4 May UTC m agl 350 m agl 1000 m agl Observed Storm Motion SFC Wind 350 m agl wind
Northeast Kansas: 4 May UTC m agl 360 m agl 1000 m agl Observed Storm Motion SFC Wind 360 m agl wind
Oklahoma: 8 May UTC m agl 350 m agl 1000 m agl Observed Storm Motion SFC Wind 350 m agl wind
Kansas/Oklahoma: 26 April UTC m agl 300 m agl 1000 m agl Observed Storm Motion SFC Wind 350 m agl wind
Ohio/Tennessee: 10 November UTC m agl 400 m agl 1000 m agl Observed Storm Motion SFC Wind 400 m agl wind
Pennsylvania/Ontario: 31 May UTC m agl 400 m agl 1000 m agl Observed Storm Motion SFC Wind 400 m agl wind
Ohio Valley Region: 3 April UTC m agl 400 m agl 1000 m agl Observed Storm Motion SFC Wind 400 m agl wind
Western Tennessee: 2 April UTC m agl 500 m agl 1000 m agl Observed Storm Motion SFC Wind 500 m agl wind
Minnesota: 16 June UTC m agl 350 m agl 1000 m agl Observed Storm Motion SFC Wind 350 m agl wind
Edmonton Alberta: 31 July UTC m agl 450 m agl 1000 m agl Observed Storm Motion SFC Wind 450 m agl wind
California (Sacramento) – 21 February UTC m agl 500 m agl 1000 m agl Observed Storm Motion SFC Wind 500 m agl wind
Question: Is the “sickle” shape to the hodograph real, or merely an artifact of data sampling? Wind Measured By Radiosonde Observed hodograph Surface wind measured by Anemometer
Question: Is the “sickle” shape to the hodograph real, or merely an artifact of data sampling? 1000 m agl SFC Wind 400 m agl wind NAM bufr forecast hodograph
Question: Is the “sickle” shape to the hodograph real, or merely an artifact of data sampling? ~350 m agl 55-60kt outbound ~350 m agl 55-60kt outbound ~350 m agl 60-65kt inbound ~350 m agl 60-65kt inbound Greensburg KS Event: 5/4/2007
Mean Parameters of the 20 Cases: Surface Temperature: 76 Surface Dewpoint: 68 Surface T/Td spread: 7.7 Surface Relative Humidity 69% LCL Height (agl): 2630 ft (802 m) LFC Height (agl): 4425 ft (1349 m) CAPE (surface parcel): 3206 j/kg CIN (surface parcel): 34 j/kg Surface Temperature: 76 Surface Dewpoint: 68 Surface T/Td spread: 7.7 Surface Relative Humidity 69% LCL Height (agl): 2630 ft (802 m) LFC Height (agl): 4425 ft (1349 m) CAPE (surface parcel): 3206 j/kg CIN (surface parcel): 34 j/kg (2/21/2005 Sacramento Case Not Included)
Operational Implications? How often do you get a warm and very humid airmass, that possesses strong instability and sufficient deep-layer shear for supercells that is also co-located with strong near-surface shear – and is nearly un-capped? From Nordin and Brooks, 2002
Some Important (and Perhaps Troublesome) Questions: 1) What do we mean when we say “elevated” vs. “surface-based” convection? 2) Do we need to consider “elevated” vs. “boundary-layer” vs. “surface- based” convection? 2) Do we need to consider “elevated” vs. “boundary-layer” vs. “surface- based” convection? 3) How do we *know* what parcels are ascending into the updraft? 4) What implications does this have for many of our near-storm environment forecast parameters? 4) What implications does this have for many of our near-storm environment forecast parameters?
Now The Dirty Details: All of this critical “stuff” is going on in a VERY shallow near-surface layer Red = SFC – 400m agl Cyan = 400m – 1000m agl Lavender = 1000m – 7000 m agl Red = SFC – 400m agl Cyan = 400m – 1000m agl Lavender = 1000m – 7000 m agl
Just Exactly How Shallow is this Layer? 457 m 1500 ft 457 m 1500 ft 553 m 1815 ft 553 m 1815 ft
Question: Is there a more effective way to examine low-level wind shear? Are We Looking Low Enough? Surface-400m shear vector Surface-1 km shear vector
Mean Parameters of the 20 Cases: Height of hodograph kink agl: 399 m Bulk Shear Vector Magnitude (sfc-kink): 18 kt Bulk Shear Vector Magnitude (sfc-1 km): 25 kt Bulk Shear Vector Ratio: 0.72 Height of hodograph kink agl: 399 m Bulk Shear Vector Magnitude (sfc-kink): 18 kt Bulk Shear Vector Magnitude (sfc-1 km): 25 kt Bulk Shear Vector Ratio: 0.72 (2/21/2005 Sacramento Case Not Included)
Central Florida – 25 December UTC m agl 300 m agl 1000 m agl Observed Storm Motion SFC Wind 300 m agl wind
Question: What is our true skill in choosing the “correct” parcel to lift in the calculation of numerous popular near-storm environment parameters and indices?
