ATMOS-02781; No of Pages 11

Atmospheric Research xxx (2012) xxx–xxx

Contents lists available at SciVerse ScienceDirect

Atmospheric Research
journal homepage: www.elsevier.com/locate/atmos

Continuous variability in thunderstorm primary electrification and an

evaluation of inverted-polarity terminology
Eric C. Bruning a,⁎, Stephanie A. Weiss b, Kristin M. Calhoun b
a
Texas Tech University, Dept. of Geosciences, Atmospheric Science Group Texas Tech University, Box 42101, Lubbock, TX 79409–2101 USA
b
Cooperative Institute for Mesoscale Meteorological Studies / University of Oklahoma 120 David L. Boren Blvd., Norman, Oklahoma 73072, USA

a r t i c l e i n f o a b s t r a c t

Article history: Several field campaigns since the year 2000 have focused on anomalously electrified or
Received 16 February 2012 “inverted polarity” thunderstorms. This study synthesizes these recent results, and considers
Received in revised form 18 September 2012 how variability in the non-inductive relative-growth rate electrification mechanism might
Accepted 5 October 2012 clarifying the meaning of “inverted polarity”. Instead of falling into two polarity classes,
Available online xxxx
electrification and charge structure in strong updrafts vary continuously, as expected if depletion
of supercooled water is a primary control on electrification. Two- or three-dimensional storm
Keywords: flows or other electrification mechanisms are required to combine one or more of these elec-
atmospheric electricity trification regimes into “inverted” or otherwise complicated local charge sequences. Cloud flashes
lightning
that result from these local charge sequences should be termed “positive” and “negative” instead
thunderstorm
of “normal” and “inverted” because cloud flashes of either polarity can occur at any altitude in
inverted polarity
electrification thunderstorms.
© 2012 Elsevier B.V. All rights reserved.

1. Notions of storm polarity is enhanced as supercooled water is collected. The NI-RGR

Is the notion of inverted-polarity meaningful in explaining
observations of electricity in thunderstorms? Of particular
concern is what physical observable is inverted. This paper
proposes a distinction between properties (such as charge
structure, electrification, and flashes, as well as storms as a
whole) that vary continuously and those that can be bifurcated
meaningfully into two polarity classes.
1.1. Electrification and Charge Structure
The primary driver of thunderstorm electrification is
thought to be non-inductive (NI) rebounding charging be-
tween ice hydrometeors in the presence of supercooled
water (Saunders et al., 2006; Emersic and Saunders, 2010).
It is thought that the relative growth rate (RGR, Baker et al.,
1987) of the two ice hydrometeors from the vapor phase
controls the sign of charging. The relative growth rate effect
⁎ Corresponding author. Tel.: +1 806 834 3120.
E-mail address: eric.bruning@ttu.edu (E.C. Bruning).
mechanism can readily explain the production of a proto-
typical “normal-polarity” tripolar charge structure (Simpson
and Scrase, 1937; Simpson and Robinson, 1941; Williams et
al., 1989), but is also effective at producing charge structures
that have been more recently referred to as “inverted-polarity”
(Marshall et al., 1995; Rust and MacGorman, 2002; Rust et al.,
2005; Wiens et al., 2005; Kuhlman et al., 2006; MacGorman et
al., 2005; Tessendorf et al., 2007; Carey and Buffalo, 2007;
Weiss et al., 2008; Tessendorf, 2009; Bruning et al., 2010).
In the normal-polarity model, precipitation carries nega-
tive charge in the midlevels of the storm and positive in
lower parts of the storm. Non-precipitating cloud ice carries
positive charge in the upper parts of the storm, and negative
charge in the midlevels of the storm. The negative regions
typically combine into a large net negative charge region
midway through the depth of the storm. A tripolar structure fits
with observations that most storms lower negative charge in
ground strikes. While observations have clearly shown that
this model in isolation is inadequate to explain all features
of in-situ measurements of the electric field in storms (Rust
and Marshall, 1996; Stolzenburg et al., 1998), it is usually

0169-8095/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.atmosres.2012.10.009

