1. Introduction
Interest in wintertime lightning and thunder dates to the Han Dynasty of ancient China (206 BC–220 AD), when emperors regarded winter thunderstorms as abnormal and evil, portending invasions or natural disasters (Wang 1980; Wang and Chu 1982). In western literature, records of winter thunderstorms first appear in the late nineteenth and early twentieth centuries (Herschel 1888; Butterworth 1895; Beldon 1927). The first scientific attempt to produce a climatology of wintertime thunderstorms for the United States was that of Curran and Pearson (1971). Since their work, a number of cold-season thunderstorm climatologies for the United States have been developed (Colman 1990; Holle et al. 1997; Holle and Cortinas 1998; Moore and Idone 1999; Hunter et al. 2001; Market et al. 2002, 2006; Crowe et al. 2006; Market and Becker 2009). A significant body of work has also been compiled for winter thunderstorms occurring along the western coast of Japan [see review by Rakov and Uman (2003), 308–316]. Together, these studies have provided an understanding of common meteorological weather patterns that support the development of wintertime thunderstorms.
The first weather pattern is associated with cold air moving over warm water, such as over and downwind of Lakes Erie and Ontario (Moore and Orville 1990; Schultz 1999; Lackmann 2001; Steiger et al. 2009), in the vicinity of the Great Salt Lake of Utah (Schultz 1999; Schultz et al. 2002), over the Gulf Stream off the U.S. East Coast (Orville 1990, 1993), and over and downwind of the Sea of Japan during Siberian cold-air outbreaks (Kitagawa and Michimoto 1994). The thunderstorms that develop in this weather pattern are surface based, triggered by strong surface heat and moisture fluxes between the warm water and the colder air overriding it. Lightning occurs more frequently in early winter when the water is warmer, although the degree of instability required to trigger convection capable of producing lightning remains somewhat uncertain because the maximum instability is realized over water, whereas thermodynamic measurements from soundings are typically made on a downwind shoreline (Schultz 1999). Steiger et al. (2009) showed that modifying model output at low levels to account for heat and moisture flux from Lakes Erie and Ontario yielded “lake induced” CAPE values in excess of 1790 J kg−1 for two lightning-producing lake-effect storms. The likelihood that lightning flashes will occur in storms over U.S. lakes and the Sea of Japan appears to be related to the height of the −10°C isotherm above the surface (Kitagawa 1992; Michimoto 1993; Steiger et al. 2009). Dual polarization measurements from Japanese storms suggest that graupel and ice occur simultaneously near the −10°C level (Maekawa et al. 1992), contributing to charge separation [see MacGorman and Rust (1998, 61–75) for a review of cloud electrification mechanisms]. An interesting anomaly of these storms is their propensity for positive lightning flashes (Takeuti et al. 1973, 1976, 1978; Brook et al. 1982; Moore and Orville 1990). Market et al. (2002) found that about 4% of all thundersnow events in the United States are associated with lake-effect storms.
Wintertime lightning also occurs associated along frontal boundaries in extratropical cyclones. Most often, lightning occurs in association with the cold frontal convection (e.g., Pessi and Businger 2009). Much less commonly, wintertime lightning involves elevated thunderstorms, forming either over a stable layer within the comma head of winter cyclones (e.g., Martin 1998; Halcomb and Market 2003; Moore et al. 2005) or over an advancing arctic front (e.g., Trapp et al. 2001). Although lightning accompanied by snowfall has been reported with other winter mesoscale phenomena such as mesoscale gravity waves (Bosart and Sanders 1986; Schneider 1990), isolated precipitation bands (Schumacher et al. 2010), and frontal squall lines (Pettegrew et al. 2009), most reports involve thunderstorms forming above a stable layer capped by a frontal surface. About 53% of all thundersnow reports in the United States were associated with continental cyclones, while about 6% were associated with arctic fronts (Market et al. 2002). Climatological studies by Curran and Pearson (1971), Colman (1990), Holle and Cortinas (1998), and Market et al. (2002) together show that cold-season (November–March) elevated thunderstorms are most common across the central plain states of the United States from Oklahoma northward into Nebraska, and eastward into Iowa and Illinois. Their analyses suggest that wintertime lightning is much less common in the northeastern United States.
A focus of investigations of elevated winter thunderstorms has been to determine the instability mechanism triggering their formation. Colman (1990) determined that elevated thunderstorms typically occurred in environments characterized by near-zero CAPE. His conclusion was based on analyses of the 850–500-hPa lapse rate (determined from 0000 and 1200 UTC weather charts) at locations where elevated thunderstorms were reported. His results and other studies (e.g., Moore and Blakly 1988; Holle and Watson 1996) raised the possibility that elevated thunderstorms might be triggered by the release of conditional symmetric instability [CSI; see review by Schultz and Schumacher (1999)]. The findings of these earlier studies were reinforced by Market et al. (2006), who examined close-proximity balloon soundings for snow events with lightning and thunder during the period 1961–1990. Although some of their soundings exhibited small values of CAPE, the majority did not, again leading the authors to conclude that weak CSI in the presence of strong frontogenetic forcing (Market et al. 2007) may be operating, at least in some of the events. For cases where instability was present, Market et al. (2006) found that the least stable (or most unstable) parcel originated about 30–50 hPa above the top of the frontal inversion and that dry air was typically present about 100 hPa above the level of the most unstable parcel. Their composite sounding showed that the temperature at the level of the most unstable parcel was −8.7°C. Charging is most common at somewhat colder temperatures, −20° ≤ T ≤ −10°C, in a variety of thunderstorm types (MacGorman and Rust 1998). Case studies of other thundersnow events (e.g., Moore et al. 1998; Martin 1998; Hunter et al. 2001; Halcomb and Market 2003) also showed CAPE above the frontal inversion.
A complication with invoking CSI to explain thunderstorm formation comes from consideration of cloud microphysics. Theoretical calculations of the maximum vertical motion expected in bands formed under inviscid conditions during the release of CSI predict maximum vertical velocities on the order of 1 m s−1 (Emanuel 1983; see also Bluestein 1993, 556–559). Modeling studies of precipitation bands, which include mixing, predict maximum vertical motions during the release of CSI on the order of 0.1 m s−1 (e.g., Bennetts and Hoskins 1979; Knight and Hobbs 1988, their Figs. 8 and 11; Xu 1992; Persson and Warner 1995, their Fig. 13; Zhang and Cho 1995, their Fig. 3). Noninductive charging in thunderstorms arises primarily from collisions between graupel and ice particles in the presence of supercooled water (MacGorman and Rust 1998, 65–70). Graupel formation requires updrafts that can both support the production of supercooled water and suspend the graupel particles during their growth. Production of supercooled water may be inhibited in snowstorms with weak updrafts because significant ice particle concentrations limit the saturation level due to the growth of ice by the Wegener–Bergeron–Findeisen precipitation mechanism (Pruppacher and Klett 1997, p. 119). Market et al. (2006) and Crowe et al. (2006) both note that thundersnow is often accompanied by moderate to heavy snowfall and significant 24-h accumulations. We note that ice–ice collisional charging in the absence of supercooled water has been shown to occur in laboratory charging studies. However, these alternate charging processes are much less efficient at separating charge than in noninductive laboratory studies in which ice crystals collide with a rimer in the presence of supercooled water [see Pruppacher and Klett (1997, 812–827) for a review of charging mechanisms].
