1. Introduction and background
Mesoscale convective systems (MCS)s are common during the summer months in central North America and are significant for their rainfall and severe weather. An MCS is a cloud system that occurs in connection with an ensemble of thunderstorms and produces a contiguous precipitation area, which has a horizontal extent in at least one direction of approximately 100 km or more (Glickman 2000). Although numerous studies have noted large variations in organizational structure, MCSs can generally be classified using radar reflectivity patterns as having either a linear or nonlinear convective band. Since most MCSs generate a stratiform precipitation region attendant to the main convective band, linear MCSs can be classified into subcategories based upon the orientation of the stratiform precipitation region with respect to the convective band: trailing stratiform, leading stratiform, and parallel stratiform (Parker and Johnson 2000). Parker and Johnson (2000) noted that 58% of all MCSs observed over the central United States in May 1996 and May 1997 were classified as trailing stratiform, making it the most frequently occurring type. Leading stratiform and parallel stratiform–type MCSs were each found to comprise less than 20% of all MCSs.
Parker and Johnson (2000) classify an MCS as parallel stratiform if most of the stratiform precipitation region moves parallel to the convective line and to the left of the line’s motion vector. Parallel stratiform MCSs generally have large reflectivity gradients that exist on both the front and rear sides of the convective line. Parker (2007a) observed that lower-tropospheric hydrometeor advection and outflow expansion is to the right of the convective line (relative to the storm motion vector), which produces back-building and generates new convection. In parallel stratiform MCSs, the front-to-rear inflow is lofted in the convective region, and upper-tropospheric storm-relative hydrometeor advection is toward the left of the convective line (relative to the storm motion vector), which produces a line-parallel stratiform precipitation region (Parker 2007a). In general, parallel stratiform MCSs have storm-relative, line-perpendicular winds that are modest at upper levels, weak in the middle-troposphere, and strong in the low levels (particularly in comparison with trailing stratiform MCSs; Parker and Johnson 2000).
Parallel stratiform MCSs have convective regions that produce outflow mostly close to the proximity of the surface outflow boundary, which leads to the rapid development of boundary layer cold pools (Parker 2007a). Strengthening and deepening cold pools provide greater lift along the main north–south outflow boundary. Halligan and Parker (2004) observed that a parallel stratiform MCS was unlikely to retain its defining characteristics because the low-level line-perpendicular shear was not sufficient to balance the strengthening cold pool. Parker and Johnson (2000) found that 58% of all parallel stratiform MCSs eventually evolved into a trailing stratiform MCS, with parallel stratiform MCSs in deep, strong line-parallel shear environments being able to retain their structure the longest before transitioning into a trailing stratiform MCS.
While parallel stratiform MCSs occur relatively infrequently, they are significant producers of severe weather. Gallus et al. (2008) reported that 80% of parallel stratiform MCSs produced at least one severe weather report, not including flooding reports. Parallel stratiform MCSs produced the highest frequency of very large hail [diameter of 2 in. (∼5 cm) or greater] and tornadoes of the nine convective modes studied. Duda and Gallus (2010) also showed that parallel stratiform MCSs have the highest frequency of tornado occurrence of all the linear MCSs. There is also a high frequency of flooding produced by parallel stratiform MCSs (Gallus et al. 2008; Duda and Gallus 2010; Schumacher and Johnson 2005). Slow-moving or stationary parallel stratiform MCSs can produce significant flooding owing to the regenerating convective cell motion is parallel to the convective line orientation, resulting in “training” cell movement over the same locations (Parker and Johnson 2000; Schumacher and Johnson 2005; Oue et al. 2014). Finally, parallel stratiform MCSs occur in environments with significant deep-layer shear and clockwise-turning hodographs, which are also known to favor the development of supercell thunderstorms (Moller et al. 1994; Parker and Johnson 2000; Parker et al. 2001). From a warning standpoint, the ability to distinguish between parallel stratiform MCS environments from supercell environments is crucial because Duda and Gallus (2010) showed that supercells produce significantly more severe weather of any type than any other convective mode, including parallel stratiform MCSs.
Despite their prevalence and severity, Parker (2007a) noted a lack of research on parallel stratiform MCSs. In addition to the seven studies mentioned in the previous subsection, Parker (2007a) highlighted five other studies that appear to document parallel stratiform MCSs but did not label them as such: Chappell (1986), Schwartz et al. (1990), Moore and Gagan (2000), Rogash and Smith (2000), and Roebber and Eise (2001). A few studies of parallel stratiform MCSs and parallel stratiform MCS-like systems have been conducted using numerical model simulations (Parker 2007a,b; Liu and Moncrieff 2017; Clark et al. 2014a,b). Only a few parallel stratiform MCSs have been observed with airborne platforms. One of these was during the Bow Echo and Mesoscale Convective Vortex Experiment (BAMEX) in the summer of 2003 (Storm et al. 2007; Davis et al. 2004; Halligan and Parker 2004) and another was a hybrid parallel–trailing stratiform MCS observed on 11 May 2011 during the Midlatitude Continental Convective Clouds Experiment (MC3E; Jensen et al. 2016). A few more parallel stratiform MCSs were observed during the Plains Elevated Convection At Night experiment in the summer of 2015 (Cui et al. 2019; Geerts, et al. 2017). The purpose herein is to study the 11 May 2011 MCS with a focus on kinematic and microphysical properties. The kinematic analysis compares observed and derived wind fields of the parallel stratiform MCS to determine the important properties of the wind field and surface cold pool. The microphysical analysis uses gamma distributions fit to the observed spectra to compare with properties of trailing stratiform MCSs observed during the BAMEX field project as described in McFarquhar et al. (2007).
2. Data and methodology
During MC3E, a wide variety of ground-based and airborne instruments were used to collect data for convective storms and their environments (Jensen et al. 2016). The instruments included seven ground-based, multi-frequency, cloud and precipitation dual-polarimetric radars, the University of North Dakota (UND) Cessna Citation II Research Aircraft (Delene et al. 2019), the Oklahoma Mesonet surface observing network, and a six-station rawinsonde network that launched four to eight soundings per day (Jensen et al. 2015).
Data from three X-Band Department of Energy Atmospheric Radiation Measurement (ARM) program radars and the Vance Air Force Base, Oklahoma National Weather Service (NWS) Weather Surveillance Radar-1988 Doppler (WSR-88D) are used to retrieve a three-dimensional wind field for a 100-km2 domain centered around the ARM Southern Great Plains (SGP) Central Facility (location shown in Fig. 1). Artifact removal and radar data quality control procedures have been applied (North et al. 2017). The multi-Doppler retrieval is performed using a three-dimensional variational analysis approach (North et al. 2017), and the retrieved three-dimensional wind components are mapped onto a 100-km2 Cartesian grid with a horizontal spacing of 500 m and a vertical spacing of 250 m. The velocity retrieved from the multi-Doppler data for this case have an uncertainty of approximately 1 m s−1 below 8-km altitude (North et al. 2017) and are only valid in regions of clouds and precipitation. The vertical velocity uncertainty was determined by using wind profiler updraft velocity retrievals (Giangrande et al. 2013) around the SGP Central Facility, which is located approximately 11 km west from the nearest radar.