Please cite this article as: Bruning, E.C., et al., Continuous variability in thunderstorm primary electrification and an evaluation of
inverted-polarity terminology, Atmos. Res. (2012), http://dx.doi.org/10.1016/j.atmosres.2012.10.009
2 E.C. Bruning et al. / Atmospheric Research xxx (2012) xxx–xxx
possible to find at least vestigial evidence of the electrification
mechanism in the form of a tripolar-like charge sequence.
The most likely place in which to find evidence of
tripolar-like structures is in 1-dimensional storms with
strong cylindrical symmetry, where precipitation trajectories
most closely follow the classic updraft-downdraft model of
(Byers and Braham, 1949). Even in storms with more
complicated flow regimes, it is usually possible to identify
a region where the boundary layer thermodynamic state
is most efficiently processed through the storm, i.e., the
“updraft core.” This region maps well on to Stolzenburg et
al. (1998)'s updraft profiles, where the NI-RGR mechanism is
strongly implicated in producing the lower three charges in
their conceptual model.
In a review of the inverted-polarity studies cited above,
Tessendorf (2009) defined inverted-polarity storms as those
exhibiting, at least somewhere within the storm, a tripolar
structure that was inverted from the normal sequence. In
updraft regions, an inverted (from the two uppermost charges
in the tripolar sequence) dipole was often observed, with
the lower negative charge of an inverted tripole absent or
reduced. If these storms produced ground strikes the majority
were observed to be of positive polarity. This study seeks to
clarify how it is that such charge sequences can become
inverted from the normal-polarity tripolar sequence given our
current understanding of the NI-RGR electrification mechanism.
We do not intend a complete characterization of thun-
derstorm charge structures in this study. Instead, our focus
is primarily on the NI-RGR electrification mechanism and
evidence of its operation in production of charge structures
within the updraft core, in a sequence of charge that some
might consider tripolar (e.g., (Rust and MacGorman, 2002;
Rust et al., 2005; Carey and Buffalo, 2007). The relative
depletion rate of cloud water has been linked to inverted
polarity hypotheses (Williams et al., 2005), and the choice
to focus on the updraft core also allows us to assume that
the relative growth rate effect is dominated by supercooled
liquid collection, and not lesser effects that might operate
in clouds at ice supersaturation in the absence of cloud liquid
water (e.g., Mitzeva et al., 2006).
We restrict our analysis of the electrification in inverted
polarity storms to a single electrification mechanism and
the tripolar charge structure that it can explain, because
tripolar language tied to the NI mechanism still dominates
wider understanding of thunderstorm charge structure and
the language used to describe it — consider its continued
appearance in introductory (Ackerman and Knox, 2007;
Aguado and Burt, 2010) and advanced (Williams, 2001)
meteorology and storm electricity (Tessendorf, 2009) texts.
By pointing out some problems with the normal/inverted
dichotomy (which implicitly references the tripole), this paper
acknowledges problems with the way a tripolar baseline is used,
and contributes to the ongoing search for a simple framework
that accounts for observed charge structures and links them in
a clear way to one or more electrification mechanisms.
The primary evidence discussed in this study comes from
recent results from the Severe Thunderstorm Electrification and
Precipitation Study (STEPS, Lang et al., 2004), the Thunderstorm
Electrification and Lightning Experiment (TELEX, MacGorman
et al., 2008), Carey et al. (2005) and Albrecht et al. (2011). These
studies have shown that storms that produce predominantly
Please cite this article as: Bruning, E.C., et al., Continuous variability in thunderstorm primary electrification and an evaluation of
inverted-polarity terminology, Atmos. Res. (2012), http://dx.doi.org/10.1016/j.atmosres.2012.10.009
positive ground strikes have shallower warm cloud depth and
more vigorous updrafts that enhance positive charging to
graupel (Williams et al., 2005). Charge inferred from in-situ
electric field and lightning mapping measurements from the
STEPS and TELEX campaigns also confirm that some form of
inverted-polarity electrical structure is present in at least some
part of the storms that produce predominantly positive ground
strikes. In these structures, the first two net regions above the
ground are inverted from those of the normal tripole. Based on
energetic arguments, the above authors have argued that the
enhanced positive charging to graupel leads to an enhancement
of the positive charge region that is the source of charge
lowered in positive ground strikes.
Recent studies also suggest that a 2D or 3D storm flow is a
minimum requirement for producing the locally inverted
structures necessary for positive ground strikes. This is
because, as we show below, the electrification mechanism
always produces a normal-polarity-like structure with pos-
itive charge lowest. Put another way, the basic non-inductive
graupel-ice electrification mechanism does not produce a
lower negative charge region in a 1D storm flow where positive
charging to graupel is enhanced, and so the interplay between
regions of simultaneously enhanced and less enhanced positive
charging rates to graupel are important.
1.2. Flashes
1.2.1. Cloud flashes
Normal- / inverted-polarity terminology has also been
used to describe cloud flash polarity. Historically, cloud
flashes were thought to be between the upper positive and
midlevel negative charge regions in the normal-polarity
tripole, (these regions are the positive dipole in the simplest
charge models, e.g., Wilson, 1916, 1925), and as such were
referred to as normal-polarity cloud flashes. The lower
positive charge center in the normal tripole was thought to
be weaker, with flashes between it and the main negative
charge center preferentially coming to ground. Low-level
flashes that did not come to ground would therefore have a
vertical dipole orientation that is inverted from upper level
cloud flashes in the normal tripole.
In storms with enhanced positive charging to graupel and
associated elevated tripole, inverted flashes become increas-
ingly common, though they remain tethered to the lower
negative dipole formed in the NI charging process. Mansell
et al. (2010) argued that the relative amounts of charge in
the tripole can vary significantly, leading to top-heavy or
bottom heavy tripole structures in normal-polarity storms.
VHF lightning mapping array data show that bottom-heavy
tripole structures often produce low-level inverted-polarity
flashes in a negative dipole. The terms normal and inverted,
while helpful shorthand for describing polarity, only indicate
“normality” relative to the previous paradigm of cloud flash
understanding.
It seems that a new characterization of the relative
frequency of normal and inverted-polarity cloud flashes is
necessary to declare normality of either polarity of cloud
flash. Likewise, the presence of cloud flashes between a
negative dipole is not necessarily evidence of an inverted-
polarity storm.
 E.C. Bruning et al. / Atmospheric Research xxx (2012) xxx–xxx
1.2.2. Ground strikes
When channel development begins within the cloud,
the polarity of ground strikes is often controlled by the
polarity of the first two charges layers near the earth's surface
(Mansell et al., 2002; Kuhlman et al., 2006; Wiens et al.,
2005; Tessendorf et al., 2007; Weiss et al., 2008; Tessendorf,
2009), though Krehbiel et al. (2008) discuss several other
discharge modes that do not fit this pattern. In most cases,
negative charge will be lowered from cloud to ground (− CG)
where the lowest charge region is positive, and positive
charge will be lowered (+CG) where the lowest charge
region is negative. Whether or not a flash comes to ground is
determined by the net energetics of the charge configuration
(Krehbiel et al., 2008; Mansell et al., 2010). The local charge
sequence provides a strong constraint on the local ground
strike polarity, since a charge region with polarity opposite
that of the eventual ground strike is required to bring the
leader sufficiently close to the surface.
A storm with a mix of inverted and normal local charge
structures could produce a continuous variation in the ratios
of ground strike polarities. However, a predominance of one
polarity may reflect net storm energetics, though the degree
of predominance must be of some importance.
In some cases, flashes may also begin at the ground, with
only a single polarity of leader traveling upward to discharge
the cloud. Ground surface irregularity (for instance, tall radio
or television towers) would to tend favor the development
of such flashes by locally enhancing the field to breakdown
thresholds. These flashes would produce ground strikes of
the opposite polarity from those discharges that begin in the
cloud.
2. Evidence from storms
Through the noninductive/relative-growth rate charging
mechanism (Saunders et al., 2006; Emersic and Saunders,
2010) it is understood that the polarity of charging in and
surrounding the main updraft core is dependent upon the
local environment in which the riming graupel and smaller
ice particles interact. Individual case studies and charge
analysis of various storms have produced a continuum of
charge profiles, but each can also provide some information
about the charging that took place in the updraft relative to the
local environment. The depth of the warm-cloud, mixed-phase,
and glaciated cloud regions, as well as trajectories surrounding
and away from the updraft core all play significant roles in
overall charge structure of a storm.
The updraft core is defined here as the trajectory from the
boundary layer to cloud top that most directly processes the
boundary layer moisture and thermal energy, and realizes
the most convective available potential energy (CAPE). The
difficulty in sampling a storm with sufficient spatial detail to
quantitatively resolve these parameters usually requires an
inference about the location of the updraft core from storm
structure. The updraft core is the most thermodynamically
“pure” region of the storm, with rather less contribution
of advected or entrained charge from other regions and a
dominance of the influence of environmental thermodynam-
ics. As such, it represents the best place to test ideas about
how storm environments affect storm charge.
Please cite this article as: Bruning, E.C., et al., Continuous variability in thunderstorm primary electrification and an evaluation of
inverted-polarity terminology, Atmos. Res. (2012), http://dx.doi.org/10.1016/j.atmosres.2012.10.009