As summarized above, a growing body of research has accumulated concerning the structure of winter cyclones that produce lightning flashes. To date, however, there have been no studies of the finescale structure of convection within the comma head of winter cyclones, and how cyclone mesostructure relates to the occurrence and distribution of lightning across the comma head. This paper focuses on these two issues. We report detailed measurements made during two winter seasons (2008/09 and 2009/10) during the Profiling of Winter Storms (PLOWS) field campaign (see also Rosenow et al. 2014; Market et al. 2012).
2. Data sources and analysis methods
PLOWS instrumentation platforms included the National Center for Atmospheric Research (NCAR) C-130 aircraft (2009–10 only), the University of Alabama–Huntsville (UAH) Mobile Integrated Profiling System (MIPS) and Mobile Alabama X-Band (MAX) scanning dual-polarization Doppler radar, the NCAR Mobile Integrated Sounding System (MISS), and the University of Missouri (UM) Sounding System. The PLOWS data were supplemented by measurements from the National Lightning Detection Network (NLDN; Orville et al. 2011) for the two winter seasons, plus two additional winter seasons (2007/08 and 2010/11). The C-130 was based near Peoria, Illinois, from 1 November to 15 December 2009, and again from 15 January to 7 March 2010. The aircraft flew stacked patterns nominally normal to the comma head precipitation bands of winter cyclones as the cyclones crossed the midcontinent (Fig. 1). The C-130 was equipped with the University of Wyoming W-Band Cloud Radar (WCR; Wang et al. 2012), the upward-looking Wyoming Cloud lidar, and a suite of microphysics probes. The WCR measurements of equivalent radar reflectivity factor Ze (hereafter simply reflectivity) and vertical radial velocity were placed in the context of storm structure by overlaying numerical model initialization gridded data from the Rapid Update Cycle (RUC) model. Details of the WCR processing procedures and RUC overlays are given in Rosenow et al. (2014).
PLOWS deployment strategies. In both strategies, the MAX radar was deployed approximately 30 km from the WSR-88D to allow for dual-Doppler scanning. The aircraft flew between two navigation beacons (vortacs), with the MISS 915-MHz profiler and MIPS 915-MHz profiler and XPR deployed along the line between the vortacs. (top) In strategy 1, the flight passes over the MAX, allowing for RHI scans directly along the flight path. (bottom) In strategy 2, the track was centered on the dual-Doppler lobe. Ideal sampling occurred when the WSR-88D was within the comma head and the profilers were deployed on opposite sides of the comma head.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
The MISS and MIPS each deployed a 915-MHz profiler. For part of the project, the MIPS also deployed a vertically pointing X-Band Profiling Doppler Radar (XPR) with a 1.3° beam and a sampling rate of 6 Hz. Rawinsondes were launched at 3-h intervals at both the MIPS and MISS locations. The deployment strategies of all these platforms are summarized in Fig. 1. During PLOWS, the MIPS profiler was operated in a three-beam mode with a dwell time of 20 s per beam, a gate spacing of 98 m, and 100 range gates per dwell period. The vertical beam was sampled every other dwell period, producing a 40-s time interval between successive vertical beams. The vertical beam of the profiler had a width of 9°, with the full power at zenith and half-power points at the nominal “edge” of the beam (4.5°), leading to a horizontal resolution at 2-, 4-, and 6-km altitudes of 310, 620, and 940 m. In this paper, we report measurements of signal-to-noise ratio (SNR, a proxy for radar reflectivity), vertical radial velocity, and spectral width. The vertical radial velocity is the sum of the vertical air velocity and the particle ensemble terminal velocity. The MIPS was also equipped with a field mill to measure local electric field variations. The field mill, borrowed from NCAR, was uncalibrated, and therefore provided only qualitative, relative measurements of surface electric field fluctuations. Nevertheless, the field mill provided clear evidence of cloud charging and lightning flashes related to convective structures detected by the profiler. The operating principles of field mills can be found in MacGorman and Rust (1998, 118–122).
NCAR deployed the truck- and trailer-mounted MISS for PLOWS. MISS includes a similar 915-MHz wind profiler radar, a rawinsonde sounding system (using Vaisala RS92 GPS rawinsondes), and a surface weather station (Cohn et al. 2011). Rawinsondes were launched at 3-h intervals at both the MIPS and MISS locations. Ground sites were chosen prior to the arrival of the cyclone, and instrument platforms remained at those locations for the duration of the storm as it passed over the site. The rawinsonde and profiler data therefore provide a time sequence of storm evolution at a specific site.
Both cloud-to-ground (CG) and intracloud flash data are used from the NLDN. These data are discussed at length by Pettegrew (2008), which we summarize herein. According to Cummins et al. (1998), the NLDN allows for grouping of flash data and flash multiplicity. A spatial and clustering algorithm is used to identify lightning flashes. Following the first stroke, any additional strokes identified within 10 km and 500 ms are considered part of one flash, with multiplicity limited to 15 strokes. Detection efficiency for the entire NLDN is determined by the individual sensor detection efficiency, the number of sensors, and the sensor baselines. More recently, Biagi et al. (2007) reported an NLDN detection efficiency of about 71% for CG strokes and greater than 90% for CG flashes during several field campaigns in 2003 and 2004 in Arizona, Texas, and Oklahoma. During this study, the best observed detection efficiency was found in Arizona at 93% with first stroke detection efficiency of 76%. The values for Texas and Oklahoma were 92% and 86%, respectively. Biagi et al. (2007) found median errors in location of 424 m (Arizona) and 282 m (Texas and Oklahoma), which are significantly smaller than that observed (500 m) from Cummins et al. (1998), and are most likely a function of NLDN detector spacing. Ward et al. (2008) found median location reporting errors to be 600–700 m on the fringe (Florida) of the NLDN; thus, the median location reporting error of the entire NLDN remains on the order of 500 m. Not only is the present study area near the heart of the NLDN, but continued implementations to algorithms (e.g., Cummins et al. 2006) have facilitated the improvements in location reporting documented by Ward et al. (2008).
The detection efficiency of cloud flashes in the NLDN is 10%–20% (Cummins and Murphy 2009). Previously, low-amplitude (<10 kA) positive CG flashes were simply removed from datasets and deemed to be cloud flashes (Orville et al. 2002). Later, Biagi et al. (2007) sorted the flash dataset by CG type and peak current values. They determined that only 1.4%–7% of identified positive CG lightning with peak currents less than 10 kA were verified, with those remaining probably being cloud flashes. However, negative CG flashes were confirmed for peak currents of 10 kA or less 50%–87% of the time. Clearly, cloud flash detection is more complex, as there is no contact with ground. Furthermore, it is commonplace, with the NLDN, to find cloud flashes located within close proximity to CG strokes.