Base radar reflectivity mosaic from the Dallas–Ft. Worth, TX; Oklahoma City, OK; Vance Air Force Base, OK; Tulsa, OK; and Wichita, KS, National Weather Service WSR-88Ds at (a) 1600, (b) 1700, (c) 1800, (d) 1900, and (e) 2000 UTC 11 May 2011. Each radar image is approximately 5 min in duration. The yellow star denotes the position of the Atmospheric Radiation Measurement Program Central Facility. Thin orange lines in (b) and (c) denote the first three and last three flight legs, respectively, of the University of North Dakota Citation II weather research aircraft flight track, defined in Table 1 and shown in Fig. 2. The black box denotes the domain of multi-Doppler wind retrieval.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
The NASA S-Band Transportable Dual-Polarimetric Radar (NPOL) was deployed in MC3E to augment the array of ARM research radars (Jensen et al. 2016). From 1700 to 1815 UTC, NPOL was used to gather range–height indicator (RHI) data from the 270° azimuth, generating vertical reflectivity and radial velocity cross sections through the parallel stratiform MCS. The 270° azimuth RHI radial velocities can be approximated as the east–west component of the wind. RHI data are plotted using the Py-ART software package (Helmus and Collis 2016).
In situ measurements of the stratiform precipitation region were obtained with the Citation Research Aircraft. Six horizontal legs were flown (Figs. 1b–d and 2), stepping down from approximately 7500 to 2800 m above mean sea level (MSL) (Table 1). Aircraft measurements of temperature, humidity, pressure, and position data are processed using the Airborne Data Processing and Analysis (ADPAA) software package (Delene 2011). Greater than 1-Hz measurements are averaged to 1 Hz. Three-dimensional winds are calculated from measurements collected with a gust probe in combination with an Applanix Position and Orientation System following the method described in Lenschow (1986).
Detailed map showing the University of North Dakota Citation II weather research aircraft flight track (thin gray line and colored lines) over north-central Oklahoma on 11 May 2011. The legend gives the leg numbers defined in Table 1. The thick dark blue line denotes the Kansas–Oklahoma state border, thin dark blue lines denote county borders, the thick black box denotes the multi-Doppler analysis domain, and the yellow star denotes the Atmospheric Radiation Measurement Program Central Facility location.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
The start time, end time, height, and temperature for flight legs of the University of North Dakota Citation II weather research aircraft. Height and temperature data are given as mean values plus or minus one standard deviation. The aircraft’s height is GPS altitude above mean sea level (MSL).
A Sky Tech Research Nevzorov Probe (Korolev et al. 1998) was used to measure liquid water content (LWC) and total water content (TWC). A “deep cone” model of the Nevzorov probe is used to reduce particle bounce and increase collection efficiency compared with earlier models (Korolev et al. 2013). It is noted that the heated wire in the collection cone can be overwhelmed in high water content conditions; therefore, while the stated range of TWC and LWC values is 0.003–3 g m−3, the maximum observed TWC values in MC3E are approximately 1.5 g m−3. It is not known if this low TWC is the maximum water content in the observed clouds or if the Nevzorov probe became saturated at 1.5 g m−3.
The Citation Research Aircraft carried a Particle Measuring Systems Inc. Two-Dimensional Cloud probe (2DC), a Droplet Measurement Technologies Cloud Imaging Probe (CIP), and a SPEC Inc. High Volume Precipitation Spectrometer version 3 (HVPS-3). The list of probe specifications is provided in Table 2. The National Center for Atmospheric Research processed the hydrometeor image data using the System for Optical Array Probe (OAP) Data Analysis version 2 (Bansemer 2019) software to obtain particle size distributions using the maximum particle length in any direction or reconstructed particle diameter (D) based upon the Heymsfield and Parrish (1978) reconstruction technique. A combined particle size distribution is created using the CIP and 2DC data for particles larger than 75 μm and smaller than 1000 μm, and HVPS-3 data for particles larger than 1000 μm and smaller than 3 cm.
List giving the specifications for the Two-Dimensional Cloud Probe (2DC), Cloud Imaging Probe (CIP), and the High-Volume Precipitation Spectrometer (HVPS-3). The diode size is the effective size of image pixels shadowed. The diodes column lists the total number of diodes in the probe’s array. The image width column lists the maximum image size along the diode array. The A.-S. tips column lists if anti-shattering tips are on the probe.
The 2DC 1-s average concentration uncertainty is approximately 50% and the particle size uncertainty is approximately 25% (Gayet et al. 2004). Artificially high concentrations of small hydrometeors are possible when larger hydrometeors impact the probe tips and shatter into smaller artifacts, which are then sampled with the probes (Gayet et al. 2004; Korolev et al. 2011). Specially designed probe tips on the 2DC and HVPS-3 help reduce artifact uncertainty by redirecting shattered hydrometeors away from the sampling area. Artifact rejection algorithms in the processing software help reduce this uncertainty, notably for times when CIP data are used rather than 2DC data (Korolev and Field 2015). The newer CIP and HVPS-3 probes used in this project contain updated photodetectors and electronics that reduce the particle undersizing errors arising from the response time of photodiodes (Baumgardner et al. 2017; Baumgardner and Korolev 1997). The HVPS-3 also benefits from a wider photodetector array, which allows for larger particles and particle concentrations to be fully imaged without the use of particle reconstruction techniques, reducing the uncertainties associated with particle size. All three probes suffer from depth of field (DOF) sizing issues for small, out-of-focus particles, but this issue is mitigated by using only particle sizes big enough that the DOF encompasses the entire distance between the probe length (Baumgardner et al. 2017). To avoid DOF errors, a particle lower-bound threshold of 75 and 1000 μm is used for the 2DC–CIP and HVPS-3 probes in the combined spectra, respectively. Given these considerations, the CIP data likely have uncertainties similar to the 2DC; however, the HVPS-3 may have similar or smaller uncertainties than the 2DC (A. Bansemer 2020, personal communication). While the CIP was the preferred instrument used to create the combined 2DC–CIP–HVPS-3 spectra, the 2DC is used 17% of the time, mostly for spectra during the two highest flight altitudes, due to intermittent data capture issues with the CIP probe. The 1-Hz combined number distributions are averaged into 10-s segments, which represents an average horizontal flying distance of 1.2 km. Spectra that coincide with an average Nevzorov TWC greater than 0.01 g m−3 are considered to be in cloud.
The Oklahoma Mesonet is a network of automated surface stations covering the state of Oklahoma that record observations every 5 min (McPherson et al. 2007; Brock et al. 1995). The network consists of 120 stations, with at least one station in every county in Oklahoma. Observed variables include air temperature, wind speed and direction, precipitation, and relative humidity. Mesonet temperature and station pressure data are used to calculate the potential temperature at a 1000-hPa surface to reduce the effect of terrain upon the analysis. Surface meteorological data from the Oklahoma Mesonet network are utilized to determine cold pool evolution.
3. Results and discussion
An MCS with trailing stratiform characteristics moved through northeastern New Mexico and southeastern Colorado in the early morning hours in local time. By 1300 UTC the southern part of the MCS moved into southwestern Oklahoma and from 1300 to 1600 the convective region lost its strength and both the convective region and the stratiform region degenerated (not shown). Between 1500 and 1600 UTC there was non-organized convective redevelopment on the southeastern side of the storm. Figure 1a shows the MCS became more organized between 1600 and 1700 UTC, with the convection on the south side of the storm with the stratiform region to the north in the parallel stratiform mode. By 1700 UTC a convective line about 70 km long had formed along the south and up into the western side of the MCS with another area of convection developing along the eastern side of the stratiform region along the leading edge of the cold pool. By MCS standards, this storm was on the small side by being only 120–150 km long from southern to northernmost extent and only 60–70 km wide at its widest point. The convection along the western side of the MCS had dissipated by 1800 UTC and the convection on the east side of the stratiform region strengthened and became connected with the convection on the MCS’s southern side (Fig. 1d). Around 1840 UTC, the MCS transitioned to a hybrid parallel–trailing stratiform MCS. After 1900 UTC, the convective line began to erode. At 2000 UTC, the remnants of the MCS, the convection to the southwest of the parallel stratiform MCS, and the stronger convection advecting northeastward out of Texas merged into one convective complex (Fig. 1f).