3
Observational studies of the updraft core region in storms
occurring prior to the year 2000 are well-summarized by
(Stolzenburg et al., 1998). One notable exception to their
results was the Dalhart, TX storm studied by Marshall et al.
(1995), which the authors noted was inverted from their
other observations. Lack of radar data and relatively slow
observed ascent rates prevented the electrical observations
from being placed in the storm context. The observational
data sets from both the STEPS and TELEX field projects
allowed for in-depth examination of charge profiles com-
bined with corresponding radar and environmental charac-
teristics of the storms (Lang et al., 2004; MacGorman et al.,
2008). Many of the storms included here occurred during
these field programs, and are combined with other recent
investigations that mention inverted polarity storms. A review
of results from these studies are included below with some
new results from cold season storms and –CG dominant
supercells. They are classified by storm type in the discussion
that follows.
2.1. Recent studies of updraft core charge structure
Rust and MacGorman (2002), Rust et al. (2005) and
Tessendorf et al. (2007) investigated low-precipitation (LP)
supercellular thunderstorms. The first of these storms was
an LP, marginally-supercellular storm on 4 June 2000 near
Bird City, KS. It was not nearly as electrically active as other
storms included in this study. The flashes that did occur
initiated near 10 km between a broad region of positive charge
(graupel) and upper region of negative charge (crystals). This
small LP storm produced no CG flashes.
Storms on 29 June 2000 and 26 May 2004, examined by
MacGorman et al. (2005), Wiens et al. (2005), Kuhlman et al.
(2006), Tessendorf et al. (2005), Bruning et al. (2010), can
be considered more typical of classic supercells, though both
produced predominately + CG flashes as a result of the
development of a lower layer negative charge region at
some distance from the updraft core. The 29 June 2000
supercell storm occurred in northwestern Kansas. It began
with a dipole charge structure with negative charge upper-
most before expanding to over five vertically sequenced
charge layers outside the main updraft core as the flash rate
increased from 10 flashes per min to nearly 300 flashes per
min over a 180 min period, both in observations and model
simulations (Wiens et al., 2005; Kuhlman et al., 2006).
Evident in the numerical simulations of this storm was the
combination of positive and negative charging to graupel in
and surrounding the updraft core, with positive charging to
graupel in the regions of higher cloud water content in the
core and negative in the surrounding periphery (Kuhlman et
al., 2006). The 26 May 2004 supercell in central Oklahoma
also exhibited a similarly complex charge structure. While
the main updraft core consisted primarily of three charge
layers (with positive above and below a middle negative),
the region to the north of the updraft core had up to six
charge layers. Another region of the same storm, east of the
updraft core into the forward-flank downdraft, had at least
four active layers of charge with negative charge at the
bottom; the +CG flashes associated with the storm were all
linked to this forward-flank region and lowest layer of
negative charge (Bruning et al., 2010, Fig. 8,).
4 E.C. Bruning et al. / Atmospheric Research xxx (2012) xxx–xxx
A high-precipitation (HP) tornadic supercell in central
Oklahoma on 29 May 2004 occurred in an environment with
CAPE in excess of 3000 J kg -1 (MacGorman et al., 2008;
Kuhlman, 2010). This storm had extremely high flash rates
approaching nearly 500 flashes per min at its peak. The
LMA-inferred charge structure primarily consisted of a large
positive charge region between 7 and 10 km below an upper
negative charge region. However, the large number of small,
compact flashes surrounding the main updraft core made
charge analysis for this storm extremely difficult. Simulations
of the storm point to a combination of positive (higher liquid
water contents in the updraft core) and negative charging of
graupel (surrounding the updraft core) with higher non-
inductive charging rates than seen during past simulations of
classic or multicell storms. While the storm was dominated
by –CG flashes, a significant percentage of + CG flashes
also occurred, and the polarity was directly relatable to
the corresponding local charge structure (similar to the 26
May 2004 studied by Bruning et al., 2010). The rear flank
downdraft and hook echo region were dominated by –CG
flashes, while roughly 30-50% of the ground flashes in the
forward flank downdraft lowered positive charge to ground.
The multicell storms that have been examined all also
depict the continuum of charging throughout the life-cycles
of the multiple updraft cores. Bruning et al. (2007) and Mansell
et al. (2010) investigated a tropical air-mass multicellular
storm that was dominated by negative ground strikes. This
particular storm from 29 June 2004 was not as electrically
active as the classic and HP supercells discussed above,
with only 30 flashes occurring in 40 min. The early pulse of
the main updraft produced a positive charge near 4 km
(graupel) with a region of negative charge above it between
4.5 and 6 km. This dipole structure was associated with the
initiation of –CG lightning in both the observations and
simulations. A later, more-vigorous, updraft pulse led to a
growth in the height of the storm through at the least the −10
to −20 °C level, an increase of graupel volume and upward
crystal flux, and the formation of an upper layer of positive
charge and enhancement of the mid-layer negative charge
region with the onset of lightning initiations near 8 km.
Weiss et al. (2008) examined a multicellular hailstorm
observed during the STEPS field program on 25 June 2000
near Haigler, Nebraska. The main body of the storm had
four vertically-sequenced charge regions in a typical arrange-
ment for a well-developed updraft (a normal tripole with a
probable negative upper screening layer above it). However,
two newer regions of convective development were also
present along outflow boundaries that originated from dying
storms to the southeast of the main storm. The new develop-
ment along the outflow boundaries consisted of a deep region
of positive charge beneath a thinner region of negative charge.
These two charge regions are consistent with positive charging
of graupel in a deep layer where a supercooled-liquid-rich
mixed phase region was enhanced by a modified outflow-
boundary air mass with a positive water vapor perturbation. As
the convection along the outflow boundaries began to interact
with the convection from the main storm (around 0224 UTC),
the lightning structure became more complex just as the charge
structure becomes more complex in supercells further from the
updraft core. The storm on 25 June 2000 developed a stratiform
region in which the charge structure was complicated by the
Please cite this article as: Bruning, E.C., et al., Continuous variability in thunderstorm primary electrification and an evaluation of
inverted-polarity terminology, Atmos. Res. (2012), http://dx.doi.org/10.1016/j.atmosres.2012.10.009
advection of charge from the main updraft region. This is
typical in mesoscale convective systems (MCSs) and is also
analogous to advection of charge into the forward anvils
of supercells (Kuhlman et al., 2009; Weiss et al., 2012).
The net result of these horizontal environmental thermody-
namic variations as well as the interaction of the differing
electrification regimes was several simultaneous charge
structures, some of opposite polarity at the same time at
the same altitude.
Emersic et al. (2011) analyzed the 15 August 2006
multicellular hail storm in central Oklahoma by combining
the LMA data with rapid scan information available from
phased array radar. They noted three phases related to the
growth of the storm. The first phase consisted of early storm
growth and interaction with existing convection from an
older dissipating storm, with a charge structure consisting of
a midlevel negative charge above a lower positive charge.
During this first phase the lower positive charge grew in size
and increased in height at the same time the height of the
peak radar echoes increased, implying positive charge to
graupel and larger ice in this region. The second phase began
6 min later, where an updraft pulse led to a large increase
in flash rate over the next 4 min to 200 flashes per min.
A three-body-scatter spike appeared, signifying wet hail
growth. The charge structure expanded to include an upper
positive charge region with flashes now initiating between
the upper two charge regions and the lower charge regions,
with two –CG flashes included in the lower two charge
regions. A third phase of peak lightning activity began 15 min
after the first pulse, and lasted for 10 min as the storm fully
realized all the CAPE in the environment. A new updraft
surge produced a different charge structure in the main
updraft core, consisting of a upper-level negative charge
above 11 km (ice crystals) over a large midlevel positive
charge (graupel), with lightning in the main updraft region
now limited to higher altitudes. Outside this region, the old
and new charge structures combined to provide a complex
charge structure including opposite polarity charge at the
same horizontal elevation. In the third stage the storm
produced four + CG flashes and no –CG flashes. LMA leaders
appeared to propagate around the side of the hail shaft
and through adjacent inferred negative charge, avoiding the
region of wet hail growth.
Rust and Trapp (2002) examined electrified winter nimbo-
stratus in Utah. If the differences in the observed temperature
profiles are accounted for, one finds charge structures that
could be compatible with the tripole produced by the non-
inductive relative-growth-rate mechanism, though the authors
did not discuss this possibility. Preliminary work by Kuhlman
and Manross (2011) and Schultz et al. (2011) using VHF
lightning mapping data is compatible with the earlier result
of Rust and Trapp (2002). When compared to warm-season
deep convection, the charge structures were closer to the
ground, due to the cold column. The cold column and slow
vertical velocities also limit the availability of supercooled
water. In cases with particularly cold columns and little upright
convective cellular structure (such as banded precipitation to
the northwest of synoptic-scale low pressure systems), the
lower positive charge center was not as apparent, perhaps
because the region of positive charging to graupel had been
shut down completely.
 E.C. Bruning et al. / Atmospheric Research xxx (2012) xxx–xxx 5