3. Case studies
a. 8–9 December 2009
The synoptic and mesoscale structure of the 8–9 December 2009 cyclone is presented in Rosenow et al. (2014). The cyclone developed east of the Rockies in association with a strong short wave that deepened over eastern Colorado. At the time of the analyses in this paper, a 990-hPa surface low pressure center was located over east-central Missouri with an open wave aloft with south-southwest 500-hPa flow over Iowa and Illinois [see Figs. 3–5 of Rosenow et al. (2014)]. Rosenow et al. showed that the comma head of the cyclone was divided into two distinct regions (Fig. 2). The northern side consisted of a deep stratiform cloud containing a steady weak updraft on the order of 0.2 m s−1, topped by cloud-top-generating cells with updrafts of the order of 1–3 m s−1. The southern side was marked by elevated convective cells originating slightly above a frontal inversion associated with the advance of a Pacific cold front aloft (Rosenow et al. 2014) and rising to near the tropopause. These cells penetrated into the drier, otherwise slowly descending air mass that was part of the cyclone's dry slot and overran Gulf air at lower levels within the comma head (Rosenow et al. 2014). Figure 3 shows composite radar analyses of the cyclone at four times. The C-130 flight track that corresponds to Fig. 2 is indicated in Fig. 3c. The curved red line across the flight track in Fig. 3c denotes the boundary between the convective and stratiform sides of the comma head located at about 180 km (0315 UTC) in Fig. 2. The locations of lightning flashes are also shown. All flashes shown were within half an hour of the time of each image. It is clear from these plots that lightning originated from the convection emerging into the dry-slot air overrunning moist Gulf air at lower altitudes on the southern side of the comma head.
(a) Ze from the WCR from 0250 to 0350 UTC 9 Dec 2009 and θei from the RUC initialization valid at 0300 UTC along the same cross section. The black dashed line is the tropopause. (b) Vertical radial velocities measured by the WCR, overlain with θei from the RUC initialization. The blue (black) lines in (a),(b) denote the 70% (100%) RHi contours. The horizontal black lines in (a),(b) are the C-130 flight track. (c) Vertical velocities from the RUC 1-h forecast overlain with θei for the same cross section as in (a),(b), valid at 0300 UTC.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
Composite radar reflectivity analysis at (a) 0100, (b) 0200, (c) 0400 and (d) 0600 UTC 9 Dec 2009 overlain with a sea level pressure analysis (2-hPa interval). The C-130 flight track is shown in yellow for 0251–0351 UTC in (c) (corresponding to Fig. 2) and 0400–0610 UTC in (d). Lightning-flash locations are designated by the white dots in red circles. The red squares denote the location of the MIPS 915-MHz profiler and the curved red line bisecting the flight track in (c) is the boundary between the convective and stratiform regions appearing in Fig. 2.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
Figure 4 shows the MIPS 915-MHz profiler SNR for the time period 1400 UTC 8 December–0800 UTC 9 December 2009. Superimposed on the figure are rawinsonde measured winds and equivalent potential temperature with respect to ice θei. Air between the dashed blue lines at the top and bottom of the diagram was saturated with respect to ice. The lower red line denotes the maxima in θei above the lower-level stable layer. The upper red line is the equilibrium level for ice-saturated ascent. The use of conditional instability relative to ice processes becomes more important at colder temperatures, since the difference between relative humidity with respect to ice and water increases at progressively colder subfreezing temperatures (Pruppacher and Klett 1997), a difference that affects both the height of the level of free convection and the CAPE. The most unstable CAPE from each of the soundings, determined by lifting air parcels from above the frontal surface, is listed across the top of the figure. The air was stable on the left side of the figure where no red lines appear. To properly interpret Fig. 4, it is important to note how the storm passed over the profiler (see red square on each panel of Fig. 3). The cyclone approached the profiler position from the southwest. Prior to about 0000 UTC, the profiler was located within the northern side of the comma head (not shown), where clouds had structure corresponding to the left side of Fig. 2. Based on the profiler SNR, the stratiform clouds were 4–5 km deep on the far northern side of the comma head (1400–2000 UTC in Fig. 4), deepening to about 7 km between 2100 and 0000 UTC. Note that the profiler (Fig. 4) is not as sensitive to small hydrometeors near cloud top as the WCR (Fig. 2). Comparison of the data from the WCR and profiler in the stratiform region (Figs. 2 and 4) shows that the cloud-top-generating cells between 7.5 and 9.0 km went undetected by the profiler. At approximately 0000 UTC, the cyclone moved sufficiently northward that the profiler was located beneath the convective region of the comma head. From this point onward, air above the frontal surface was conditionally unstable (with respect to ice processes) and convective cells erupted above the frontal surface. Convective cells passed directly over the profiler between 0200 and 0300 UTC, and again between 0430 and 0600 UTC. These cells extended upward to about 8 km, closely matching the top altitudes of the cells appearing in Fig. 2. Snow was falling throughout the comma head region in the vicinity of the profiler and occasionally lightning flashes occurred both west and east of the profiler location during this period (Fig. 3).
Time section of the MIPS 915-MHz profiler SNR, θei, and winds (flag = 25 m s−1, full barb = 10 m s−1, half barb = 5 m s−1) derived from sequential rawinsondes from 1400 UTC 8 Dec to 0800 UTC 9 Dec 2009. The lower and upper red dashed lines denote, respectively, the midlevel θe maxima and the equilibrium level for ice-saturated ascent. The region between the upper and lower blue dashed lines denotes the region where air is saturated with respect to ice. The black dashed line is the 0°C isotherm. Temperatures across all of the cross section except within the black dashed line were below 0°C and precipitation fell as snow. The most unstable CAPE, determined from each of the soundings by lifting parcels above the frontal inversion, is listed across the top of the figure.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
Figure 5 shows a detailed view of the profiler radial velocity, spectral width, and SNR, as well as the field mill data during the passage of stratiform clouds over the profiler between 2200 and 2300 UTC. The finescale periodic oscillations in the field mill data are due to the operation of the mill and not to any fluctuations in the background electric field. These data clearly confirm that the cloud was stratiform. There were no temporal fluctuations in the vertical particle motion, spectral width, SNR, or electric field.
(a) Vertical radial velocity, (b) spectral width, and (c) SNR from the vertical beam of the MIPS 915-MHz profiler for the period 2200–2300 UTC 8 Dec 2009. Sounding-derived temperatures are overlaid. (d) Surface electric field from the field mill for the same time period.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
Figure 6 shows the same fields, but for the period 0200–0300 UTC when a convective cell moved across the profiler. The strongest convective velocities (both updrafts and downdrafts) occurred between −20° and −30°C. Updrafts and downdrafts of 4–5 m s−1 occurred within this region. The spectral width values increase to 4–6 m s−1 from background stratiform values of about 1–2 m s−1. Spectral width increases in response to both increased turbulence and spread in particle terminal velocities. Both of these processes enhance the probability of collisions between particles and the likelihood of charge separation in the updrafts. The electric field mill data show clear evidence that charge separation was occurring, based on the fluctuations in the electric field. Several lightning flashes indeed occurred in nearby cells during this time period. Figure 7 shows the same fields, but for cells passing over the profiler between 0415 and 0545 UTC. Here the updrafts emerge upward from approximately the −10°C level. The stronger updrafts are again marked by increased spectral width. The greatest fluctuations in the surface electric field occurred with the passage of these cells. The specific convective cells observed by the profiler and aircraft did not produce lightning. However, this same time period coincided with lightning flashes from cells within about 100-km radius of the profiler. In the interpretations presented here, based on the proximity of lightning-producing cells and the similarity of the environment, it is inherently assumed that the actual thunderstorm cells share similar properties to those observed and analyzed.