What about the mixed boundary layer? Question: What is our true skill in choosing the “correct” parcel to lift in the calculation of numerous popular near-storm environment parameters and indices?
Question: Do we need to re-evaluate our use of the terms “elevated” and “surface-based” convection?
What are the “correct” parcels with this thermodynamic profile? Question: Do we need to re-evaluate our use of the terms “elevated” and “surface-based” convection? How do we define “surface-based” DMC? Theta-e decreases rapidly with height
Question: What is the importance of surface heating in the contribution to instability on significant tornado days? How does the atmosphere produce/maintain this thermodynamic profile in the near-surface layer near max heating time?
Calculation of both of these indices for some useful purpose requires an accurate input value of total CAPE and shear over the appropriate layer (0-1 km/0-3 km/etc.)… …but how do we know what is the appropriate parcel to choose for an accurate value of CAPE? – and therefore… …how do we know what effective shear the storm is tapping? Calculation of both of these indices for some useful purpose requires an accurate input value of total CAPE and shear over the appropriate layer (0-1 km/0-3 km/etc.)… …but how do we know what is the appropriate parcel to choose for an accurate value of CAPE? – and therefore… …how do we know what effective shear the storm is tapping? Question: Can we improve on the utility of the two near- storm environment significant tornado parameters that have shown the most promise: namely surface-1km EHI and surface- 3km VGP?
Implications for NSE Parameters: 100 mb Mean-Layer CAPE (MLCAPE) 100 mb Mean-Layer CIN (MLCIN) Lowest 100 mb Averaging is “safer” - well-mixed BL should have uniform thetae Averaging is dangerous!! - thetae decreases rapidly with height in BL Difference in computed CAPE is small Difference in computed CAPE can be large - ~ j/kg! **(VGP/EHI)**
Implications for NSE Parameters: 0-1 km and 0-3 km Energy-Helicity Index (EHI) 0-3 km Vorticity Generation Potential (VGP) Lowest 100 mb If the storm isn’t tapping *surface* parcels (i.e. below ~ m) – it isn’t realizing the full effect of the calculated EHI or VGP! Might this explain in part why VGP in particular is plagued by high false alarm ratios (>80%)?
Question: If a systematic search of the historical upper air database was performed, would a superposition of low-level shear and thermodynamic profiles presented here be present in a majority of significant tornado events? Question: Would a systematic search of the historical upper air database also identify null cases?
Climatological Frequency - Hodographs ONLY:
Final Thought... Superposition of these profiles appears to be critical – NOT only the “sickle” hodograph Red = SFC – 400m agl Cyan = 400m – 1000m agl Lavender = 1000m – 7000 m agl Red = SFC – 400m agl Cyan = 400m – 1000m agl Lavender = 1000m – 7000 m agl
Acknowledgements David Andra: NWS/WFO Norman OK Michael Foster: NWS/WFO Norman OK Rich Thompson: NWS/SPC Norman OK Dr. Bob Conzemius: WindLogics Grand Rapids MN Dr. Bruce Lee: WindLogics Grand Rapids MN Doug Speheger: NWS/WFO Norman OK Kevin Scharfenberg: NSSL Norman OK Bob Johns: former SOO SPC Norman OK Jon Davies: Private Meteorologist Kansas City MO Todd Lindley: NWS/WFO Lubbock TX Dr. Chris Weiss: Texas Tech University Lubbock TX Dr. Matt Bunkers: NWS/WFO Rapid City SD Dr. David Blanchard: NWS/WFO Flagstaff AZ David Andra: NWS/WFO Norman OK Michael Foster: NWS/WFO Norman OK Rich Thompson: NWS/SPC Norman OK Dr. Bob Conzemius: WindLogics Grand Rapids MN Dr. Bruce Lee: WindLogics Grand Rapids MN Doug Speheger: NWS/WFO Norman OK Kevin Scharfenberg: NSSL Norman OK Bob Johns: former SOO SPC Norman OK Jon Davies: Private Meteorologist Kansas City MO Todd Lindley: NWS/WFO Lubbock TX Dr. Chris Weiss: Texas Tech University Lubbock TX Dr. Matt Bunkers: NWS/WFO Rapid City SD Dr. David Blanchard: NWS/WFO Flagstaff AZ
Questions? Thanks for your attention! Photo Credit: Todd Lindley