Regional Variability and Frequency of
Thundersnow over the U.S.
Kyle Meier, Lance Bosart, and Dan Keyser
Department of Atmospheric and Environmental Sciences
University at Albany, State University of New York
CSTAR Focal Point: Michael Jurewicz
National Weather Service WFO, Binghamton, NY
NROW XIV 10–11 December 2013
Importance
o Thundersnow events can produce mesoscale regions
of locally enhanced snowfall accumulations (6–12 in.)
and intense snowfall rates (2–4 in. h−1)
o Relatively little is known about thundersnow
•
•
Rarity of such storms compared to both non-thundering
snowstorms and summertime thunderstorms
Lack of direct observations inside clouds that produce
lightning
o The rarity of thundersnow events presents a
significant forecasting challenge when they do occur
Background
o Convective storms require the collective contribution of:
•
•
•
Moisture
Instability
Lift
o Fourth ingredient needed specifically for thundersnow to
occur: cold air
•
Below-freezing temperatures within clouds and near the surface
o Lightning production requires an interaction between
different types of ice in clouds
•
Separation between ice crystals (+) and graupel (−) can result in an
electric field that becomes large enough to produce an electrical spark
(i.e. the lightning stroke)
Background
 Thundersnow can occur in a variety of mesoscale and
synoptic-scale settings:
o Lake-effect
o Orographic lift
o Coastal storms and coastal fronts
o Thundersnow ahead of warm fronts (elevated
convection)
o Thundersnow in the vicinity of cold fronts (anafronts)
o Thundersnow associated with Alberta Clippers
Background
 Thundersnow can occur in a variety of mesoscale and
synoptic-scale settings:
o Lake-effect
o Orographic lift
o Coastal storms and coastal fronts
o Thundersnow ahead of warm fronts (elevated
convection)
o Thundersnow in the vicinity of cold fronts (anafronts)
o Thundersnow associated with Alberta Clippers
TSSN Climatology (Methodology)
o Objective: Construct a thundersnow climatology in order to
establish the spatial and temporal distribution of TSSN
reports across the contiguous U.S.
o Period of Study
•
•
19 years: 1994–2012
Cool season: October–March
o Dataset: Total Surface Archives (Weather Graphics
Technologies)
•
•
•
Comprehensive archive of hourly METAR surface observations
Off-hour (SPECI) surface observations also included
All valid AWOS and ASOS stations
TSSN Climatology (Methodology)
o Scan observations for all instances of TSSN, VCTSSN,
TSPL, and TSGS during the period of study
o Manually eliminate reports from Alaskan stations,
Canadian stations, and eliminate “false positives”
o Tsurface < 4°C
o Multiple consecutive reports (i.e. separated by less than
12 h) at a single station constitute one count in the
climatology
o Compile statistics
o Plot the reports spatially on a map of the U.S.
TSSN Climatology (Overview)
o 2667 reports extracted
o TSSN was reported at 680 stations in the contiguous
U.S.
o 46 of 48 states reported thundersnow (exceptions:
Delaware and Florida)
o Single, isolated TSSN reports were common
•
Reinforces the notion that TSSN is a fairly localized
phenomenon of limited duration
o Other instances where TSSN occurred at several
adjacent stations and/or for several consecutive hours
1994–2012 Thundersnow Climatology
Legend
1–2 reports
3–5 reports
6–10 reports
10–20 reports
20+ reports
1994–2012 Thundersnow Climatology
Intermountain
West
Legend
1–2 reports
3–5 reports
6–10 reports
10–20 reports
20+ reports
1994–2012 Thundersnow Climatology
Intermountain
West
Central
U.S.
Legend
1–2 reports
3–5 reports
6–10 reports
10–20 reports
20+ reports
1994–2012 Thundersnow Climatology
Intermountain
West
Central
U.S.
Northeast
Coast
Legend
1–2 reports
3–5 reports
6–10 reports
10–20 reports
20+ reports
1994–2012 Thundersnow Climatology
Great
Lakes
Intermountain
West
Central
U.S.