As in Fig. 5, but for 0200–0300 UTC 9 Dec 2009.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
As in Fig. 5, but for 0415–0545 UTC 9 Dec 2009.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
After the flight leg in Fig. 2, the C-130 made repeated passes through cells on the southern side of the comma head (Fig. 3d). Figure 8 shows the WCR data for a direct pass through a developing and mature convective cell by the C-130. The upward vertical air velocities in these cells, after accounting for particle terminal velocities (Rosenow et al. 2014) were similar to that reported by Rosenow et al., generally 1–8 m s−1. Based on the environmental CAPE associated with these cells (50–250 J kg−1; see Fig. 4), the updrafts are likely reduced by entrainment to about half their theoretical maximum values based on adiabatic parcel ascent. The frontal surface at about 2.5-km altitude in these figures is marked by significant shear—the falling streamers of particles reverse direction toward the south as they fall into the air mass below the front. Noninductive charging within storms requires the presence of graupel and ice particles in the presence of supercooled water (MacGorman and Rust 1998). Figures 9a and 9b show particle images from the 2DC spectrometer probe between 0602 and 0603 UTC within the developing cell on the northern (left) side of Fig. 8. The data were collected as the aircraft descended from the −14° to the −10°C level where charge separation commonly occurs in convective storms. The images are characteristic of both rimed particles and smaller ice crystals. Figure 9c shows the liquid water content measured by two probes on the aircraft, the King hot wire probe and Gerber Particle Volume Monitor (PVM) probe. Both show that supercooled liquid water was present. The magnitude of the liquid water content differed on the probes, but even with the lower measurement, the water content exceeded 0.1 g m−3 in some locations. The presence of supercooled water was confirmed independently with the Rosemount Icing Probe. These conditions were observed in several cells penetrated by the aircraft. These data provide evidence that noninductive charging likely occurred in these elevated convective cells.
(a) Equivalent radar reflectivity factor and (b) vertical radial velocity through a developing and mature elevated convective cell at 0600 UTC 9 Dec 2009. The black dashed line is the flight track.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
2DC images of rimed particles at (a) 0602:12 and (b) 0602:51 UTC 9 Dec 2009. (c) Liquid water content measured by the King (blue) and Gerber Particle Volume Monitor (red) probes.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
b. 24 November 2009
A weak cyclone passed across the central plains on 24 November 2009 (Fig. 10). This event was quite different from the previous case in that the storm was a cutoff low rather than an open wave (Fig. 11), and the convection was near-surface-based rather than elevated. The cyclone produced little precipitation, and what fell was in the form of rain rather than snow. The comma head of the storm consisted of two narrow convective bands (Fig. 10). The southern band produced a number of lightning flashes very close to the profiler between 1230 and 1430 UTC (Figs. 10a,b), and again between 1730 and 1930 UTC (Figs. 10c,d). The C-130 also sampled the band near the profiler location between 1955 and 2010 UTC (see track in Fig. 10d).
Composite radar reflectivity analysis at (a) 1300, (b) 1400, (c) 1800 and (d) 1900 UTC 24 Nov 2009 overlain with sea level pressure analyses (2-hPa interval). The C-130 flight track between 1909 and 2055 UTC is shown in yellow in (d). Lightning flashes are denoted by the white dots in red circles, and the red square is the location of the MIPS 915-MHz profiler.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
The 500-hPa analysis of height and absolute vorticity at 1800 UTC 24 Nov 2009.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
Figure 12 shows the MIPS 915-MHz profiler SNR for the time period 0900 UTC 24 November–0000 UTC 25 November 2009. Superimposed on the figure are rawinsonde measured winds and θei (θe below the melting level). Air was saturated with respect to ice between the melting level and the blue dashed line. The level of maximum θei and the equilibrium level are shown as the two dashed red lines. Note the transition from southerly winds to deep weak easterly winds, to northerly winds on the soundings, as the center of the cutoff low passed south of the profiler. During the period of southerly winds, the most unstable CAPE was 639 J kg−1. The CAPE values decreased substantially with the onset of easterly winds. Two periods of convection over the profiler are evident. During these periods, the CAPE values ranged from 13 to 163 J kg−1.
Time section of the MIPS 915-MHz profiler SNR, θei (θe below the melting level), and winds (full barb = 10 m s−1, half barb = 5 m s−1) derived from sequential rawinsondes from 0900 UTC 24 Nov to 0000 UTC 25 Nov 2009. The lower and upper red dashed lines denote, respectively, the low-level θe maxima and the equilibrium level for ice-saturated ascent. The region between the upper blue dashed line and dashed black line denotes the region where air is saturated with respect to ice. The black dashed line is the 0°C isotherm. All precipitation fell as rain. The most unstable CAPE determined from each of the soundings is listed across the top of the figure.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
Figure 13 shows a detailed view of the profiler radial velocity, spectral width, and SNR, as well as the field mill data for the passage of the first convective band. During this time, three lightning flashes were visible from location of the profiler, and several others occurred in the near vicinity. The bright band is evident on the figures both in the SNR and vertical radial velocity fields. The 5–6 m s−1 updraft and 5–6 m s−1 maxima in spectral width occurred nearly simultaneously with two lightning flashes near the profiler. Again, the stronger updrafts appear to be originating near the −10°C level.
As in Fig. 5, but for 1230–1430 UTC 24 Nov 2009.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
Figure 14 shows the profiler data for the second band. In this case, parcels entering the convective cells likely entered from near the surface, and updrafts appear in some cases to extend below the melting level. Updrafts of 2–3 m s−1 were still present above the −10°C level. The number of flashes in the general vicinity of the profiler decreased during this period (cf. 1300–1400 and 1800–1900 UTC in Fig. 10), consistent with the weaker vertical motions. All of the data presented for this case again are consistent with noninductive charging through graupel–ice particle collisions.
As in Fig. 5, but for 1730–2000 UTC 24 Nov 2009.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
Figure 15 shows the WCR data during the pass through the convective band. The pass was made near −20°C, based on aircraft measurements within the precipitation. The temperatures shown on the figure were based on the nearest sounding and are about 2°C warmer than measurements made by instruments on the aircraft at the same level. The data show vertical particle motions ranging up to 5 m s−1 across the band. Accounting for the terminal velocities of the graupel (see microphysical data in Fig. 16), the maximum updrafts sampled by the aircraft were probably on the order of 6–8 m s−1. Turbulence during this pass caused the aircraft roll angle to vary slightly more than normal. Although the vertical radial velocities were corrected for roll using techniques described in Rosenow et al. (2014), any residual error would cause some uncertainty in these values.
(a) Ze from the WCR from 1955 to 2010 UTC 24 Nov 2009 and θei from the RUC initialization valid at 2000 UTC along the same cross section. The blue (black) line denotes the 70% (100%) RHi contour above the melting level. (b) Vertical radial velocities measured by the WCR, overlain with θei from the RUC initialization. The horizontal black line in both panels is the C-130 flight track.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
2DC images of rimed particles at (a) 1956:30 and (b) 1956:40 UTC. (c) Liquid water content measured by the King (blue) and Gerber Particle Volume Monitor (red) probes during the pass through one of the cells in Fig. 15.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
Figure 16 shows 2DC particle images and supercooled liquid water measurements during the passage through a developing cell between 1956 and 1958 UTC. Graupel and ice crystals are clearly present simultaneously with supercooled water, again providing conditions where noninductive charging can occur within the clouds. The attenuation of the radar signal above and below the aircraft in Fig. 15 is associated with the supercooled water present in this cell. The microphysical structure of this wintertime convective storm therefore had similarities to the previous case, although the liquid water content of the clouds was larger in this case by a factor of 3.
4. Climatology of lightning flashes within the comma head
The 8–9 December 2009 cyclone discussed in the previous section was a deep cyclone that produced over 30–40 cm (12–16 in.) of snow beneath the comma head region of the storm. Although many cyclones were observed during PLOWS, no others had this intensity and produced lightning near the instrumentation platforms. However, there is reason to believe that other lightning-producing winter storms have similar structure in the comma head region. For example, one year after the field campaign, a blizzard of comparable magnitude and structure crossed the midwestern United States. The 1–2 February 2011 blizzard was a disastrous storm, paralyzing the city of Chicago and stranding cars for miles along Chicago’s Lake Shore Drive. Over 15 cm of ice pellets fell in central Illinois and deep heavy snow fell in a large swath across the Midwest. Figure 17 shows a visible satellite image of the storm at 2130 UTC before darkness set in. The satellite image clearly shows two different cloud types in Illinois. North of the red line on the figure, the clouds have a stratiform appearance, while south of the line they are bright and appear more convective in character. Over the 2 h following this image, the cyclone moved northward (Fig. 18), along with the position of the boundary between the cloud types.
Visible satellite image of the 1–2 Feb 2011 cyclone at 2130 UTC 1 Feb 2011. The red dots show the locations of lightning flashes near the dry slot–comma head interface between 2000 and 2200 UTC 1 Feb. The yellow dots are between 0000 and 0200 UTC 2 Feb 2011, 2.5–4.5 h after the image was taken. No flashes were recorded near the interface between 2200 and 0000 UTC. The cyclone was propagating northwestward during the period. The red line marks the boundary between clouds that appear stratiform and those appearing convective. The black line denotes the cross section location for Fig. 19.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
WSR-88D composites of the 1–2 Feb 2011 cyclone at (a) 2200 UTC 1 Feb, (b) 0000 UTC 2 Feb, and (c) 0200 UTC 2 Feb 2011, corresponding to the cross sections shown in Fig. 19. The red lines denote the location of each cross section.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
Lightning flashes occurred at the locations of the red dots between 2000 and 2200 UTC, and at the location of the yellow dots between 0000 and 0200 UTC. No flashes occurred between 2200 and 0000 UTC in the region near Illinois. The series of cross sections in Fig. 19 shows that the increase in lightning frequency after 0000 UTC was coincident with the advance of low-θei air associated with the cyclone’s dry slot over lower-level moist air, creating elevated instability. These data suggest that the advancement of upper-tropospheric dry air within the cyclone’s dry slot over the low-level moist air in the comma head, in a manner similar to the 8–9 December 2009 storm, may be key to creating elevated instability and the triggering of convection and occasional wintertime lightning in this storm.
Cross sections of vertical air velocity overlain with θei from 1-h RUC forecasts valid at (a) 2200 UTC 1 Feb, (b) 0000 UTC 2 Feb, and (c) 0200 UTC 2 Feb 2011 along the cross sections shown in Fig. 18. The black dashed (dotted) line marks a maximum (minimum) in θei. The blue (black) thick lines denote the 70% (100%) RHi contours.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
This scenario is consistent with the sounding observations of Market et al. (2006), who noted that during thundersnow events, the most unstable parcel was found 30–50 hPa above the top of the frontal inversion accompanied by significantly drier air about 100 hPa above the level of the most unstable parcel. The overrunning of dry air is also consistent with the findings of Grim et al. (2007). In that study, high-resolution dropsonde measurements were made across the comma head of two strong winter cyclones. In both cases, the most significant features were the advance of the dry air intrusion over moist air, which isolated the deeper warm, moist airflow to the north and created a sharp relative humidity gradient at the leading edge of the dry-air intrusion.
The evidence presented in previous case studies and in this paper suggest that lightning within the comma head region of winter cyclones should preferentially occur on the southern side of the comma head where dry air aloft may overrun low-level moist air and create convective instability. Below, the distribution of lightning flashes across the comma head region of cyclones during 4 years, the two PLOWS observation periods, and the winter seasons bounding the PLOWS campaign, 2007/08 and 2010/11, are examined to determine if lightning flashes occur preferentially on the southern side of the comma head.
The procedure used to create the climatology is shown in Fig. 20 for the 1–2 February 2009 cyclone. Lightning flashes were considered only when the surface low pressure center was within the red outline in Fig. 20. This region was chosen because it conforms to the most common region for elevated winter thunderstorms, as defined by the studies of Curran and Pearson (1971), Colman (1990), Holle and Cortinas (1998), and Market et al. (2002). A line (line y) was then drawn northward from the center of low pressure, and a second line (line x) was drawn along the southern boundary of the radar echo comprising the comma head. Line x was drawn subjectively by eye, and ignored small-scale variability in the echo along the southern boundary (e.g., Fig. 20). Only lightning flashes west of line y and north of line x were considered in the climatology. Note that this criterion excluded some of the flashes reported in the case study that were somewhat east of the low pressure center. The positions of lines x and y moved with time as the cyclone progressed across the domain defined by the red border. These procedures were adopted to insure that the flashes were confined to the comma head and were in storms traversing the northern half of the country, where colder weather is common in the winter season. Arctic fronts and inverted troughs also produced lightning flashes during the study period. These were not included in this study. Also, two cyclones (11 February and 16 November 2009) developed lines of convection extending from the comma head and into the cyclone’s dry slot. Both lines produced considerable lightning. These cyclones were eliminated from the climatology because the lines were not features of the comma head, but rather were forced by other processes. Subject to satisfying all of the criteria listed above, cyclones were required to produce at least one lightning flash (cloud to ground, in-cloud, of either polarity) to be included in the climatology. Sixteen cyclones met these criteria. The tracks of these cyclones across the analysis domain are shown in Fig. 21.
Radar reflectivity, surface pressure (hPa), and frontal analysis for the 1–2 Feb 2011 cyclone. The red dot represents the position of a lightning flash. Lines x, y, A, and B are described in the text.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
Tracks of all surface low pressure centers for all cyclones included in the climatology.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
Weather Surveillance Radar-1988 Doppler (WSR-88D) level-2 data and lightning data were imported into Google Earth for all cyclones. Using Google Earth mapping capabilities, a line was drawn normal to line x through the position of the flash for each lightning flash in each cyclone. The distance from line x to the lightning flash (line A in Fig. 20) and the total distance across the comma head (lines A and B in Fig. 20) were recorded. The northern boundary of the echo (the terminus of line B) was defined by the 5-dBZ contour. The data for 2636 lightning flashes were recorded using this procedure.
The NLDN (Orville et al. 2011) reports both cloud-to-ground and in-cloud (IC) flashes. The CG flashes are reported as either positive polarity (PP) or negative polarity (NP). CG flashes often fork so that ground flashes may occur in more than one location with the same flash. The NLDN reports each ground flash as a separate event although several may be components of the same visual event. For recording purposes, we counted events using four different methods. First, each individual event recorded by the NLDN (CG flash or IC flash) was counted. These are labeled as “Total flashes” in Table 1. Recounting was performed after eliminating duplicate records of CG flashes occurring at the exact same time and location (see column labeled “Total flashes, 0 s, 0 m” in Table 1). An additional recount was made after further eliminating duplicate records for CG flashes occurring within 0.5 s and 110 m of each other. Finally, a fourth recounting was made, this time eliminating duplicate records where two more CG flashes occurred within 1 s and 220 m of each other. The data after each elimination process are compiled in Table 1, as well as the number of PP and NP events for all CG flashes. Distributions of flash positions for each cyclone were then developed to determine their relative positions within each storm’s comma head structure.
Lightning flash statistics for 16 cyclones included in the climatology.
Figure 22 shows the distribution of lightning flashes as a function of the fractional distance [A/(A + B) in Fig. 20] across the comma head for the 16 cyclones. The panels are ordered according to the number of flashes that were observed. The distributions show a strong bias toward the southern side of the comma head region of the cyclones, consistent with the distribution of convection implied by the case studies in the previous section. About 55% of the flashes were associated with cloud-to-ground flashes while 45% were in-cloud flashes. Of the ground flashes, 96% had negative polarity. This value is comparable with the United States annual (all seasons) climatology of Orville and Huffines (2001), who reported that 91% of all ground flashes in the United States lower the negative charge to ground, but somewhat higher than the 80% negative polarity value for winter storms reported by Market and Becker (2009). The common occurrence of negative-polarity lightning differs substantially from wintertime lightning associated with lake-effect storms and suggests that elevated convection associated with cyclone comma head circulations are not sufficiently sheared and that flashes between the storm anvil and the ground are uncommon. The WCR and profiler data for the case studies presented provides support for this interpretation, although more data are needed to definitively determine if this is the case.
The fractional distance across the comma head of lightning flashes in 16 cyclones.
Citation: Journal of the Atmospheric Sciences 71, 5; 10.1175/JAS-D-13-0253.1
5. Summary and conclusions
This paper presented analyses of the finescale structure of convection within the comma head of winter cyclones, and a climatology analyzing the distribution of lightning across the comma head. The work was based on detailed measurements from the Profiling of Winter Storms (PLOWS) field campaign. The analyses were based on data from the University of Wyoming Cloud Radar and microphysics probes on the C-130 aircraft, the University of Alabama Mobile Integrated Profiling System 915-MHz profiler, and the NCAR Mobile Integrated Sounding System and University of Missouri Sounding Systems. The PLOWS data were supplemented by measurements from the National Lightning Detection Network for the two winter seasons, plus two additional winter seasons outside the PLOWS field campaign.
These analyses, considered in the context of past studies, support the following points regarding the formation and nature of lightning within the comma head of continental winter cyclones:
Based on data presented here and in Rosenow et al. (2014), case studies presented by Grim et al. (2007), and the sounding analysis of Market et al. (2006), upper-tropospheric dry air associated with a cyclone’s dry slot frequently intrudes over lower-level Gulf air in the comma head of strong cyclones, creating two zones of precipitation within the comma head: a northern zone characterized by deep stratiform clouds and topped by cloud-top-generating cells, and a southern zone marked by elevated convection. Lightning, when it occurs, appears to originate from the elevated convection within the southern zone.
Radar measurements of vertical radial velocity within the convection from both airborne radar and ground-based profilers, after adjustment to account for particle terminal velocities, show that the convection in winter storms is associated with updrafts that can approach 6–8 m s−1. The CAPE measured in the environments in which these cells developed was typically 50–250 J kg−1 (see also Market et al. 2012), suggesting that the updrafts were reduced by entrainment to about half their theoretical maximum values based on parcel ascent. These vertical velocities measured in the wintertime convection are somewhat below thresholds for rapid electrification proposed by Zipser and Lutz (1994) for oceanic convection. In their work, they found that rapid electrification occurs with mean updrafts exceeding 6–7 m s−1 and peak updrafts of 10–12 m s−1. However, over land, communication towers and manmade structures may provide focal points for lightning and can be expected to reduce the threshold required for cloud-to-ground flashes to occur. For example, Lyons et al. (2012) showed that the lightning flashes occurring in the 1 February 2011 storm discussed in section 4 all occurred in association with tall objects such as towers and buildings. They termed these “self-initiated upward lightning” events, found that nearly all were negative as reported by the NLDN, and all likely occurred as the towers and buildings launched upward positive leaders into what they presumed were negatively charged cloud bases.
Microphysical measurements within convective updrafts showed the simultaneous presence of graupel, ice particles, and supercooled water within the temperature range from −10° to −20°C. These conditions support noninductive charging as an important mechanism charging winter storms.
Over 90% of the cloud-to-ground flashes in the 16 winter cyclones studies in the climatology had negative polarity. This finding differs significantly from winter lightning associated with lake-effect storms, and suggests that most cells are not sufficiently sheared aloft to create conditions supporting positive polarity ground flashes.
About 55% of the flashes were cloud-to-ground while 45% were in-cloud flashes.
The applicability of these findings to the comma head region of cyclones in other locations is not clear. Climatological studies of lightning in winter cyclones on the U.S. East Coast, for example, suggest that lightning is uncommon. Nevertheless, there is some suggestion from satellite imagery that air within the dry slot may overrun low-level moist air and trigger convection and lightning on the southern side of the comma head in some storms (e.g., Fig. 5 of Stuart 2001). Evidence from studies of East Coast cyclones also suggests that stability varies across the comma head, from a conditionally unstable environment on the equatorward side to a stable environment on the poleward side (e.g., Nicosia and Grumm 1999; Novak et al. 2008, 2009, 2010). Future research using finescale measurements similar to those reported here will be needed to clarify these and other outstanding questions remaining about wintertime lightning.
Acknowledgments
The authors thank the staff at the National Center for Atmospheric Research Environmental Observing Laboratory, particularly Alan Schanot and the Research Aviation Facility staff for their efforts with the C-130, the staff of the University of Wyoming King Air facility for their support of the WCR deployment, and William Brown and the staff of the In-situ Sensing Facility for their help with the MISS deployment. We thank Major Donald K. Carpenter and the U.S. Air Force Peoria National Guard for housing the C-130 during the project. The composite radar analyses appearing in Figs. 3, 10, and 18 were provided by the Iowa Environmental Mesonet maintained by the Iowa State University Department of Agronomy. This work was funded under National Science Foundation Grants ATM-0833828 and AGS-1247404 to the University of Illinois, ATM-0833995 and AGS-1247412 to the University of Alabama-Huntsville, and AGS-1247473 to the University of Wyoming. Special thanks go to Ron Holle and Vaisala, Inc. for sharing data from the National Lightning Detection Network.
REFERENCES
Beldon, W. S., 1927: A thunderstorm with rain, hail, sleet and snow. Mon. Wea. Rev., 55, 133.
Bennetts, D. A., and B. J. Hoskins, 1979: Conditional symmetric instability—A possible explanation for frontal rainbands. Quart. J. Roy. Meteor. Soc., 105, 945–962.
Biagi, C. J., K. L. Cummins, K. E. Kehoe, and E. P. Krider, 2007: National Lightning Detection Network (NLDN) performance In southern Arizona, Texas, and Oklahoma in 2003–2004. J. Geophys. Res., 112, D05208, doi:10.1029/2006JD007341.
Bluestein, H. B., 1993: Synoptic–Dynamic Meteorology in Midlatitudes. Vol. 2, Observations and Theory of Weather Systems, Oxford University Press, 594 pp.
Bosart, L. F., and F. Sanders, 1986: Mesoscale structure in the megalopolitan snowstorm of 11–12 February 1983. Part III: A large-amplitude gravity wave. J. Atmos. Sci., 43, 924–939.
Brook, M., M. Nakano, P. Krehbiel, and T. Takeuti, 1982: The electrical structure of the Hokuriku winter thunderstorms. J. Geophys. Res., 87, 1207–1215.
Butterworth, J., 1895: Thunder and lightning in December. Mon. Wea. Rev., 23, 464.
Cohn, S. A., V. Grubišić, and W. O. J. Brown, 2011: Wind profiler observations of mountain waves and rotors during T-REX. J. Appl. Meteor. Climatol., 50, 826–843.
Colman, B. R., 1990: Thunderstorms above frontal surfaces in environments without positive CAPE. Part I: A climatology. Mon. Wea. Rev., 118, 1103–1121.
Crowe, C. C., P. S. Market, B. P. Pettegrew, C. J. Melick, and J. Podzimek, 2006: An investigation of thundersnow and deep snow accumulations. Geophys. Res. Lett.,33, L24812, doi:10.1029/2006GL028214.
Cummins, K. L., and M. J. Murphy, 2009: An overview of lightning locating systems: History, technique, and data uses, with an in-depth look at the U.S. NLDN. IEEE Trans. Electromagn. Compat., 51, 499–518.
Cummins, K. L., M. J. Murphy, E. A. Bardo, W. L. Hiscox, R. B. Pyle, and A. E. Pifer, 1998: A combined TOA/MDF technology upgrade of the U.S. National Lightning Detection Network. J. Geophys. Res., 103, 9035–9044.
Cummins, K. L., J. A. Cramer, C. J. Biagi, E. P. Krider, J. Jerauld, M. A. Uman, and V. A. Rakov, 2006: The U.S. National Lightning Detection Network: Post-upgrade status. Extended Abstracts, Second Conf. on Meteorological Applications of Lightning Data, Atlanta, GA, Amer. Meteor. Soc., 6.1. [Available online at https://ams.confex.com/ams/Annual2006/techprogram/paper_105142.htm.]
Curran, J. T., and A. D. Pearson, 1971: Proximity soundings for thunderstorms with snow. Preprints, Seventh Conf. on Severe Local Storms, Kansas City, MO, Amer. Meteor. Soc., 118–119.
Emanuel, K. A., 1983: The Lagrangian parcel dynamics of moist symmetric instability. J. Atmos. Sci., 40, 2368–2376.
Grim, J. A., R. M. Rauber, M. K. Ramamurthy, B. F. Jewett, and M. Han, 2007: High resolution observations of the trowal–warm frontal region of two continental winter cyclones. Mon. Wea. Rev., 135, 1629–1646.
Halcomb, C. E., and P. S. Market, 2003: Forcing, instability, and equivalent potential vorticity in a Midwest USA convective snowstorm. Meteor. Appl., 10, 273–280.
Herschel, A. S., 1888: Lightning in snowstorms. Quart. J. Roy. Meteor. Soc., 14, 222–225.
Holle, R. L., and A. I. Watson, 1996: Lightning during two central U.S. winter precipitation events. Wea. Forecasting, 11, 599–614.
Holle, R. L., and J. V. Cortinas Jr., 1998: Thunderstorms observed at surface temperatures near and below freezing across North America. Preprints, 19th Conf. on Severe Local Storms, Minneapolis, MN, Amer. Meteor. Soc., 705–708.
Holle, R. L., R. E. López, K. W. Howard, K. L. Cummins, M. D. Malone, and E. P. Krider, 1997: An isolated winter cloud-to-ground lightning flash causing damage and injury in Connecticut. Bull. Amer. Meteor. Soc., 78, 437–441.
Hunter, S. M., S. J. Underwood, R. L. Holle, and T. L. Mote, 2001: Winter lightning and heavy frozen precipitation in the southeast United States. Wea. Forecasting, 16, 478–490.
Kitagawa, N., 1992: Charge distribution of winter thunderclouds. Res. Lett. Atmos. Elec., 12, 143–153.
Kitagawa, N., and K. Michimoto, 1994: Meteorological and electrical aspects of winter thunderclouds. J. Geophys. Res., 99, 10 713–10 721.
Knight, D. J., and P. V. Hobbs, 1988: The mesoscale and microscale structure and organization of clouds and precipitation in midlatitude cyclones. Part XV: A numerical modeling study of frontogenesis and cold-frontal rainbands. J. Atmos. Sci., 45, 915–931.
Lackmann, G. M., 2001: Analysis of a surprise western New York snowstorm. Wea. Forecasting, 16, 99–116.
Lyons, W. A., and Coauthors, 2012: Different strokes: Researching the unusual lightning discharges associated with sprites and jets and atypical meteorological regimes. Proc. 22nd Int. Lightning Detection Conf., Broomfield, CO, Vaisala. [Available online at http://www.vaisala.com/en/events/ildcilmc/Documents/Different%20Strokes%20Researching%20the%20Unusual%20Lightning%20Discharges%20Associated%20with.pdf.]
MacGorman, D. R., and W. D. Rust, 1998: The Electrical Nature of Storms. Oxford University Press, 422 pp.
Maekawa, Y., S. Fukao, Y. Sonoi, and F. Yoshino, 1992: Dual polarization radar observations of anomalous wintertime thunderclouds in Japan. IEEE Trans. Geosci. Remote Sens., 30, 838–844.
Market, P. S., and A. E. Becker, 2009: A study of lightning flashes attending periods of banded snowfall. Geophys. Res. Lett., 36, L01809, doi:10.1029/2008GL036317.
Market, P. S., C. E. Halcomb, and R. L. Ebert, 2002: A climatology of thundersnow events over the contiguous United States. Wea. Forecasting, 17, 1290–1295.
Market, P. S., and Coauthors, 2006: Proximity soundings of thundersnow in the central United States. J. Geophys. Res., 111, D19208, doi:10.1029/2006JD007061.
Market, P. S., R. L. Ebert-Cripe, and M. Bodner, 2007: Case study of a long-lived thundersnow event. Nat. Wea. Dig.,31 (2), 103–119; Corrigendum, 32 (1), 77–86.
Market, P. S., K. Crandall, and R. Rauber, 2012: High-resolution rawinsonde observations of the cold-sector precipitation regions in transient mid-latitude extratropical cyclones. Natl. Wea. Dig., 36, 3–8.
Martin, J. E., 1998: The structure and evolution of a continental winter cyclone. Part II: Frontal forcing of an extreme snow event. Mon. Wea. Rev., 126, 329–348.
Michimoto, K., 1993: A study of radar echoes and their relation to lightning discharges of thunderclouds in the Hokuriku District. Part II: Observation and analysis of “single-flash” thunderclouds in midwinter. J. Meteor. Soc. Japan, 71, 195–204.
Moore, J. T., and P. D. Blakley, 1988: The role of frontogenetical forcing and conditional symmetric instability in the Midwest snowstorm of 30–31 January 1982. Mon. Wea. Rev., 116, 2155–2171.
Moore, J. T., A. C. Czarnetzki, and P. S. Market, 1998: Heavy precipitation associated with elevated thunderstorms formed in a convectively unstable layer aloft. Meteor. Appl., 5, 373–384.
Moore, J. T., C. E. Graves, S. Ng, and J. L. Smith, 2005: A process-oriented methodology toward understanding the organization of an extensive mesoscale snowband: A diagnostic case study of 4–5 December 1999. Wea. Forecasting, 20, 35–50.
Moore, P. K., and R. E. Orville, 1990: Lightning characteristics in lake-effect thunderstorms. Mon. Wea. Rev., 118, 1767–1782.
Moore, P. K., and V. P. Idone, 1999: Cloud-to-ground lightning at low surface temperatures. Proc. 11th Int. Conf. on Atmospheric Electricity, Guntersville, AL, International Commission on Atmospheric Electricity, 472–475.
Nicosia, D. J., and R. H. Grumm, 1999: Mesoscale band formation in three major northeastern United States snowstorms. Wea. Forecasting, 14, 346–368.
Novak, D. R., B. A. Colle, and S. E. Yuter, 2008: High-resolution observations and model simulations of the life cycle of an intense mesoscale snowband over the northeastern United States. Mon. Wea. Rev., 136, 1433–1456.
Novak, D. R., B. A. Colle, and R. McTaggart-Cowan, 2009: The role of moist processes in the formation and evolution of mesoscale snowbands within the comma head of northeast U.S. cyclones. Mon. Wea. Rev., 137, 2662–2686.
Novak, D. R., B. A. Colle, and A. R. Aiyyer, 2010: Evolution of mesoscale precipitation band environments within the comma head of northeast U.S. cyclones. Mon. Wea. Rev., 138, 2354–2374.
Orville, R. E., 1990: Winter lightning along the East Coast. Geophys. Res. Lett., 17, 713–715.
Orville, R. E., 1993: Cloud-to-ground lightning in the blizzard of ’93. Geophys. Res. Lett., 20, 1367–1370.
Orville, R. E., and G. R. Huffines, 2001: Cloud-to-ground lightning in the United States: NLDN results in the first decade, 1989–98. Mon. Wea. Rev., 129, 1179–1193.
Orville, R. E., G. R. Huffines, W. R. Burrows, R. L. Holle, and K. L. Cummins, 2002: The North American Lightning Detection Network (NALDN)—First results: 1998–2000. Mon. Wea. Rev., 130, 2098–2109.
Orville, R. E., G. R. Huffines, W. R. Burrows, and K. L. Cummins, 2011: The North American Lightning Detection Network (NALDN)—Analysis of flash data: 2001–09. Mon. Wea. Rev., 139, 1305–1322.
Persson, P. O. G., and T. T. Warner, 1995: The nonlinear evolution of idealized, unforced, conditional symmetric instability: A numerical study. J. Atmos. Sci., 52, 3449–3474.
Pessi, A. T., and S. Businger, 2009: The impact of lightning data assimilation on a winter storm simulation over the North Pacific Ocean. Mon. Wea. Rev., 137, 3177–3195.
Pettegrew, B. P., 2008: Analysis of cloud and cloud-to-ground lightning in winter convection. Ph.D. dissertation, University of Missouri, 204 pp.
Pettegrew, B. P., P. S. Market, R. A. Wolf, R. L. Holle, and N. W. S. Demetriades, 2009: A case study of severe winter convection in the Midwest. Wea. Forecasting, 24, 121–139.
Pruppacher, H. R., and J. D. Klett, 1997: Microphysics of Clouds and Precipitation. Kluwer Academic, 954 pp.
Rakov, V. A., and M. A. Uman, 2003: Lightning: Physics and Effects. Cambridge University Press, 687 pp.
Rosenow, A. R., D. M. Plummer, R. M. Rauber, G. M. McFarquhar, B. F. Jewett, D. Leon, and J. M. Keeler, 2014: Vertical motions within generating cells and elevated convection in the comma head of winter cyclones. J. Atmos. Sci., 71, 1538–1558.
Schneider, R. S., 1990: Large-amplitude mesoscale wave disturbances within the intense Midwest extratropical cyclone of 15 December 1987. Wea. Forecasting, 5, 533–558.
Schultz, D. M., 1999: Lake-effect snowstorms in northern Utah and western New York with and without lightning. Wea. Forecasting, 14, 1023–1031.
Schultz, D. M., and P. N. Schumacher, 1999: The use and misuse of conditional symmetric instability. Mon. Wea. Rev., 127, 2709–2732.
Schultz, D. M., and Coauthors, 2002: Understanding Utah winter storms: The Intermountain Precipitation Experiment. Bull. Amer. Meteor. Soc., 83, 189–210.
Schumacher, R. S., D. M. Schultz, and J. A. Knox, 2010: Convective snowbands downstream of the Rocky Mountains in an environment with conditional, dry symmetric, and inertial instabilities. Mon. Wea. Rev., 138, 4416–4438.
Steiger, S. M., R. Hamilton, J. Keeler, and R. E. Orville, 2009: Lake-effect thunderstorms in the lower Great Lakes. J. Appl. Meteor. Climatol., 48, 889–902.
Stuart, N., 2001: Multi-dimensional analysis of an extreme thundersnow event during the 30 December 2000 snowstorm. Preprints, 18th Conf. on Weather Analysis and Forecasting, Fort Lauderdale, FL, Amer. Meteor. Soc., 223–226.
Takeuti, T., M. Nakano, M. Nagatani, and H. Nakada, 1973: On lightning discharges in winter thunderstorms. J. Meteor. Soc. Japan, 51, 494–496.
Takeuti, T., M. Nakano, and Y. Yamamoto, 1976: Remarkable characteristics of cloud-to-ground discharges observed in winter thunderstorms in Hokuriku area, Japan. J. Meteor. Soc. Japan, 54, 436–440.
Takeuti, T., M. Nakano, M. Brook, D. J. Raymond, and P. Krehbiel, 1978: The anomalous winter thunderstorms of the Hokuriku coast. J. Geophys. Res., 83, 2385–2394.
Trapp, R. J., D. M. Schultz, A. V. Ryzhkov, and R. L. Holle, 2001: Multiscale structure and evolution of an Oklahoma winter precipitation event. Mon. Wea. Rev., 129, 486–501.
Wang, P.-K., 1980: On the relationship between winter thunder and the climatic change in China in the past 2000 years. Climatic Change, 3, 37–46.
Wang, P.-K., and J.-H. Chu, 1982: Unusual lightning events in ancient Chinese literature. Weatherwise, 35, 119–122.
Wang, Z., and Coauthors, 2012: Single aircraft integration of remote sensing and in situ sampling for the study of cloud microphysics and dynamics. Bull. Amer. Meteor. Soc., 93, 653–668.
Ward, J. G., K. L. Cummins, and E. P. Krider, 2008: Comparison of the KSC-ER cloud-to-ground lightning surveillance system (CGLSS) and the U.S. National Lightning Detection Network (NLDN). Extended Abstracts, 13th Conf. on Aviation, Range and Aerospace Meteorology, Amer. Meteor. Soc., New Orleans, LA, P2.6. [Available online at https://ams.confex.com/ams/88Annual/techprogram/paper_130694.htm.]
Xu, Q., 1992: Formation and evolution of frontal rainbands and geostrophic potential vorticity anomalies. J. Atmos. Sci., 49, 629–648.
Zhang, D.-L., and H.-R. Cho, 1995: Three-dimensional simulation of frontal rainbands and conditional symmetric instability in the Eady-wave model. Tellus, 47A, 45–61.
Zipser, E. J., and K. R. Lutz, 1994: The vertical profile of radar reflectivity of convective cells: A strong indicator of storm intensity and lightning probability? Mon. Wea. Rev., 122, 1751–1759.