2.2. New case study: 6 Feb 2008

To add diversity to the set of supercells studied with an

LMA and weather radar data, analysis of the updraft core
charge structure of another supercell storm dominated by
negative ground strikes is desirable. This study presents data
from an HP supercell storm on 6 Feb 2008 over northern
Alabama with these characteristics. An HP storm would
be expected to produce more precipitation by warm rain
processes than a classic or LP supercell, while having a
more-vigorous updraft and higher likelihood of wet growth
of precipitation in the mixed-phase region than a tropical
multicell storm. This study follows the analysis methodology
of Bruning et al. (2010), which used storm structure inferred
from radar reflectivity and the presence of new clusters or
arcs of lightning to identify the “updraft core.”
The large-scale meteorological environment of the storm
in question is described further in Darden et al., 2010; their
study focused primarily on the southern, Jackson County
storm, while the analysis below focuses on the northern,
Lawrence County storm. While no in-storm soundings are
available, the 00 UTC 6 Feb 2008 sounding from Birmingham,
AL indicates the freezing level was at 4.0 km for a moist
adiabatic parcel, and that the environment was characterized
by less than 1000 J kg-1 of CAPE.
Fig. 1 shows one-minute snapshots of lightning source
counts from the storm. At 0942 UTC, a new cluster of
lightning activity forms on the southern flank of the storm.
The source count (related to the flash rate) and extent of
this cluster increases with time. By 0953 UTC the peak source
count begins to approach the peak source count of the earlier,
mature cluster of lightning activity to the north of the new
cluster. Fig. 1. Time history of VHF source density for the 6 Feb 2008 storm in
All lightning flashes in the 0942 UTC cluster were analyzed Northern Alabama. Each panel shows one minute of data in plan view,
(Fig. 2), and a normal-polarity tripole structure was inferred. reading from left to right and down. The charge analysis in Fig. 2 follows the
cluster of lightning activity that develops at 0942 UTC at (−2 km East, 7 km
Most flashes were cloud flashes between negative charge
North) and moves to (5 km, 15 km) by 0951 UTC.
centered at 6 km altitude, and positive charge centered at
8–9 km altitude. There were about twenty cloud flashes
between the lower charge pair, with the lower positive parameter space into two regions that separate the polarity
charge centered at 4 km altitude. of charging to graupel, with warmer and wetter regions
As in the case analyzed by (Bruning et al., 2010), the new charging positively. The water content axis might instead
cluster of lightning activity developed in vaulted precipita- be rime accretion rate (Saunders and Peck, 1998) or another
tion above a low-level weak echo region (Fig. 2). The upper suitable parameter, and the position of the reversal line might
cloud flashes were located primarily in this lofted precipitation. vary, depending on entrainment and mixing conditions
Low-level flashes that propagated into the lower positive experienced by the parcel as it rises (Saunders et al., 2006).
charge region indicated that the lower positive charge center The portion of the parameter space near freezing with
was displaced to the north, closer to or within the mature abundant supercooled water is the most likely region in
precipitation shaft that was already present when the new which wet growth conditions would be found; these condi-
cluster formed at 0942 UTC. tions would inhibit electrification by preventing rebounding
collisions (Emersic et al., 2011).
3. Processes resulting in storms’ net charge structures Fig. 3 also plots conceptual scenarios for several updraft
trajectories through the parameter space, each starting with
3.1. The electrification continuum the same amount of liquid water content. The trajectories
reflect varying rates of depletion of supercooled liquid
Each of the storms above provides some information available for riming, and visually depict recent hypotheses
about the charging that took place in the updraft core relative for positive-strike dominant storms (Williams et al., 2005;
to temperature. Fig. 3 illustrates a method of thinking about Carey and Buffalo, 2007). Some parcels deplete faster as
electrification in the updraft core that explicitly depicts warm processes deplete liquid water at a rapid rate, which we
the liquid water depletion rate as a parcel rises and cools in consider to be most similar to tropical multicellular or mountain
the updraft. The foundation of the diagram is a reversal line thunderstorm convection due to the role that warm (non-ice
that partitions an effective water content and temperature phase) precipitation growth plays in depleting low-level

Please cite this article as: Bruning, E.C., et al., Continuous variability in thunderstorm primary electrification and an evaluation of
inverted-polarity terminology, Atmos. Res. (2012), http://dx.doi.org/10.1016/j.atmosres.2012.10.009
6 E.C. Bruning et al. / Atmospheric Research xxx (2012) xxx–xxx

16
Inferred
Storm Charge
+ – ?
12
Altitude (km)

0
0943 0944 0945 0946 0947 0948 0949 0950 0951 0952 0953 0954
Time (UTC)

14

10
Cloud
Altitude (km)

B
25
6
35

35
A 2

50 55
-20 -10 0
A North Distance (km) B
Fig. 2. Updraft core charge structure inferred from VHF Lightning Mapping Array (LMA) data on 6 Feb 2008 for a storm in Northern Alabama. Red, blue and gray
points indicate positive, negative, and unanalyzed storm charge. (top) Time-height view of inferred charge for all flashes from 0942–0954 UTC in the cluster
described in Fig. 1. (middle) 0.9° elevation angle scan from the Hytop, AL WSR-88D radar overlaid with coincident LMA data, 0952:14–31 UTC. Red, blue and
green points indicate positive, negative, and unanalyzed storm charge inferred from data. Triangles indicate location of negative ground strikes from the National
Lightning Detection Network. Location of this scan is part of a wider view (bottom left) of the storm. A simplified summary of radar reflectivity contours and inferred
gross charge regions in the vertical cross-section A-B are shown (lower left), with positive, LMA-inferred storm charge in pink and negative charge in blue.

liquid water. The slowest depletion rate shares many similar- deep into the mixed phase region. The available cloud water
ities with storms with a high cloud base, where most of is shown to fully deplete at very cold temperatures for all
the available boundary layer liquid water content is maintained trajectories, consistent with the idealized processes implied by

Please cite this article as: Bruning, E.C., et al., Continuous variability in thunderstorm primary electrification and an evaluation of
inverted-polarity terminology, Atmos. Res. (2012), http://dx.doi.org/10.1016/j.atmosres.2012.10.009
 E.C. Bruning et al. / Atmospheric Research xxx (2012) xxx–xxx
Variation of cloud water and
temperature in updraft core
4
Effective Water Content g m-3
3 The early- and late-stage tropical multicellular storm
2
Wet growth or
precip ice absent
1
+ to graupel – to graupel
0 the cloud; in all winter cases, weak updrafts provide little
0 -10 -20
Temperature oC
Fig. 3. Idealized variation of effective riming rate (a proxy for cloud water
content) with temperature in storm's updraft trajectory as non-precipitating
cloud water is depleted by precipitation growth processes (dotted line).
Polarity of charge gained by graupel is separated by a solid line, which is
one of many possible reversal lines from Saunders et al. (2006). Initially the
trajectories traverse a part of the temperature / water content space where
little precipitation-sized ice is present, or where wet growth retards
rebounding collisions (dashed line). The bottom dotted curve represents
the fastest initial depletion of liquid water, while the upper curve represents
the slowest initial depletion. Solid black dots indicate the reversal temperature
at which a given trajectory crosses from positive to negative polarity.
measures like equivalent potential temperature. The slower
the depletion of liquid water, the colder the temperature at
which each curve depicted in Fig. 3 crosses (black dots in the
figure) the electrification reversal temperature line. Therefore,
Fig. 3 shows how a shift of the reversal line to a colder
temperature (higher altitude) for a given updraft leads to a
deeper region of positive charging.
Fig. 4a extends the ideas of Fig. 3 from the electrification
parameter space into an explicit tripolar charge structure
produced by the NI electrification mechanism. Each of the
four updraft trajectories from bottom to top in Fig. 3 cor-
responds, generally speaking, to left-to-right movement
across Fig. 4a. The charge structures on the left correspond
to conditions poor in supercooled liquid. In the middle
reside moderate depletion rates with possible wet growth.
On the right are supercooled-water-rich conditions in
storms with high cloud bases or small warm cloud depth.
The charge structure is assumed to arise entirely from within
the mixed phase region, though warm cloud and fully glaciated
cloud regions are also indicated for reference. Updraft charge
structures inferred in various observational studies are also
indicated on the plot. In many cases, individual storms are
represented by two or more charge structures that were
present at different stages or storm-relative positions of the
updraft core in the storm.
It is interesting to note how the storms summarized earlier
span this diagram, ranging from weak tropical multicellular
convection on the left to stronger multicellular or supercellular
thunderstorms on the right. The results show that the tripolar
electrical structure due to the NI charging mechanism can be
inferred to operate in all storms, and varies continuously from
storm to storm. The interplay of convective available potential
energy (CAPE), available liquid water, pre-existing hydrome-
teors, and precipitation trajectories (including the influence
of wind shear) all contribute to the liquid depletion rate around
which this diagram is built. These are familiar parameters
Please cite this article as: Bruning, E.C., et al., Continuous variability in thunderstorm primary electrification and an evaluation of
inverted-polarity terminology, Atmos. Res. (2012), http://dx.doi.org/10.1016/j.atmosres.2012.10.009
7
from Carey and Buffalo (2007). For warm-season convection
there is a progression from moist, weak tropical multicellular
convection to well-organized supercells on the HP / classic / LP
spectrum (Rasmussen and Straka, 1998).
studied by Bruning et al. (2007) and Mansell et al. (2010), as
well as winter storms studied by Rust and Trapp (2002),
Kuhlman and Manross (2011) and Schultz et al. (2011), exist
on the fast-depletion side of the diagram. The lower charging
pair may not be present in some winter storms due to the
presence of precipitation growth only in the coldest parts of
-30 -40
new supercooled water which is rapidly scavenged by large
number concentrations of ice hydrometeors. In warm-season
convection, a relatively large percentage of the cloud water is
consumed by warm rain processes prior to reaching the
mixed phase region (for multicellular convection).
The updraft core of the –CG dominant supercell summa-
rized in this study sits in the middle of the diagram, where
wet growth is assumed to suppress the relative importance
of the lower positive charge region. The diagram also notes a
relatively large warm cloud depth. As warm cloud depth
decreases, the near-complete dominance of the upper charging
pair reverses, giving storms with the smallest warm cloud
depth and dominant positive charge to graupel.
The far right of the diagram is characterized by the 4
June 2000 storm studied by Rust and MacGorman (2002) and
Rust et al. (2005), where a small low-precipitation supercell
apparently only had charge regions associated with the lower
charging pair in a normal tripole. No CGs were reported
with this storm.
Supercells on 28–29 June 2000 (MacGorman et al., 2005;
Wiens et al., 2005; Tessendorf et al., 2005; Kuhlman et al.,
2006), 26 May 2004 (Bruning et al., 2010) and 29 May 2004
(Kuhlman, 2010) are characterized by two main-updraft charge
structures. A supercooled-rich environment favored the
lower charging pair in the updraft core, while both charging
pairs were active on the updraft periphery where entrain-
ment and precipitation detrainment were taking place.
The multicellular hailstorm studied by Emersic et al.
(2011) occupies three different positions, with an early stage
dominated by relatively weak convection that was quite
similar to the convection in Bruning et al. (2007). Later, the
storm broke a midlevel capping inversion, resulting in a more-
vigorous updraft that shifted electrification into a balanced
tripole with –CGs. Finally, a particularly strong updraft per-
turbation associated with full realization of the available
potential energy and liquid water content in the boundary
layer resulted in a liquid-water-rich environment. Radar data
showed evidence of wet growth and lightning mapping data
indicated that the lower charging pair had been elevated and
favored. Positive CGs were associated with this last phase in
the storm's growth. A similar pattern was also seen in the
strong multicellular storm on 25 June 2000 studied by Weiss et
al. (2008), who found that the lower charging pair was elevated
and favored for parcels that experienced enhanced forcing by
an outflow boundary and the best boundary layer trajectory,
when compared to the initial convection associated with the
storm complex.
In the strong or severe multicellular storms, the strength of
individual cellular pulses, the extent to which they fully realize
8 E.C. Bruning et al. / Atmospheric Research xxx (2012) xxx–xxx

p
cloud top

(a) Fractional
ctional positive + and negative – charging along updraft trajectories

cloud base
Fastest-depleted
eted effective water content, Slowest-depleted effective
effe water content,
supercooled-liquid-poor
ed-liquid-poor mixed phase supercooled-liquid-rich
supercooled-liquid mixed-phase

Glaciated cloud
Charge
carrier

ice crystals

graupel Mixed-phase
region
ice crystals

graupel
wet growth suppresses lower charging pair

Warm cloud
Relative warm

cloud depth

2000/6/4
LP storm with few CGs
(Rust and MacGorman 2002,
Rust et al. 2005)
Winter Storms
(Rust and Trapp 2002, entraining updraft 2004/5/29
updraft core
Kuhlman and Manross 2010, mostly –CG, HP supercell
Schultz et al., 2011 ) (Kuhlman 2010)

entraining updraft
2000/6/29, 2004/5/26
2004/6/29 2008/2/6 updraft core +CG dominant
–CG dominant HP supercell updraft core Classic/LP supercell
Early (right) & late (left)
(Wiens et al. 2005, Tessendorf et al.
(Bruning et al. 2009)
stage tropical multicell 2005, MacGorman et al. 2005,
Kuhlman et al 2006, Bruning et al. 2010)
(Bruning et al. 2007, Mansell et al. 2011)
midlevel cap full CAPE realized,
early broken, -CG +CG. Did cloud base rise? 2006/8/15
Multicellular
hailstorm
(Emersic et al. 2011)

2000/6/25
initial convection outflow-modified thermodynamics
Multicellular storm
(Weiss et al. 2008)

(b) (c)
+ – + –
+ + – +

– – + –
–
– Advection,
+ –
+
+ Sedimentation,
+ +
+ Lightning,
Other electrification
– –
mechanisms +
Various charge sequences produced by Inverted Normal,
the NI-RGR mechanism tripole (outside updraft)

Fig. 4. (a) Observed charge structures in updraft core regions for various storms, which arise from a continuum of positive and negative charging along updraft
trajectories. The white dashed line within the negative charge region separates the relative fraction of charging to negative graupel vs. ice. (b) Local charge
structures that arise easily from the continuum of non-inductive relative-growth-rate electrification, and which might be found in 1D, 2D, or 3D characteristic
storm flows. Such charge structures might exhibit vertical shifts and deactivate charging zones, as shown in (a). The local charge sequences in (b) might combine
through advection or sedimentation to produce (c) other observed local sequences of charge, which might also be influenced by screening layer formation,
deposition of charge by lightning, or other electrification mechanisms. The charge structures in (c) likely require a minimum of 2D or 3D characteristic storm flow.

the convective available potential energy, and any variations in multicellular parcels and supercellular updraft edges exhibiting
ingested environmental thermodynamics for each cell show more effects of entrainment.
diversity in the resulting charge structure that mimics the It is particularly interesting to compare the two HP storms,
quasi-steady charge structures that are produced in supercells. which have very different inferred updraft core charge
In the supercellular case, the variation is mostly spatial relative structures. In the Alabama storm, the lower positive charge
to the updraft core, while for muticells the variation is along an center was relatively weak or absent, perhaps due to the
unsteady temporal coordinate. In both cases, entrainment may absence of precipitation sized ice and the presence of ice wet
play a role in controlling the supercooled water availability (as growth and warm rain conditions that retarded rebounding
well as the electrification reversal line itself), with smaller collisions and prevented charge transfer. These same processes