Northeast
Coast
Legend
1–2 reports
3–5 reports
6–10 reports
10–20 reports
20+ reports
NLDN Data to Supplement Reports
Legend
METAR report
NLDN lightning
flash
12 February 2006
Hypotheses
o The intermountain west maximum in TSSN reports is likely due
in part to orographic forcing (and lake-enhanced effects near
the Great Salt Lake)
o The maximum in the central U.S. is likely associated with the
relatively high frequency of extratropical cyclone activity
o Reports near the Great Lakes stations suggest a lake influence
(if not actual lake-effect events)
o Some East Coast events may have benefited from mesoscale
forcing provided by coastal fronts associated with coastal
cyclones and their ability to tap warm, moist oceanic air
Thundersnow Annual Distribution
800
700
19 Year Counts
600
500
400
300
200
100
0
October
November
December
January
February
Month
**Cursory analysis suggests there is not a strong diurnal
preference for TSSN to occur
March
Thundersnow Capital of the U.S.?
Copper Mountain, CO
135 reports (5.07%)
Thundersnow Capital of the U.S.?
Wolf Creek Pass, CO
114 reports (4.28%)
Rounding out the top 10…
3. Beckley, WV
50 reports
4. Salida Mountain, CO
48 reports
5. Pagosa Springs, CO
29 reports
6. Telluride, CO
28 reports
7. Ely, NV
27 reports
8. Sunlight Mountain, CO
24 reports
9. Ogden Hill, UT
19 reports
10. Bedford, MA
19 reports
Thundersnow Reports by State
1. Colorado
516 reports
2. Minnesota
379 reports
3. Illinois
219 reports
4. Oklahoma
197 reports
5. Nebraska
100 reports
6. Texas
99 reports
7. Wisconsin
98 reports
8. Michigan
84 reports
9. West Virginia
73 reports
10. New York
71 reports
**The top 10 states comprise ~68% of the total reports
Two Case Studies from the
2012–2013 Winter Season
o 8–9 February 2013
• A historic blizzard associated with a deep cyclone produced
widespread snowfall totals of 20–40 inches across parts of New
England. TSSN was reported in five states: NY, CT, MA, RI, and NH.
o 16–17 February 2013
• A strong cold front moved across the Carolinas on the morning of
16 February. Later in the afternoon, the main upper-level trough
moved across the Carolinas, resulting in a second round of
precipitation. TSSN was reported across several locations in NC
and SC.
Methodology
o Objectives: (1) Compare the synoptic-scale and
mesoscale features associated with the two events and
(2) Identify the relevant dynamical and thermodynamic
reasons for the observed thundersnow
o Datasets
•
Plan view charts (0.5° GFS data)
•
Cross-sections and proximity soundings (13-km RUC)
•
0.5° WSR-88D radar reflectivity mosaics and NLDN data
Plan View Analysis
0000 UTC 9 February 2013
0000 UTC 17 February 2013
1000–500-hPa thickness (dashed, every 6 dam), mean sea level pressure (black, every 4 hPa), and 250hPa wind speed (filled, every 10 m s−1 starting at 40 m s−1)
Plan View Analysis
0000 UTC 9 February 2013
* Coupled jet system
with a strong jet
entrance region to
the north
0000 UTC 17 February 2013
* Strong jet core (90 m s−1)
well to the east of
observed TSSN
1000–500-hPa thickness (dashed, every 6 dam), mean sea level pressure (black, every 4 hPa), and 250hPa wind speed (filled, every 10 m s−1 starting at 40 m s−1)
Plan View Analysis
0000 UTC 9 February 2013
0000 UTC 17 February 2013
* Strong jet core (90 m s−1)
well to the east of
observed TSSN
* Coupled jet system
with a strong jet
entrance region to
the north
* 984 hPa surface cyclone
SE of Long Island, NY, and
a 1036 hPa anticyclone in
Quebec
* Absence of a strong cyclone. 1024
hPa anticyclone over WI advecting
cold air into the Southeast
1000–500-hPa thickness (dashed, every 6 dam), mean sea level pressure (black, every 4 hPa), and 250hPa wind speed (filled, every 10 m s−1 starting at 40 m s−1)
Plan View Analysis
0000 UTC 9 February 2013
0000 UTC 17 February 2013
500 hPa geopotential height (black, every 6 dam), geostrophic absolute vorticity (filled, every 4 x 10-5 s−1),
and wind barbs (kts)
Plan View Analysis
0000 UTC 9 February 2013
* Northern and southern stream
short-wave troughs merge ~ 0200
UTC, coincident with a maximum in