The average MCS motion is 11 m s−1 from 231° between 1600 and 2100 UTC while the average motion of individual convective cells is 22 m s−1 from 202°. Both the MCS aggregate motion and the individual cell motions are determined using the Gibson Ridge Level II Analyst radar-viewing application (GR2Analyst), version 2.92 (http://www.grlevelx.com). The MCS system motion is found by using GR2Analyst to map out vertices of a rough polygon of the MCS shape every hour, at the top of the hour for consistency. These vertices are averaged to determine an approximate centroid position of the MCS. From the centroid positions of the MCS an average system motion is computed. For the individual cell motions, a plethora of convective cells within the MCS are tracked from between 20 and 60 min, depending on cell longevity, their motion vectors recorded and averaged over the 1600–2000 UTC timeframe. Figure 3 shows an example of a convective cell moving northward from the convective region into the stratiform region and into the Citation’s flight path from 1722 to 1744 UTC. Some of these cells were brief, lasting less than 30 min before they dissipated into the stratiform region.
Radar reflectivity from the Vance Air Force Base, Oklahoma, National Weather Service WSR-88D at 1.3° elevation tilt at (a) 1722, (b) 1727, (c) 1731, (d) 1735, (e) 1740, and (f) 1744 UTC. University of North Dakota Citation II weather research aircraft position at time of plot is given with a black circle and the aircraft track for the previous 5 min and next 1 min are shown in the brown line. Black circles denote a convective cell of interest.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
Figure 4a shows the 1700 UTC proximity sounding from the ARM Central Facility (SGP), which was launched approximately 70 min before the parallel stratiform MCS moved over the ARM Central Facility (yellow star in Fig. 1), and plotted using the SHARPpy software package (Blumberg et al. 2017). Several small inversions are present from the surface to just above 800 hPa, with near-dry adiabatic lapse rates between and above these inversions. The sounding contains over 1900 J kg−1 of surface-based convective available potential energy and close to 0 J kg−1 of surface-based convective inhibition, indicating that the thermodynamic environment over northern Oklahoma is conducive to maintaining the ongoing convection.
(a) Atmospheric sounding launched from SGP at 1700 UTC 11 May 2011. Wind speed is in meters per second with a full barb equal to 5 m s−1. Only 1 out of every 60 wind barbs are shown for clarity. Derived surface-based thermodynamic indices include convective potential available energy (CAPE), convective inhibition (CIN), lifted condensation level (LCL), level of free convection (LFC), level of neutral buoyancy (LNC), and the lifted index (LI). (b) Hodograph from the 1700 UTC SGP sounding with the data color coded based on altitude (darker colors closer to the ground) for clarity and altitude in kilometers marked with black dots and numerical labels for reference.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
a. Kinematic properties of the MCS
The use of multiple in situ and remote sensing platforms to observe the 11 May 2011 MCS enables an in-depth analysis of both how the internal wind field changed during the transition from parallel stratiform to trailing stratiform mode and the role the cold pool played in that transition. The data from the various platforms are not all continuously available between 1600 and 2100 UTC (Table 3) while the MCS is in the observational area; therefore, this analysis attempts to utilize the strengths of each dataset while minimizing the uncertainties that are inherently present.
Analysis times when data from the University of North Dakota Citation II weather research aircraft (Citation), NASA’s NPOL Doppler radar (NPOL), multi-Doppler winds analyses (multi-doppler), and the Oklahoma Mesonet (Mesonet) are available as denoted by the × symbol.
1) Anvil region
Between 1640 and 1706 UTC, the Citation Research Aircraft observations indicate that in-cloud winds (7.5 km MSL) are from the south and south-southwest between 20 and 30 m s−1 (Fig. 5). The aircraft data are consistent with the 1700 UTC SGP sounding (Fig. 4). The anvil-level wind direction and wind speed show minimal dependence relative to position within and near the parallel stratiform MCS; the anvil-level wind velocities reflect the large-scale, upper-level synoptic flow (not shown). The multi-Doppler wind retrieval indicates the anvil region south-southwesterly ground-relative flow (not shown) is persistent between 1800 and 2036 UTC (the ARM radar observation period). The MCS at 1700 UTC is likely parallel stratiform in nature, but as time goes on, the MCS becomes a hybrid parallel–trailing stratiform MCS and it gains trailing stratiform characteristics, but still retains the strong southerly storm-relative flow that gives it the parallel stratiform traits.
Wind direction and speed from the University of North Dakota Citation II weather research aircraft on 11 May 2011 as a function of altitude and distance from NPOL. The times of level flight segment midpoints are shown to the right of the data for reference. The 1700 UTC SGP rawinsonde winds from altitudes between 2 and 8 km are included along the right-hand side. Full wind barb is equal to 5 m s−1 and flag is equal to 25 m s−1. Every 30th aircraft wind measurement and every 20th sounding measurement are shown for clarity. Estimated maximum westward and eastern extent of the MCS at the time of the flight leg is given by the vertical black lines.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
The storm-relative wind field is computed by taking the vector difference of the ground-relative flow and the average MCS system motion from 1600 to 2100 UTC (11 m s−1 from 231°). Prior to about 1830 UTC, the storm-relative flow is mostly parallel to the convection at the leading edge of the stratiform region. Figure 6 shows that at 1655 and 1808 UTC there is not any continuous rear-to-front flow through the MCS and flow pattern is mostly parallel stratiform in nature. [Figures 8a and 9a show vertical cross sections of the MCS along the line “A” and “B,” respectively (lines shown in Fig. 7).] At 1821 UTC where the main updraft branch from along the eastern convection is fairly upright, disorganized, and only extends 15–30 km rearward of the leading edge of the precipitation. The updraft branch from the southern edge of the convection does not extend through the stratiform region south–north either, but the southerly flow is established throughout the analysis domain (Fig. 9a).
NASA S-band transportable dual-polarimetric radar (NPOL) range–height indicator (RHI) scans showing ground-relative radial velocity (in color, m s−1) and storm-relative radial velocity (arrows, m s−1) along the 270° radial (due west) for (a) 1655 and (b) 1808 UTC 11 May 2011. Plot orientation is such that east is to the right and north is into the image.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
The 8 km AGL multi-Doppler retrieval for radar reflectivity (color) and storm-relative horizontal winds (wind barbs) at (a) 1842, (b) 1918, and (c) 2000 UTC 11 May 2011. Full wind barb equals 5 m s−1. The domain of the multi-Doppler retrieval is shown in Fig. 1, the bold line “A” denotes the location of the cross section shown in Figs. 8 and 11, and the bold line “B” denotes the location of the cross section shown in Fig. 9.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
Based on the anvil region storm-relative flow, the storm could be considered a hybrid parallel stratiform–trailing stratiform–type MCS after 1834 UTC. By 1842 UTC, the main ascending updraft branch extends continuously upward and rearwards through the stratiform anvil region to the rear of the precipitation shield (Fig. 11a). At this time, the storm-relative flow in the anvil is 15–20 m s−1, predominately from the south and southeast (Fig. 7a), likely contributing to the hybrid parallel–trailing stratiform region to the north and west through hydrometeor advection. Near the convective cores there is a diffluent signature in the wind field, probably caused by convective upwelling. Interestingly, there is some southwesterly storm-relative flow to the northeast of the convective cores, which may have led to increased anvil coverage ahead (to the east) of the MCS.