Please cite this article as: Bruning, E.C., et al., Continuous variability in thunderstorm primary electrification and an evaluation of
inverted-polarity terminology, Atmos. Res. (2012), http://dx.doi.org/10.1016/j.atmosres.2012.10.009
 E.C. Bruning et al. / Atmospheric Research xxx (2012) xxx–xxx
would have depleted much liquid water along the updraft
trajectory, and resulted in moderately low liquid water contents
in the colder portion of the mixed phase region. In the
Oklahoma storm, electrification was also suppressed at low
altitudes where wet growth took place, but the supply of
liquid (controlled by some combination of environmental
moisture supply, existing hydrometeors, CAPE, and dynamic
pressure perturbation forcing) was apparently still quite high
in the coldest portions of the mixed phase region, and favored
a very deep region of positive charging. Future studies should
give attention to observing and modeling these liquid deple-
tion processes. While instantaneous microphysical state is
an important check on hydrometeor type, the rate of consump-
tion of liquid water available for later riming seems to be
key, and as yet is poorly constrained by observational studies.
This is doubtless due in part to the difficulty in measuring
supercooled liquid along thunderstorm updraft trajectories.
3.2. Charge structures
Having discussed recent observed results for electrifica-
tion in updraft cores and the updraft-core charge structures,
it should be evident that the local charge structure produced
by active electrification is a direct response to the environ-
mental thermodynamics (Williams et al., 2005) and storm
cycling. How the non-inductive relative-growth-rate mecha-
nism within various environments might produce the re-
sultant charge structures is discussed below. The aim is to
circumscribe the possibilities (and attendant representative
terminology) that can be explained through use of the
NI-RGR mechanism alone.
Fig. 4b shows various charge sequences that can result
directly from the non-inductive relative growth rate mech-
anism. The configuration of the environmental thermody-
namics and relative shifts in vertical displacement of the
maximum updraft result in a continuum of vertical shifts in
the position of these charge sequences. Sometimes one of
the charges may be eliminated by vertical displacements that
move that polarity of graupel electrification outside the mixed
phase parameter space. These basic charge sequences may be
found in storms characterized by precipitation trajectories
that are one-, two- or three-dimensional.
The individual, local sequences of charge that arise
directly from electrification in Fig. 4b likely must be com-
bined in some way to yield inverted or more complicated
charge structures, as in Fig. 4c. The combination of charge
layers occurs by advection and sedimentation along hydro-
meteor trajectories, with the influence of new in-situ
charging playing a role. (Neglected here for simplicity are
the role of screening layer formation, inductive charging,
melting layer charging, and lightning deposition which all
likely play a minor role compared to initial NI charging in
the updraft core.) In all but the simplest storms (those with
1-D characteristic flow and strong radial symmetry about
the updraft core), many electrification regimes likely exist,
due to variations in updraft thermodynamics, supercooled
water supply, and the balance of those properties with
existing precipitation. Therefore, inverted structures, or the
more-complicated six-layer normal-polarity outside-updraft
structures of Stolzenburg et al. (1998) require two-dimensional
or three-dimensional storm flows. In one-dimensional warm-
Please cite this article as: Bruning, E.C., et al., Continuous variability in thunderstorm primary electrification and an evaluation of
inverted-polarity terminology, Atmos. Res. (2012), http://dx.doi.org/10.1016/j.atmosres.2012.10.009
9
season cases, there is no means for the non-inductive mech-
anism to produce the low-level negative charge required for
an inverted tripole. The single reversal line in the non-inductive
mechanism also precludes formation of six charge layers
without some spatial variation in the thermodynamics for
the single electrification mechanism under consideration.
4. Proposed use of “inverted”
4.1. Electrification and Charge Structure
Based on charge structures inferred in regions where the
base environmental state is most efficiently processed through
the storm (“updraft core”), it is clear that environmental
thermodynamic and aerosol controls on the non-inductive
charging mechanism vary continuously, and as such the
“inversion point” is unidentifiable.
In updraft cores, charge structures consistent with a pre-
dominant influence of the relative-growth-rate NI mechanism
have been observed. Those charge structures may be effectively
considered part of a normal tripole that is shifted continuously
relative to constant temperature levels. This continuum may
exist within the same storm, as has been inferred in STEPS,
TELEX, and other recent storms with “inverted polarity”
characteristics. These studies have noted that environmental
thermodynamics, updraft strength (and perhaps dilution by
entrainment), and aerosol/cloud condensation nucleus effects
may conspire to shift the relative importance of charge gen-
erator regions. These may be superimposed for particularly
broad updrafts and be rearranged substantially by multi-
dimensional storm flows, producing charge structures that
look like inverted-polarity tripoles. Instead of saying that a
storm becomes inverted with time, it is better to say that
the storm precipitation structure, dynamics, and kinematics
evolve to produce inverted regions in parts of storms.
Therefore, the temptation to conceptualize entire storms
as an inverted tripole (or, worse, dipole) should be resisted.
Stolzenburg et al. (1998) emphasized the problem of oversim-
plification of charge structures normal-polarity storms, and it
likewise is important for so-called inverted storms, as in Weiss
et al. (2008). Rust and MacGorman (2002) and Rust et al.
(2005) proposed that positive ground strikes resulted from
inverted-polarity electrical structures, and there is growing
consensus (Mansell et al., 2002; Kuhlman et al., 2006; Wiens et
al., 2005; Tessendorf et al., 2007; Weiss et al., 2008; Tessendorf,
2009) that a lower charge region opposite the ground strike
polarity (negative in this case) is necessary. The observations
show that a single characteristic inverted structure is an in-
appropriate simplification except in local regions. Interleaving
of charge from multiple sources (an essentially 2D or 3D
process tied to hydrometeor trajectories, as in Bruning et al.,
2010) must be necessary to explain the formation of lower
negative charge, especially if non-inductive charging is the
most significant source of net charge regions.
4.2. Flashes
The polarity of cloud and ground flashes is primarily
determined by the electrification and charge structure mech-
anisms previously discussed, with a secondary effect due to
local influences like terrain (including man-made structures
10 E.C. Bruning et al. / Atmospheric Research xxx (2012) xxx–xxx
such as radio and television towers) and charge deposited by
flashes themselves (e.g., Coleman et al., 2003). As this review
of results from recent studies focused on the inverted polarity
question shows, the normal / inverted cloud flash terminol-
ogy references a false dichotomy in storm charge structures
that vary continuously in their height of production and in
their interleaving within individual storms (Fig. 5a,b).
It is still useful to refer in shorthand to the orientation
of the vertical dipole that is discharged by a typical flash that
propagates into two horizontally stratified charge regions.
(This leaves aside the issue of what to call flashes that discharge
horizontally adjacent regions, which have also been observed.)
The convention used for discharges where a channel connects
to ground is recommended to resolve this problem. A negative
ground flash lowers negative charge to ground, while a positive
ground flash lowers positive charge to ground. What has been
referred to as an “inverted-polarity cloud flash” becomes a
“negative cloud flash” and a “normal-polarity cloud flash”
becomes a “positive cloud flash,” or more compactly, –IC and
+IC as counterparts to –CG and +CG.
There is some precedent for use of the proposed IC and CG
polarity terminology in extant studies, for example Lu et al.
(2012). Furthermore, while low-frequency (LF) sferics loca-
tion systems such as the National Lightning Detection Network
primarily detect ground strikes, some have recently started
reporting cloud flashes as well (Cummins and Murphy, 2009),
and in doing so have adopted a similar convention.
The proposed cloud flash terminology naturally results in
the correct description of the physics involved (Fig. 5c,d).
Consider the lower dipole of a horizontally stratified tripolar
charge structure that directly results from the non-inductive
relative-growth-rate mechanism. Both cloud flashes and ground
flashes may be observed to propagate through these two layers.
Under the old naming convention, these low-altitude flashes
would have been termed “inverted” if they remained in the
cloud, while becoming a negative strike (the “normal” polarity)
if ground contact was made. Such confusion is eliminated under
(a) (b)
– Walt Petersen, Chris Schultz, Elise Schultz, and Rich Blakeslee
Normal
+ Inverted
+ vided helpful comments on an early version of thispaper. We
Inverted,
sometimes comes
– Normal,
sometimes comes
to ground as –CG + – to ground as +CG
(c) (d)
– GOES-R GLM Science Program helped bring this work to
+IC
+ –IC
+
sometimes comes
–IC – +IC
sometimes comes
to ground as –CG + – to ground as +CG
Fig. 5. “Inverted” and “normal” polarity cloud flash terminology as applied to
(a) “normal” and (b) “inverted” polarity tripolar charge structures. The
“normal” flash in (a) is usually thought of as the prototypical cloud flash.
This study supports use of the terms +IC and –IC, which have pleasing
symmetry with existing ground flash terminology, as in panels (c) and (d).
Please cite this article as: Bruning, E.C., et al., Continuous variability in thunderstorm primary electrification and an evaluation of
inverted-polarity terminology, Atmos. Res. (2012), http://dx.doi.org/10.1016/j.atmosres.2012.10.009
the new naming convention, where the same in-cloud charge
configuration that produces a negative ground flash if the
channel reaches the ground will produce a negative in-cloud
flash if it does not.
4.3. Net Storm Polarity
What, then, of “inverted-polarity storms?” The production
of predominantly positive strikes by certain storms (e.g., Carey
and Rutledge, 2003) must be accounted for, and indeed the
terminology is helpful shorthand for that observation (Williams
et al., 2005; Carey and Buffalo, 2007), though those ratios
are themselves a continuum. From an energetic perspective,
Krehbiel et al. (2008) have argued that such storms have a net
excess of positive charge, much like negative ground-strike pro-
ducing storms have an excess of negative charge (Mansell et al.,
2010), where screening layer formation at storm top and pre-
cipitation fluxes can bias potentials to allow lightning channels
to escape the cloud. The polarity of the net contribution of
whole storms to the global circuit (Mach et al., 2011) might be
another appropriate storm-total metric.
5. Concluding Remarks
This study proposes a flexible view of thunderstorm elec-
trification and charge structure. A continuum of storm types
has been observed that suggests a wide range of traversals of
the cloud water vs. temperature electrification parameter space.
The liquid depletion rate along updraft trajectories is a helpful
way to think about this traversal. Different flash polarities and
their vertical positions within the thunderstorm are the natural
result of shifts and combinations of charge structures produced
along this continuum. Such ideas may be useful in future
analyses of datasets pertaining to the storm polarity question,
including those collected in the Deep Convective Clouds and
Chemistry (DC3) program during May and June 2012.
Acknowledgments
The authors wish to thank Don MacGorman, Larry Carey,
for providing data for this study. Christopher Emersic pro-
thank the three anonymous reviewers for their insightful
reviews that led to further refinement of this study. Some of
the ideas in this paper are based upon work conducted while
the authors were supported by NSF Grant ATM-0233268 and
a National Science Foundation Graduate Research Fellowship.
Recent support from UCAR/COMET award Z11-91820 and the
completion.
References
Ackerman, S.A., Knox, J.A., 2007. Meteorology: Understanding the Atmo-
sphere, second edition. Thomson Higher Education, Belmont, CA.
Aguado, E., Burt, J.E., 2010. Understanding Weather and Climate, fifth
edition. Prentice Hall, New York.
Albrecht, R.I., Morales, C.A., Silva Dias, M.A.F., 2011. Electrification of
precipitating systems over the Amazon: Physical processes of thunder-
storm development. J. Geophys. Res. 116.
Baker, B., Baker, M.B., Jayaratne, E.R., Latham, J., Saunders, C.P.R., 1987. The
influence of diffusional growth rates on the charge transfer accompanying
 E.C. Bruning et al. / Atmospheric Research xxx (2012) xxx–xxx
rebounding collisions between ice crystals and soft hailstones. Q. J. Roy.
Met. Soc. 113, 1193–1215.
Bruning, E.C., Rust, W.D., Schuur, T.J., MacGorman, D.R., Krehbiel, P.R., Rison,
W., 2007. Electrical and polarimetric radar observations of a multicell
storm in TELEX. Mon. Weather Rev. 135, 2525–2544. http://dx.doi.org/
10.1175/MWR3421.1.
Bruning, E.C., Rust, W.D., MacGorman, D.R., Biggerstaff, M.I., Schuur, T.J.,
2010. Formation of charge structures in a supercell. Mon. Weather Rev.
138, 3740–3761.
Byers, H.R., Braham, R.R., 1949. The Thunderstorm: Report of the Thunder-
storm Project. U.S. Govt. Print. Off. (287 pp.).
Carey, L.D., Buffalo, K.M., 2007. Environmental control of cloud-to-ground
lightning polarity in severe storms. Mon. Weather Rev. 135, 1327–1353.
Carey, L.D., Rutledge, S.A., 2003. Characteristics of cloud-to-ground lightning
in severe and nonsevere storms over the central United States from
1989–1998. J. Geophys. Res. 108, 4483.
Carey, L.D., Murphy, M.J., McCormick, T.L., Demetriades, N.W., 2005.
Lightning location relative to storm structure in a leading-line trailing
stratiform mesoscale convective system. J. Geophys. Res. 110.
Coleman, L.M., Marshall, T.C., Stolzenburg, M., Hamlin, T., Krehbiel, P.R.,
Rison, W., Thomas, R.J., 2003. Effects of charge and electrostatic potential
on lightning propagation. J. Geophys. Res. 108.
Cummins, K.L., Murphy, M.J., 2009. An overview of lightning locating
systems: History, techniques, and data uses, with an in-depth look at
the U.S. NLDN. IEEE Trans. Electromagn. Compat. 51, 499–518.
Darden, C.B., Nadler, D.J., Carcione, B.C., Blakeslee, R.J., Stano, G.T., Buechler,
D.E., 2010. Using total lightning information to diagnose convective
trends. Bull. Am. Meteorol. Soc. 91, 167–175.
Emersic, C., Saunders, C.P.R., 2010. Further laboratory investigations into the
Relative Diffusional Growth Rate theory of thunderstorm electrification.
Atmos. Res. 98, 327–340.
Emersic, C., Heinselman, P.L., MacGorman, D.R., Bruning, E.C., 2011.
Lightning activity in a hail-producing storm observed with phased-
array radar. Mon. Weather Rev. 139, 1809–1825.
Krehbiel, P.R., Riousset, J.A., Pasko, V.P., Thomas, R.J., Rison, W., Stanley, M.A.,
Edens, H.E., 2008. Upward electrical discharges from thunderstorms.
Nat. Geosci. 1, 233–237.
Kuhlman, K.M., 2010. Evolution of charge and lightning throughout an observed
and simulated supercell storm. 156 leaves. University of Oklahoma.
Kuhlman, K.M., Manross, K., 2011. Lightning and polarimetric signatures of
two electrified winter storms in central Oklahoma. Fifth Conference on
Meteorological Applications of Lightning Data. American Meteorological
Society, Seattle, WA, USA, p. 7.4.
Kuhlman, K.M., Ziegler, C.L., Mansell, E.R., MacGorman, D.R., Straka, J.M.,
2006. Numerically simulated electrification and lightning of the 29 June
2000 STEPS supercell storm. Mon. Weather Rev. 134, 2734–2757.
Kuhlman, K.M., MacGorman, D.R., Biggerstaff, M.I., Krehbiel, P.R., 2009.
Lightning initiation in the anvils of two supercell storms. Geophys. Res.
Lett. 36.
Lang, T.J., Miller, L.J., Weisman, M.L., Rutledge, S.A., Barker III, L.J., Bringi, V.N.,
Chandrasekar, V., Detwiler, A., Doesken, N., Helsdon, J.H., Knight, C.,
Krehbiel, P.R., Lyons, W.A., MacGorman, D.R., Rasmussen, E.N., Rison, W.,
Rust, W.D., Thomas, R.J., 2004. The Severe Thunderstorm Electrification
and Precipitation Study. Bull. Am. Meteorol. Soc. 85, 1107–1125.
Lu, G., Cummer, S.A., Blakeslee, R.J., Weiss, S.A., Beasley, W.H., 2012. Lightning
morphology and impulse charge moment change of high peak current
negative strokes. J. Geophys. Res. 117, D04213.
MacGorman, D.R., Rust, W.D., Krehbiel, P.R., Rison, W., Bruning, E.C., Wiens,
K., 2005. The electrical structure of two supercell storms during STEPS.
Mon. Weather Rev. 133, 2583–2607.
MacGorman, D.R., Rust, W.D., Schuur, T.J., Biggerstaff, M.I., Straka, J.M.,
Ziegler, C.L., Mansell, E.R., Bruning, E.C., Kuhlman, K.M., Lund, N.R.,
Biermann, N.S., Payne, C., Carey, L.D., Krehbiel, P.R., Rison, W., Eack, K.B.,
Beasley, W.H., 2008. TELEX: The Thunderstorm Electrification and
Lightning Experiment. Bull. Am. Meteorol. Soc. 89, 997–1013.
Mach, D.M., Blakeslee, R.J., Bateman, M.G., 2011. Global electric circuit
implications of combined aircraft storm electric current measurements
and satellite-based diurnal lightning statistics. J. Geophys. Res. 116.
Mansell, E.R., MacGorman, D.R., Ziegler, C.L., Straka, J.M., 2002. Simulated
three-dimensional branched lightning in a numerical thunderstorm
model. J. Geophys. Res. 107.
Please cite this article as: Bruning, E.C., et al., Continuous variability in thunderstorm primary electrification and an evaluation of
inverted-polarity terminology, Atmos. Res. (2012), http://dx.doi.org/10.1016/j.atmosres.2012.10.009
11
Mansell, E.R., Ziegler, C.L., Bruning, E.C., 2010. Simulated electrification of a
small thunderstorm with two-moment bulk microphysics. J. Atmos. Sci.
67, 171–194.
Marshall, T.C., Rust, W.D., Stolzenburg, M., 1995. Electrical structure and updraft
speeds in thunderstorms over the southern great plains. J. Geophys. Res.
100, 1001–1015.
Mitzeva, R., Saunders, C., Tsenova, B., 2006. Parameterisation of non-inductive
charging in thunderstorm regions free of cloud droplets. Atmos. Res. 82,
102–111.
Rasmussen, E.N., Straka, J.M., 1998. Variations in supercell morphology. part i:
Observations of the role of upper-level storm-relative flow. Mon. Weather
Rev. 126, 2406–2421.
Rust, W.D., MacGorman, D.R., 2002. Possibly inverted-polarity electrical
structures in thunderstorms during STEPS. Geophys. Res. Lett. 29. http://
dx.doi.org/10.1029/2001GL014303.
Rust, W.D., Marshall, T.C., 1996. On abandoning the thunderstorm tripole-
charge paradigm. J. Geophys. Res. 101, 23499–23504.
Rust, W.D., Trapp, R.J., 2002. Initial balloon soundings of the electric field in
winter nimbostratus clouds in the USA. Geophys. Res. Lett. 29.
Rust, W.D., MacGorman, D.R., Bruning, E.C., Weiss, S.A., Krehbiel, P.R.,
Thomas, R.J., Rison, W., Hamlin, T., Harlin, J., 2005. Inverted-polarity
electrical structures in thunderstorms in the Severe Thunderstorm
Electrification and Precipitation Study. Atmos. Res. 76, 247–271.
Saunders, C.P.R., Peck, S.L., 1998. Laboratory studies of the influence of the
rime accretion rate on charge transfer during crystal/graupel collisions.
J. Geophys. Res. 103, 13,949–13,956.
Saunders, C.P.R., Bax-Norman, H., Emersic, C., Avila, E.E., Castellano, N.E.,
2006. Laboratory studies of the effect of cloud conditions on graupel/
crystal charge transfer in thunderstorm electrification. Q. J. Roy. Met.
Soc. 132, 2653–2673.
Schultz, C.J., Bruning, E.C., Carey, L.D., Petersen, W.A., Heckman, S., 2011.
Total lightning within electrified snowfall using LMA, NLDN and WTLN
measurements. Eos Trans. AGU, Fall Meet. Suppl.
Simpson, G., Robinson, G.D., 1941. The distribution of electricity in thunder-
clouds, II. Proc. Roy. Soc. London A 177, 281–329.
Simpson, G., Scrase, F.J., 1937. The distribution of electricity in thunder-
clouds. Proc. Roy. Soc. London A 161, 309–352.
Stolzenburg, M., Rust, W.D., Marshall, T.C., 1998. Electrical structure in
thunderstorm convective regions 3. Synthesis. J. Geophys. Res. 103,
14097–14108.
Tessendorf, S.A., 2009. Lightning: Principles, Instruments and Applications.
(chapter 4) Characteristics of Lightning in Supercells. Springer, pp. 83–114.
Tessendorf, S.A., Miller, L.J., Wiens, K.C., Rutledge, S.A., 2005. The 29 June
2000 supercell observed during STEPS. Part I: Kinematics and micro-
physics. J. Atmos. Sci. 62, 4127–4150.
Tessendorf, S.A., Rutledge, S.A., Wiens, K.C., 2007. Radar and lightning
observations of normal and inverted polarity multicellular storms from
STEPS. Mon. Weather Rev. 135.
Weiss, S.A., Rust, W.D., MacGorman, D.R., Bruning, E.C., Krehbiel, P.R., 2008.
Evolving complex electrical structure of the STEPS 25 June 2000
multicell storm. Mon. Weather Rev. 136, 741–756.
Weiss, S.A., MacGorman, D.R., Calhoun, K.M., 2012. Lightning in the anvils of
supercell thunderstorms. Mon. Weather Rev. 140, 2064–2079.
Wiens, K.C., Rutledge, S.A., Tessendorf, S.A., 2005. The 29 June 2000 supercell
observed during STEPS. Part II: Lightning and charge structure. J. Atmos.
Sci. 62, 4151–4177.
Williams, E.R., 2001. Severe Convective Storms. (chapter 13) The Electrification
of Severe Storms. American Meteorological Society, Boston, pp. 527–561.
Williams, E.R., Weber, M.E., Orville, R.E., 1989. The relationship between
lightning type and convective state of thunderclouds. J. Geophys. Res. 94,
13213–13220.
Williams, E., Mushtak, V., Rosenfeld, D., Goodman, S., Boccippio, D., 2005.
Thermodynamic conditions favorable to superlative thunderstorm up-
draft, mixed phase microphysics and lightning flash rate. Atmos. Res. 76,
288–306.
Wilson, C.T.R., 1916. On some determinations of the sign and magnitude of
electric discharges in lightning flashes. Proc. Roy. Soc. London A 92, 555–574.
Wilson, C.T.R., 1925. The electric field of a thundercloud and some of its
effects. Proc. Phys. Soc. Lond. 37, 32D–37D.