observed lightning
0000 UTC 17 February 2013
* Deep longwave trough
associated with
lowest heights of
~528 dam
500 hPa geopotential height (black, every 6 dam), geostrophic absolute vorticity (filled, every 4 x 10-5 s−1),
and wind barbs (kts)
Radar and Observed Lightning
0230 UTC 9 February 2013
2145 UTC 16 February 2013
B
A
++
++
+
+++
+++ +
A’
B’
0.5° radar reflectivity mosaics and observed CG lightning flashes (black plus signs)
Cross-section Analysis
0300 UTC 9 February 2013
2100 UTC 16 February 2013
Cross sections of θe (red, every 4 K), absolute geostrophic momentum (black, every 10 m s−1), and relative
humidity (filled, every 10% starting at 80%)
Cross-section Analysis
0300 UTC 9 February 2013
WMSS
2100 UTC 16 February 2013
WMSS
Cross sections of θe (red, every 4 K), absolute geostrophic momentum (black, every 10 m s−1), and relative
humidity (filled, every 10% starting at 80%)
Cross-section Analysis
0300 UTC 9 February 2013
2100 UTC 16 February 2013
Cross sections of θes (black, every 4 K), negative ω (dashed, every 3 μbar s−1 starting at −12 μbar s−1),
frontogenesis [filled, every 2 K (100 km)−1 (3 h)−1], and the −10°C and −20°C isotherms
Cross-section Analysis
0300 UTC 9 February 2013
2100 UTC 16 February 2013
Mixed-phase
region
Mixed-phase
region
Cross sections of θes (black, every 4 K), negative ω (dashed, every 3 μbar s−1 starting at −12 μbar s−1),
frontogenesis [filled, every 2 K (100 km)−1 (3 h)−1], and the −10°C and −20°C isotherms
Cross-section Analysis
0300 UTC 9 February 2013
2100 UTC 16 February 2013
ω = −21 μbar s−1
ω = −33 μbar s−1
Mixed-phase
region
Mixed-phase
region
Cross sections of θes (black, every 4 K), negative ω (dashed, every 3 μbar s−1 starting at −12 μbar s−1),
frontogenesis [filled, every 2 K (100 km)−1 (3 h)−1], and the −10°C and −20°C isotherms
Sounding Analysis
0300 UTC 9 February 2013
2100 UTC 16 February 2013
Vertical temperature and dewpoint profiles at
New Haven, CT (HVN)
Vertical temperature and dewpoint profiles at
Rock Hill, SC (UZA)
Sounding Analysis
0300 UTC 9 February 2013
* CAPE: 0 J kg−1
* 700−500 hPa
lapse rate: 6.3°C
km−1
* LCL: 669 hPa
* Winds veer with
height in lower
troposphere
Vertical temperature and dewpoint profiles at
New Haven, CT (HVN)
2100 UTC 16 February 2013
* CAPE: 210 J kg−1
* 700−500 hPa
lapse rate: 7.4°C
km−1
* LFC: 940 hPa
* EL = 535 hPa
* Winds back with
height in the lower
troposphere
Vertical temperature and dewpoint profiles at
Rock Hill, SC (UZA)
Sounding Analysis
0300 UTC 9 February 2013
* CAPE: 0 J kg−1
* 700−500 hPa
lapse rate: 6.3°C
km−1
* LCL: 669 hPa
* Winds veer with
height in lower
troposphere
2100 UTC 16 February 2013
* CAPE: 210 J kg−1
* 700−500 hPa
lapse rate: 7.4°C
km−1
* LFC: 940 hPa
* EL = 535 hPa
* Winds back with
height in the lower
troposphere
* Entire
tropospheric
column below
freezing
0°C Isotherm
Vertical temperature and dewpoint profiles at
New Haven, CT (HVN)
Vertical temperature and dewpoint profiles at
Rock Hill, SC (UZA)
Conclusions
o Two thundersnow events from February 2013 occurred in
very dissimilar synoptic-scale environments
•
Case #1: NW quadrant of a strong coastal cyclone and was
associated with the merger of two shortwave troughs
•
Case #2: Post-cold-frontal environment and associated with a deep
500 hPa trough
o Similarities: Near-saturated conditions, weak MSS, and
strong updrafts in the lower-to-middle troposphere over the
range of temperatures corresponding to the mixed-phase
region of a thundercloud
Future Work
o Surface observations alone will not reveal all TSSN events
•
NLDN observations can fill in these gaps (provided the lightning
strokes are CG)
o Generate constant-pressure and vertical-profile composites of
the environment preceding and during the occurrence of
thundersnow
•
The composites may help determine the dynamical and
thermodynamic processes that contribute to regional TSSN frequency
and variability
o Conduct representative case studies of the various TSSN
pathways
o Determine discriminators between TSSN events and nonthundering snow events