Radar observations show that by 1900 UTC the parallel stratiform region has diminished greatly in size (Fig. 1d); however, there is only a modest increase in size of the trailing stratiform region. Figure 8b shows the front-to-rear flow through the MCS, although the flow no longer continuously ascends through the stratiform region. The horizontal multi-Doppler wind field at 1918 UTC (Fig. 7b) shows that the anvil region storm-relative flow has backed slightly more to the south-southeast, so any hydrometeor advection contributes equally to a parallel stratiform region as to a trailing stratiform region. This is likely why the parallel stratiform region diminishes and the trailing stratiform region slowly expands between 1900 and 2000 UTC (Figs. 1d,e). The strong southerly flow present throughout the observation period suggests that the MCS retains a parallel-stratiform flow component despite the transition toward the trailing stratiform mode (Figs. 7, 9b).
Vertical cross section along line “A” in Fig. 7 of w component of the storm-relative wind (color), radar reflectivity (line contours), and u–w components of the storm-relative wind (vectors) for (a) 1821 and (b) 1900 UTC from the multi-Doppler retrieval. Representative wind speed vectors are given in the top-right-hand corner of the figure.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
Vertical cross section along line “B” in Fig. 7 of w component of the storm-relative wind (color), radar reflectivity (line contours), and υ–w components of the storm-relative wind (vectors) for (a) 1821 and (b) 2000 UTC from the multi-Doppler retrieval. Representative wind speed vectors are given in the top-right-hand corner of each panel.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
2) Mid-cloud level
Below the anvil region, between 1731 and 1803 UTC, the two aircraft flight segments at 4.7 and 3.7 km MSL contained southwesterly ground-relative winds at the rear of parallel stratiform MCS, which transitioned to southerly and then southeasterly from the rear to the front of the MCS (Fig. 5). The change in wind direction between the eastern (leading) portion of the MCS and the western (trailing) portion is also observed within the NPOL radial velocity data. At 1655 UTC, there is an enhanced region of inbound wind components (from the west and heading east) between 4 and 8 km in altitude at 120–150 km (Fig. 6a). Farther to the east, winds away from NPOL (heading west) are observed, which is consistent with a wind direction change from southwest to southeast.
From the UND Citation Research Aircraft data, wind speeds at the rear of the MCS exceeded 30 m s−1, which are greater than the 11 m s−1 mean translational speed of the MCS in approximately the same direction, which means this feature is an area of rear inflow. Regions of rear inflow have been observed in the MCS of Clark et al. (2014a) and were a feature of a simulation of the same MCS (Clark et al. 2014b), but have not been shown or observed within other studies of parallel stratiform MCS, including the in-depth modeling studies of Parker (2007a,b). However, rear inflow is a common feature with trailing stratiform MCSs (Houze et al. 1989). The rear inflow may be caused by midlevel convergence, which accompanies all regions of stratiform precipitation (e.g., Houze et al. 1989).
NPOL data at 1808 UTC (Fig. 6b) and multi-Doppler analysis data (Figs. 10 and 11) show that as the MCS transitioned from parallel to trailing stratiform mode, the rear storm-relative inflow is a semi-persistent feature between 3 and 6 km AGL at the rear edge of the parallel stratiform region of the MCS. The southwesterly storm-relative inflow is flowing at an angle into the stratiform region and did not follow directly behind the convective region relative to the MCS’s path at 1842 UTC (Fig. 10a). The rear-inflow region is quite limited in areal extent, being approximately 60 km long and 30 km wide with an orientation of northwest to southeast. The area of rear-inflow moved north out of the analysis domain with the parallel stratiform region by 1918 UTC (Fig. 10b) and at this time, the storm-relative flow is south-southwesterly with only a very minor component of the wind moving from rear-to-front (Fig. 11b). By approximately 1940 UTC, another region of southwesterly rear-inflow started to develop on the rear side of the stratiform region to the southwest of the convective line (not shown). The rear-inflow increased in speed through 2000 UTC (Figs. 10c and 11c) and followed the trailing stratiform region as it exited the analysis domain by 2036 UTC.
The 4 km AGL multi-Doppler retrieval for radar reflectivity (color) and storm-relative horizontal winds (wind barbs) at (a) 1842, (b) 1918, and (c) 2000 UTC 11 May 2011 for 4 km AGL. Full wind barb equals 5 m s−1. Black dashed ovals highlight areas of rear inflow into the mesoscale convective system (MCS). The multi-Doppler retrieval domain is shown in Fig. 1, the bold line “A” denotes the location of the cross section shown in Figs. 8 and 11, and the bold line “B” denotes the location of the cross section shown in Fig. 9.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
Vertical cross section along line “A” in Fig. 7 of radar reflectivity (color), υ component of the storm-relative wind (contours), and u–w components of the storm-relative wind (vectors) for (a) 1842, (b) 1918, and (c) 2000 UTC 11 May 2011 from the multi-Doppler retrieval. Representative wind speed vectors are given in the top-right-hand corner of each panel. Red dashed ovals highlight areas of rear inflow into the mesoscale convective system (MCS).
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
As in the anvil region of the MCS, persistent line-parallel flow was present in nonconvective regions (Figs. 9 and 10). Despite the attention given to the area of rear-inflow, the MCS retained some parallel-stratiform characteristics through 2000 UTC.
3) Sub-cloud level
At 1600 UTC, the parallel stratiform MCS was collocated with a cold pool in western Oklahoma that developed from a previous convective system (Fig. 12). From 1600 UTC and onward through 2100 UTC the cold pool associated with the MCS expanded northeastward as it drifted into central Oklahoma. After 1800 UTC, the cold pool began to merge with the cold pool associated with convection advecting northeastward from northern Texas. This was a strong cold pool with a potential temperature perturbation of 6°–8°C compared to the pre-MCS environment at 1600 UTC (Fig. 12). The potential temperature perturbation increased slightly throughout the day to 8°–10°C. Figure 12 shows that the minimum temperature of the cold pool remained rather steady, but the increase of the temperature perturbation is likely the result of radiational heating of the pre-MCS environment.
Surface potential temperature (filled color contours) and ground-relative surface wind over Oklahoma from 1600 to 2100 UTC as measured by the Oklahoma Mesonet. Full wind barb denotes 5 m s−1 and half barb equals 2.5 m s−1. The star denotes location of Atmospheric Radiation Measurement Program Central Facility and the black box shows the domain of multi-Doppler wind retrieval.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
The outflow associated with the cold pool plays a role in the transition from the parallel stratiform mode to the trailing stratiform mode (Parker 2007a). Parker (2007a) found that the cold pools in parallel stratiform MCSs develop near the convective cells, and expand to the right and rearward in a storm-relative sense. The rearward expansion of the cold pool is seen in Fig. 12 with the expanding breadth of the cold pool and the easterly surface outflow to the south and west of the MCS. Rotunno et al. (1988) showed that the cold pool produces rearward accelerations of air parcels that are lifted by the cold pool, which suggests that the cold pool influences the MCS transition to the trailing stratiform mode, in line with the conclusions of Parker (2007a). This may be what caused the line of convection on the eastern side of the stratiform region to develop around 1700 UTC (Fig. 1b). Figure 12 shows there is some surface convergence along the east side of the cold pool, denoting the boundary of the gust front. This particular gust front did not produce strong winds along this leading edge, like that seen at 2100 UTC in southern Oklahoma from the MCS to the south (Fig. 12).
The depth of the cold pool is estimated to be approximately 2 km based on the depth of the low-level rear-to-front flow in Fig. 11. The control volume used in this exercise extends along line “A” in Fig. 7 from beyond the east (right-hand) side of the multi-Doppler analysis westward to the intersection of lines “A” and “B” in Fig. 7 from the surface to 2.5 km AGL. The values of
Analysis times, the square of the u component of the wind at an altitude of 0 km AGL, the square of the u component of the wind at 2.5 km AGL, and the cold pool strength, all in meters per second, at the location where line “A” meets line “B” in Fig. 7.
The calculated cold pool strength, generally between 6 and 10 m s−1, shows a minor time dependence (Table 4). This time dependence seems to be due to a similar trend in the
The 0 km AGL multi-Doppler retrieval, as described by North et al. (2017), for radar reflectivity (color) and ground-relative horizontal winds (wind barbs) at (a) 1842, (b) 1900, and (c) 2000 UTC 11 May 2011. Full wind barb equals 5 m s−1. The multi-Doppler retrieval domain is shown in Fig. 1, the bold line “A” denotes the location of the cross section shown in Figs. 8 and 11, and the bold line “B” denotes the location of the cross section shown in Fig. 9.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
Overall, the small range of calculated cold pool strength appears to signify that the cold pool did not strengthen appreciably during the analysis time. Based on the potential temperature data shown in Fig. 12, the cold pool was relatively stable. The minimum potential temperature of the cold pool did not decrease through the analysis time, nor did the cold pool expand significantly prior to 2000 UTC.
b. Microphysical properties of the MCS
A second goal of this study is to compare microphysical properties in the rarely studied parallel stratiform MCS to those in the more common trailing stratiform MCS. The oft-cited Parker (2007a,b) used a numerical simulation with a microphysics parameterization based on Lin et al. (1983) and Tao and Simpson (1993) to study parallel stratiform microphysical processes while other studies (e.g., Cui et al. 2019) rely on remote sensing techniques.
The McFarquhar et al. (2007, hereafter M07) study is used to compare parallel stratiform (herein) and trailing stratiform (M07) characteristics. M07 analyzed 17 in situ spiral descents from BAMEX in order to further the role of microphysics in MCS and bow echo evolution. This study aims to show that microphysical processes are similar between this parallel stratiform MCS and trailing stratiform MCSs. The descents in M07 were conducted in the trailing stratiform region of MCSs where the temperatures were between −10° and +10°C. Bulk microphysical properties are examined along with particle size distributions.
1) Parallel stratiform region bulk microphysics
Figures 14 through 16 show the environmental relative humidity characteristics, total water content, and total hydrometeor concentrations of particles larger than 100 μm, respectively, of the parallel stratiform region as derived from Citation Research Aircraft data collected during the six level flight segments (Table 1). The aircraft flight pattern on 11 May 2011 did not contain a Lagrangian flight profile, thus advection may play a role in the differences between flight legs. The Citation Research Aircraft sampled the stratiform region starting at an altitude of 7.5 km before convection on the eastern side of the stratiform region developed. The aircraft finished sampling the stratiform region at an altitude of 2.8 km about 30 min prior to the front-to-rear flow becoming established through the stratiform region.
While sampling the stratiform region, the Citation Research Aircraft encountered regions of near-saturated and supersaturated air. Figure 14 shows that for most of the first (top) leg, the relative humidity with respect to ice (RHi) at 7.5 km is between 95% and 120%, which are similar to conditions found in cirrus (Ström et al. 1997; Garrett et al. 2004; Gayet et al. 2004; Jensen et al. 2005) and in thunderstorm anvils near convective cores (Heymsfield et al. 2005). In the western half of the MCS at 3.7 and 4.7 km, RHi is primarily between 90% and 110% (Fig. 14). Figure 3 shows that near the western edge of the MCS there were convective elements in the vicinity of the Citation flight path during the beginning of the fourth flight leg (4.7 km). These convective elements had disintegrated and merged into the stratiform region by the time the Citation was flown back into the area at 3.7 km during the fifth flight leg (not shown). The remnants of this convection explain the saturated air found during the fifth flight leg.
Relative humidity (colors) with respect to ice when T < 0°C and with respect to liquid water when T ≥ 0°C from the University of North Dakota Citation II weather research aircraft during the six level legs (Table 1) when the Nevzorov total water content was greater than 0.01 g m−3 on 11 May 2011. Estimated maximum westward extent of the mesoscale convective system (MCS) anvil at the time of the flight leg is given by the vertical black lines. Average temperature for each flight leg is given to the right of the respective flight leg.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
The Citation research aircraft also encountered regions of sub-saturated air within the parallel stratiform region. The second flight leg at 6.6 km and the third flight leg at 5.6 km both show broad areas of moderately sub-saturated air, with RHi in the 85%–95% region, which is dry enough to cause sublimation of cloud and precipitating ice particles (Fig. 14). Figure 5 shows that the sampling area for the second and third flight legs was confined to the western part of the MCS, where the winds were from the southwest and part of the rear inflow region. In the eastern half of the MCS at 3.7 and 4.7 km, RHi ranges from 50% to 90% (Fig. 14). The low relative humidity air is probably due to entrainment of pre-MCS environmental air into the eastern side of the MCS due to the approximately 20 m s−1 storm-relative front-to-rear flow (Fig. 6). At 2.8 km (sixth flight leg), the very sub-saturated air shown in Fig. 14 coupled with the raindrop sized particles on the OAP imagery suggests that the Citation was below cloud base with precipitation falling into the sub-saturated boundary layer.
Moreover, both TWC and total hydrometeor number concentration (Nt) show a trend of decreasing with decreasing height (increasing temperature; Figs. 14 and 15), which is consistent with the results of M07 for trailing stratiform MCSs1. M07 found that, averaged over all spirals, Nt decreased at a rate of 19% ± 10% °C−1 (i.e., Nt decreased with increasing temperature) and TWC decreased by 10% ± 7% °C−1. In M07, the higher rate of decrease in the total number concentration than TWC was thought to be a result of aggregation and sublimation occurring in the stratiform region. In this study, the median TWC and Nt (Figs. 14 and 15) values for each flight leg are used to compute the vertical rate of decrease. For flight legs at altitudes with temperatures greater than −10°C, Nt decreases at a rate of 20% ± 8% °C−1 and the TWC decreases at a rate of 9% ± 2% °C−1. Like in M07, the Nt decreases at a faster rate than the TWC, which is indicative of aggregation and sublimation. Both processes are likely in play, given the subsaturated conditions that generally occurred for the flight legs that are at air temperatures greater than −10°C (Fig. 14) and the large, lumpy, irregular shaped particle images in the OAP data (Fig. 17).
Box-and-whisker plot showing variation of total water content as a function of temperature between 1640:00 and 1811:40 UTC 11 May 2011. The flight leg segments are defined in Table 1. The top and bottom of a box represent the 75th and 25th percentiles, respectively; the middle line represents the median; the bottom and top whisker represents the smallest and largest data point within 1.5 times the interquartile range (IQR), respectively; and the small circles represents data outside 1.5 times the IQR. The fifth leg was divided according to when the temperature was <0°C (leg 5a) and when the temperature was ≥0°C (leg 5b).
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
Icing detector data (not shown) indicate that supercooled liquid drops were present during the western half of the fourth flight leg (air temperature around −6°C; Table 1), with liquid water contents measured with the Nevzorov probe as high as 0.28 g m−3. At the start of this leg, the Citation flew through an area of supercooled liquid water (Nevzorov liquid water concentrations peaking over 0.1 g m−3) for about 2 min and then another area of higher concentrations (0.05–0.28 g m−3) for up to a minute, followed by decreasing liquid water concentrations for the next 3 min. Unfortunately, the vertical wind retrieval is unavailable during this portion of the Citation dataset so it cannot be stated with utmost certainty whether the presence of the supercooled liquid water was due to upward advection or condensation. Given that the specific regions where the icing detector registered supercooled liquid water were small and do not correspond well with localized areas of supersaturated air, it seems likely that the nearby convection was bringing cloud droplets that formed below the melting level up into the subfreezing midlevel air.
2) Unimodal particle spectra
The 10-s-averaged spectra are fit to three types of distributions: the gamma function (3) fit over the entire spectrum (unimodal), a gamma function fit over the entire spectrum where the order of fit (μ) is set to zero (an exponential function), and a bimodal gamma distribution. For each new gamma distribution fit, the smallest χ2 value is saved. The fit with the smallest χ2 value among the three is used as the final fit for the spectrum.
Each spectrum is tested for bimodality based on the method described in Mace et al. (2002), where a spectrum is considered bimodal if the decrease in slope is more than 25% of the slope between the next highest and lowest bins. For bimodal spectra, two gamma functions are fit—one to each segment. The segments match along bin boundaries and only two modes are allowed. A valid gamma distribution fit occurs when there are four or more size bins present within a spectrum, the Levenberg–Marquardt technique can fit a distribution to the data, N0 is greater than zero but less than 104 cm−(3+μ) μm−1, μ is less than 104, and λ is less than 103 cm−1. The thresholds used to determine valid N0, μ, and λ are arbitrary, but a visual inspection of the distribution fits shows that of the spectra with four or more size bins and whose fit does not match the spectrum, one or more of the fit parameters exceed the limits given above. If both parts of the bimodal spectrum have a valid gamma fit, the χ2 values for each part are averaged to create a combined χ2. Otherwise, the χ2 value for the part of the distribution that does have a valid fit is used for the whole spectrum. An example of unimodal and bimodal spectra and their respective gamma distribution fits are shown in Fig. 18.
As in the measurement of particle concentrations, the fitting of gamma distributions to the particle spectra is subject to uncertainty. An anomalous concentration of a single size bin, whether high or low, could cause the algorithm to flag the spectrum as bimodal when a human observer would classify the spectra as unimodal. Or vice versa, a human observer could see a slight bump or dip in the particle spectra and classify it as bimodal, but if the change in slope was less than 25%, the fitting algorithm would consider it a unimodal spectrum. A visual inspection of the gamma distribution fits shows that the fitting algorithm works well overall, with perhaps a few misclassifications here and there. Similarly, by including the first couple of concentration bins, which have midpoint diameters of 75 and 125 μm, additional uncertainty may have been introduced to the gamma distribution fits. Several studies (e.g., Heymsfield 1985; Baumgardner and Korolev 1997; Strapp et al. 2001; Wu and McFarquhar 2016) have determined that a combination of sizing errors due to using only a few image pixels for particle size determination and limited depth of field for small particles leads to large uncertainties in the computation of number concentrations for particles less than 100–150 μm in diameter. The added uncertainty included by using the first couple of size bins should not detract from the results since a visual inspection of the gamma distribution fits shows good representation along the whole length of the spectrum and that the discussion below is focused on the trends within the gamma distribution parameters and not the individual values of the fit parameters.
The distribution fits range from four to 35 bins. Unimodal distributions fits are mostly longer than 10 bins, with only a couple less than that. For bimodal distribution fits, the length of the first piece (small-sized particles) ranges from four to 13 bins, with fits of 5 bins or less happening about one fifth of time. A length of 5 bins would cover particles in the size range of 75–275 μm, while a length of 13 bins would cover particles up to 850 μm in diameter. The length of the second piece (large-sized particles) of bimodal distributions are mostly greater than 10 bins with the majority of the fits that are less than 10 bins in length occurring during the fifth and sixth flight legs where sampling raindrops instead of large snow aggregates leads to a smaller range of particle sizes.
The unimodal distributions are compared to M07 microphysical characteristics of the trailing stratiform MCSs. M07 fit gamma distributions (3) to 60-s-averaged hydrometeor spectra2 during the 16 spiral descents obtained from both trailing stratiform and leading stratiform MCSs.
Figure 19 shows the variation of gamma-fit parameters for unimodal cloud particle spectra with respect to temperature. N0 and λ decrease with increasing temperature above the melting layer (Fig. 19), which is probably due to aggregation and sublimation of ice particles as they fall through the stratiform region. The decrease in λ indicates that there is an increasing fraction of large particles compared with the total distribution (broader spectrum). The decrease in N0 indicates that there are fewer total particles at lower altitudes than at higher altitudes. There is an increase in μ with increasing temperature, which also indicates a broadening of the cloud particle spectra. Evidence of aggregation at or above the melting layer is seen in HVPS-3 images collected during the first five level flight legs (a sample of which is shown in Fig. 17). M07 also showed a trend of decreasing λ with increasing temperature in trailing stratiform MCSs, concluding that the decrease in λ was caused by aggregation occurring in the stratiform regions of trailing stratiform MCSs. Sublimation also contributes to a decrease in N0 and λ due to the removal of small particles from the spectrum in regions of subsaturated air (Fig. 14).
M07 found that the slope parameter λ ranged between 0 and 20 cm−1, N0 ranged between 10−8 and 10−5 cm−(3+μ) μm−1, and μ ranged from −2 to 0 for 17 spirals in BAMEX stratiform regions. In this study, 35 distributions are classified as unimodal (Table 5). For all the distributions that are located above the melting layer, N0 ranges from 1.6 × 10−8 to 2.1 × 10−6 cm−(3+µ), µ from −1.9 to −0.4, and λ from 1.3 to 9.6 cm−1, which fall within the ranges of N0, μ, and λ given by M07. There are not enough samples below the melting layer to provide a reliable indication of the properties of the drop size distributions there. The consistency between these results and those of M07 is further evidence that microphysical characteristics in this MCS and those studied in BAMEX are similar.
Number of unimodal distributions and bimodal distributions during flight legs defined in Table 1. The “valid first piece” column lists the number of the bimodal distributions where the first (smaller-sized) piecewise function is valid, while the “valid second piece” column lists the number of the bimodal distributions where the second (larger-sized) piecewise function is valid. The fifth flight leg is broken into two sections for when the air temperature was less than 0°C (leg 5a) and greater than (leg 5b) 0°C.
3) Bimodal particle spectra
In the literature, bi or multimodal particle distributions have been observed in a wide variety of cloud phenomena, but primarily through use of surface distrometer, radar, or wind profiler data. Bimodal particle distributions can be generated through several different mechanisms. Mixed-phase clouds have been shown to produce such spectra through riming and/or secondary ice production (Zawadzki et al. 2001) and can be transient features (Radhakrishna and Rao 2009). Bimodality has been reported within the melting layer due to the coexistence of distinct raindrop and melting snowflake modes (Gossard et al. 1990) and is also an indication of hydrometeor aggregation (Field 2000). Such spectra have been reported in cirrus, with aggregation and diffusional growth being the two primary mechanisms causing bimodality (Mitchell et al. 1996). Diffusional growth contributes to bimodality in cirrus when excess water vapor cannot all be deposited upon the existing ice crystals and new particles are created; this mechanism is not applicable to this study.
M07 did not observe bimodality in the data; thus, they fit only a single gamma distribution to the entire spectrum. It is unknown why this study saw bimodality in the particle size distributions and M07 did not. Some possibilities that may fully or partially explain this variance include the difference in averaging periods (10 s in this study vs 60 s in M07), differences in the software used to create the particle size distributions (Wu and McFarquhar 2016), or perhaps differences in microphysical processes resulted in bimodal distributions in this study and unimodal in M07.
In this study, there are likely a few causes of particle spectra bimodality present. The first cause is from the presence of mixed-phase clouds, especially during the fourth flight leg that occurred in a region of supercooled liquid water. The bimodality could be picking up the presence of the smaller liquid drops (first mode) and the larger snowflakes falling through this layer (second mode). Particle growth through riming is probably occurring here. Another possible source of bimodality that may have been encountered during the fourth flight leg are small ice splinters caused by the Hallett-Mossop process (Hallett and Mossop 1974). However, a visual inspection of the 2DC and CIP images show hardly any columns or needles, making it doubtful that the Hallett–Mossop process was contributing to the bimodality in the spectra at that time. The third cause is from sampling the melting layer during the fifth flight leg as this leg sampled the top of the melting layer. The smaller mode would indicate the presence of melted snow along with larger snowflakes undergoing the melting process (second mode). The fourth mechanism, and the focus of the discussion below, is the indication of aggregation occurring above the melting layer (even within the mixed-phase layer of the fourth flight leg).
N01, μ1, and λ1 show very little dependence on temperature (Fig. 20). The median values of N01, μ1, and λ1 are given in Table 6. This may be due to the limited number of bimodal distributions that have a valid first piece (Table 5). Most of the invalid first pieces of the bimodal distributions either have too few data points (less than four) to produce a valid gamma distribution fit, or the Levenberg-Marquardt technique converged on a solution that results in one of the gamma distribution parameters exceeding the valid bounds of either N0, μ, or λ. About 56% of the spectra have too few size bins in the first bimodal piece for a gamma distribution to be fit. Due to uncertainties within the measurements, some of the spectra that have invalid first pieces may be unimodal distributions that have been misclassified, which may skew the statistics for the gamma distribution fit parameters for the second bimodal piece presented below.
Median values of the intercept parameter (N0), order of fit (μ), and slope parameter (λ) of gamma distribution fits to particle spectra observed during flight legs defined in Table 1.
N02 also shows very little dependence on temperature at altitudes above the 0°C isotherm but decreases with increasing temperature in the melting layer and below (Fig. 20). This may be due to the decreased particle number concentrations at the lower altitudes as shown in Fig. 16. The slight decrease in N02 in the melting layer while μ2 and λ2 remain nearly constant suggests the sublimation of small particles. Figure 20 shows μ2 increases and λ2 decreases with increasing temperature, which indicates that there was a broadening in the distribution to larger sizes (μ2) and a slight increase in the percentage of large particles compared to small particles (λ2). Again, this implies that aggregation occurred in the parallel stratiform region. The trend of decreasing λ is also noted in Heymsfield et al. (2015) in areas just above the melting layer during other flights in the MC3E field campaign.
As in Fig. 15, but for variation of the combined total hydrometeor number concentration of particles with diameters greater than 100 μm from the combined particle distribution created from the Two-Dimensional Cloud probe (2DC), Cloud Imaging Probe (CIP), and the High-Volume Precipitation Spectrometer Version-3 (HVPS-3).
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
It is unlikely that riming was the principal cause of particle growth. The icing detector does indicate the presence of supercooled liquid water after the third flight leg had been completed and into fourth flight leg. An analysis of particle habit in the 2DC and CIP imagery shows irregular, oblong particles, such as those shown in the 2DC imagery in Fig. 17, were common throughout the first four flight legs. These particles are more likely to have formed through aggregation attaching appendages to bigger snowflakes than for asymmetrical riming to have occurred.
Examples of (left) HVPS-3 and (right) 2DC probe images and (from top to bottom) mean temperature of the corresponding leg in order of descending height on 11 May 2011. The width of the HVPS-3 and 2DC image panels are 19.2 and 0.96 mm, respectively.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
Plot showing the observed unimodal spectrum at 1648:22 UTC (plus signs) and the observed bimodal spectrum (squares) at 1654:32 UTC 11 May 2011. Also plotted are the fitted unimodal gamma distribution (blue) and the fitted piecewise bimodal fit (red).
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
As in Fig. 15, but for variation of the (a) gamma intercept (N0), (b) gamma function breadth (μ), and (c) gamma function slope (λ) parameters with temperature.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
As in Fig. 15, but for (a)–(c) small-particle and (d)–(f) large-particle plots of piecewise gamma distribution fit.
Citation: Monthly Weather Review 150, 4; 10.1175/MWR-D-21-0009.1
The goal of M07 was to further the role of microphysics in MCS and bow echo evolution. The goal here is to compare a parallel stratiform MCS case with trailing stratiform MCSs to determine if the microphysical processes are similar. So far, the evidence says that this particular MCS has similar microphysical processes occurring as trailing stratiform MCSs. Snowflakes fall from the anvil region down through the stratiform region, growing in size as they collect other snowflakes in the aggregation process. In areas close to convection, riming of supercooled liquid water adds a second hydrometeor growth mechanism. In regions of sub-saturated air, sublimation removes small particles and particle mass. At the time of the in situ sampling of the stratiform layer (1640–1811 UTC), the effects of the leading convective line upon the stratiform region appear to be limited (Fig. 6) and the parallel convective region more influential.
4. Conclusions
Parker and Johnson (2000) classified MCSs into three types: trailing stratiform, leading stratiform, and parallel stratiform. Only a small fraction of studies has focused on parallel stratiform MCS systems. The 11 May 2011 hybrid parallel–trailing stratiform MCS was well sampled as it traversed across the MC3E experimental domain, which enabled a detailed case study.
Conclusions from this study are as follows:
-
The southerly winds present in the atmospheric column over Oklahoma and southern Kansas were oriented parallel to the developing/reorganizing convective line, initially producing a parallel stratiform region and pushing convective elements into the parallel stratiform region.
-
Anvil-level front-to-rear flow developed over time even as strong southerly flow through the stratiform region persisted, producing a hybrid parallel–trailing stratiform region.
-
In the western (trailing) half of the MCS, an area of line-perpendicular rear-to-front flow signaled the presence of rear-inflow, a feature only observed in a few recent studies of parallel stratiform MCS, but common in trailing stratiform systems.
-
The wind field in and in front of the parallel stratiform MCS (the upper-level wind parallel to the orientation of convective region, the southwesterly to south-southwesterly storm-relative inflow, and the southwesterly storm-relative surface outflow) is similar to those published in Parker (2007a) and Clark et al. (2014a,b).
-
Based on the surface meteorological measurements and the multi-Doppler wind field analysis, it appears that the cold pool influenced the evolution of the 11 May 2011 MCS through the generation of front-to-rear flow.
-
There are indications of sublimation and aggregation of ice particles occurring in the parallel stratiform region. These include the following:
-
Median Nt value decreasing at a faster rate than that of TWC decline.
-
The presence of bimodal spectra.
-
Increasing µ (dispersion) and decreasing λ (slope) with respect to increasing temperature in both the unimodal and bimodal gamma distributions.
-
-
Ice particle growth through riming probably occurring in regions of supersaturated air near convective cores.
-
The unimodal gamma distribution parameters in the parallel stratiform region follow similar trends with respect to temperature as those found in M07 and are indicative of ice particle aggregation and sublimation.
-
While M07 did not report any bimodality of ice particle spectra within their study, the bimodal gamma distribution parameters observed in the parallel stratiform region also indicate aggregation. It is uncertain why this study saw bimodality in the particle size distributions while M07 did not. Some possibilities explaining the difference include the longer averaging period in M07 compared to this study (60 s versus 10 s), differences in instrumentation and/or software, or different microphysical processes were present in this study.
Regardless of the differences, the similarities between the microphysical properties of this parallel stratiform MCS and the BAMEX trailing stratiform MCSs suggest there likely are many similarities in the dynamical and microphysical processes in the parallel stratiform and trailing stratiform regions.
This study uses the Nevzorov probe to measure the total water content of the MCS, whereas M07 use mass–dimensional relationships applied to the particle size distributions, constrained by estimates of radar reflectivity, to derive the mass of the particles. Since the TWC in this study and the TWC in M07 are derived from two different methods, differences between these two quantities do not necessarily mean that the clouds are dissimilar.
M07 used a two-dimensional Particle Measurement System precipitation probe (2DP) to measure the number distributions of large hydrometeors whereas this study uses a HVPS-3, which has a sample volume 1.8 times larger than the sample volume of the 2DP. The HVPS-3 also has a wider laser (19.2 mm) than the 2DP (6.4 mm; M07), allowing the HVPS-3 to fully image large particles and thus reduce sizing uncertainty associated with partially imaged hydrometeors and particle reconstruction algorithms. While both studies used similar particle rejection algorithms, the 2DC and HVPS-3 in MC3E were outfitted with ice shattering-resistant tips, to which M07 make no reference; thus, this study assumes that the 2DC and 2DP in BAMEX were not outfitted with these tips.
Acknowledgments.
This research was supported by NASA Grant NNX11AP12G. We gratefully acknowledge Aaron Bansemer of the National Center for Atmospheric Research as he processed the 2DC, CIP, and HVPS-3 data, and created the HVPS-3 images. Our thanks to the UND Citation fight crew and ground support team for their work in collecting the data.
Data availability statement.
The majority of the data and all of the software used in this study are open source and freely accessible. An archive of the Citation 1-Hz in-situ and NPOL data can be found at the NASA Global Hydrology Resource Center (Delene and Poellot 2012; Gerlach and Petersen 2012). The ARM archive website contains the rawinsonde data (Jensen and Toto 2014). The multi-Doppler wind retrieval data used in this study are available at the Stony Brook University Academic Commons (Oue et al. 2020). NWS WSR-88D data can be accessed via the National Centers for Environmental Information website (NCEI 2020). The Oklahoma Mesonet data can be requested through a web interface (https://www.mesonet.org/index.php/past_data/data_request_form) and may require a service fee. The specific datasets used to generate the figures shown in this study are available online at the UND Chester Fritz Library (Neumann et al. 2020). The in situ data are processed using the open-source ADPAA and SODA2 software packages (Delene et al. 2015; Bansemer 2019). A software repository housed at SourceForge (https://sourceforge.net/projects/neumannetal2021/) and archived via Zenodo (Neumann 2020) include all code to generate the figures used in this study. The wind retrieval code is available upon request by contacting Mariko Oue (mariko.oue@stonybrook.edu).
REFERENCES
Bansemer, A., 2019: SODA-2 Data Processing Software. GitHub, accessed 13 October 2016, https://github.com/abansemer/soda2.
Baumgardner, D., and A. Korolev, 1997: Airspeed corrections for optical array probe sample volumes. J. Atmos. Oceanic Technol., 14, 1224–1229, https://doi.org/10.1175/1520-0426(1997)014<1224:ACFOAP>2.0.CO;2.
Baumgardner, D., and Coauthors, 2017: Cloud ice properties: In situ measurement challenges. Ice Formation and Evolution in Clouds and Precipitation: Measurement and Modeling Challenges, Meteor. Monogr., No. 58, Amer. Meteor. Soc., 9.1–9.23, https://doi.org/10.1175/AMSMONOGRAPHS-D-16-0011.1.
Blumberg, W. G., K. T. Halbert, T. A. Supinie, P. T. Marsh, R. L. Thompson, and J. A. Hart, 2017: SHARPpy: An open-source sounding analysis toolkit for the atmospheric sciences. Bull. Amer. Meteor. Soc., 98, 1625–1636, https://doi.org/10.1175/BAMS-D-15-00309.1.
Brock, F. V., K. C. Crawford, R. L. Elliott, G. W. Cuperus, S. J. Stadler, H. L. Johnson, and M. D. Eilts, 1995: The Oklahoma Mesonet: A technical overview. J. Atmos. Oceanic Technol., 12, 5–19, https://doi.org/10.1175/1520-0426(1995)012<0005:TOMATO>2.0.CO;2.
Chappell, C. F., 1986: Quasi-stationary convective events. Mesoscale Meteorology and Forecasting, P.S. Ray, Ed., Amer. Meteor. Soc., 289–310.
Clark, P. A., K. A. Browning, C. J. Morcrette, A. M. Blyth, R. M. Forbes, B. Brooks, and F. Perry, 2014a: The evolution of an MCS over southern England. Part 1: Observations. Quart. J. Roy. Meteor. Soc., 140, 439–457, https://doi.org/10.1002/qj.2138.
Clark, P. A., K. A. Browning, R. M. Forbes, C. J. Morcrette, A. M. Blyth, and H. W. Lean, 2014b: The evolution of an MCS over southern England. Part 2: Model simulations and sensitivity to microphysics. Quart. J. Roy. Meteor. Soc., 140, 458–479, https://doi.org/10.1002/qj.2142.
Cui, W., X. Dong, B. Xi, J. Fan, J. Tian, J. Wang, and T. M. McHardy, 2019: Understanding ice cloud-precipitation properties of three modes of mesoscale convective systems during PECAN. J. Geophys. Res. Atmos., 124, 4121–4140, https://doi.org/10.1029/2019JD030330.
Davis, C., and Coauthors, 2004: The Bow Echo and MCV Experiment: Observations and opportunities. Bull. Amer. Meteor. Soc., 85, 1075–1094, https://doi.org/10.1175/BAMS-85-8-1075.
Delene, D., 2011: Airborne data processing and analysis software package. Earth Sci. Inform., 4, 29–44, https://doi.org/10.1007/s12145-010-0061-4.
Delene, D., and M. R. