The 4 Sep 2011 Tornado in Eastern New York: An Example for Updating Tornado Warning Strategies Brian J. Frugis NOAA/NWS Albany, NY NROW XIII 2-3 November 2011
Dec 14, 2015
The 4 Sep 2011 Tornado in Eastern New York: An Example for Updating Tornado Warning
Strategies
Brian J. FrugisNOAA/NWS Albany, NY
NROW XIII 2-3 November 2011
CSTAR IV & Motivation for Study
• New Tornado Climatology was developed over the summer of 2010 for the Northeastern US with CSTAR work by undergrad UAlbany students
• A goal of CSTAR IV is to develop/update tornado warning strategies using the new 8 bit high resolution radar data– V-R shear relationship has been an effective method
for predicting tornadic development, however, this is based off of 4-bit radar data
Updated Tornado Climatology by CWANumber of Tornadoes for 1981 – 2010 Normal Period
100
47
19
23
6312467
222205
81
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec0
5
10
15
20
25
30
35
40
45
50
Number of Tornadoes in ALY CWA by month from 1981-2010
Number of tornadoes from 1981-2010=100 Average number per year is 3.33
F0/EF0 F1/EF1 F2/EF2 F3/EF3 F4/EF4 F5/EF50
10
20
30
40
50
60
Tornadoes by Ranking in ALY CWAfrom 1981-2010
Number of Tornadoes = 100
79% are weak (EF0/EF1), 16% are strong (EF2/EF3) and only 5% are violent(EF4/EF5)
Previous Work
• LaPenta (2000) led a COMET study on V-R shear relationship for Northeastern United States tornadoes– Collaborate project between UAlbany and NWS
Albany• This study created nomograms for operational
use based on a linear relationship between gate-to-gate shear and the strength of the rotational velocity of the mesocyclone
V-R Shear Relationship
• Maximum observed gate-to-gate shear below 3 km was found to be useful in identifying tornadic storms (LaPenta et al. 2000)
• S=Vr/(D*1800)• Shear (S) is measured in units of s-1, rotational
velocity (Vr) in knots and D is the diameter of which S is calculated in n mi.
Accounting for Varying Shear Values
With 4-bit data, adjacent pixels are 0.5 n mi apart within 30 n mi from the radar.
This distance becomes 1.0 n mi at 60 n mi away from the radar due to beam spreading
Because of this, D is variable depending on range from radar and must be normalized
In the LaPenta study, D was set to 0.5 n mi for areas within 30 n mi of radar and adjusted for areas further away (LaPenta et al, 2000)
Maximum Velocity Differential of Mesocyclone (Vm)
• Determine strength of large-scale mesocyclone
• Based off “Mesocyclone Recognition Guidelines” (Andra et al. 1994)– Uses a 3.5 nm
mesocyclone width Image from 1998 Mechanicville, NY F3 Tornado, courtesy of LaPenta, et al. 2000
Legacy 4-bit V/SRM Products
• Resolution: 1 km (0.54 nm) by 1 degree
• Range: 124 nm • 16 data levels: -64
kts to +64 kts
An example of 4-bit reflectivity (Z) image from KENX from 2123z 4 Sept 2011
8 Bit V/SRM Products
• Introduced in late 2002/early 2003 with AWIPS build 5.2.1
• Resolution: 0.25 km (0.13 nm) by 1 degree
• Range: 124 nm (no change)
• 256 data levels (28): -123 kts to +123 (using standard setup) An example of 8-bit reflectivity (Z) image from
KENX from 2123z 4 Sept 2011
Super Resolution 8 bit V/SRM Products
• Introduced in Spring-Summer 2008 with RPA/RDA Build 10.0
• Resolution: 0.25 km (0.13 nm) by ½ degree
• Range: 162 nm• 256 data levels (28): -
123 kts to +123 (using standard setup) An example of 8-bit super resolution reflectivity
(Z) image from KENX from 2123z 4 Sept 2011
Advantages of New High Res Data
• With the 8-bit high res data, we no longer have to adjust D for range when using the V-R shear technique– This is because going from the mid point to mid point of
adjacent pixels is 0.5 n mi or less for up to 60 n mi from the radar
• The higher resolution data allows more subtle features to be resolved– Higher velocity values can be calculated (up to 128 kts
in normal setup for 8 bit as compared to 64 kts for 4 bit)
4 Sept 2011 Tornado
• Rated EF1 with top winds of 110 mph by NWS Albany Survey Team
• Formed at 2120z near Florida, NY
• Up to a half-mile wide through Cranesville, NY
• Dissipated at 2135z near Glenville, NY
• On the ground for 7 miles KENX 0.5° Z Loop from 4 Sept 2011 between 2109z and 2146z
KENX 0.5° Reflectivity Loop w/TVS KENX 0.5° Spectrum Width Loop
Using Spectrum Width as an Indicator of Tornadic Development
Calculating Legacy V-R Shear
• 4 bit KENX SRM 4 Sept 2011 – 2123z
• Since distance from RDA=20 nm, shear can be measured with 0.5 n mi diameter
• Vr =43.0 kts• S=0.0450 s-1
Maximum Velocity Differential of Mesocyclone (Vm)
Tornadic Circulation
MesocycloneCirculation
SRM Cross-Section from KENX Radar 2123z 4 Sept 2011 – About 20 n mi from RDA
Maximum Velocity of Mesocyclone (Vm) = 55 kts
Image is from Vertical Slice Pane from AWIPS FSI
4 Sep 2011 2123Z KENX SRM
Legacy 4-bit: (1 km x 1 degree)Super Resolution 8 bit: (0.25 km x 0.5 degree)
Vr=43.0 kts, D=0.5 n mi, S=0.0450 s-1 Vr=53.9 kts, D=0.5 n mi, S=0.0626 s-1
Comparing Storm Types
• LaPenta et al. (2000) found that supercells were responsible for producing tornadoes in about half (49%) of the storms examined
• 67% of the storms were associated with a bow echo and 19% had a boundary interaction– The 4 Sept 2011 storm is a good example of a
boundary interaction and a bow echo as well
Future Work
• Higher resolution data will require a new nomogram to be created– Scale of shear values (S) will likely change
• With 8 bit data, gate-to-gate shear can be calculated without having to adjust for range from radar– Will allow for faster identification and quicker warnings
• Future work over the next year will study all tornadoes since start of 8-bit data– Also, null cases (non-verified TOR) will be examined as well
• Storm type will be examined as well to see if our trends fit what LaPenta et al. found in 2000
Limitations/Potential Issues
• Knowing what was really a tornado can be somewhat subjective– Weak tornadoes common to this area don’t always show
much different damage than thunderstorm winds/microbursts– NCDC StormData doesn’t allow for “landspouts”
• Radar doesn’t always show rotation– Not all classified tornadoes show classic rotation couplet on
radar• Radar data can be limited or affected by terrain
– Beam may overshoot low-level features farther away from RDA
– Many parts of the Northeast are affected by beam blockage
Items to Track in Spreadsheet for Study
• Date/Time/Location of Tornado• Radar Site Used/VCP/Range from Radar• Vr, S, Vm
• Storm Type and Tornado Formation– Boundary Interaction? Supercell with hook echo?
• Meso/TVS Present?• Other Unique/Curious Items– Large Hail, Strong LLJ, Max Reflectivity, etc.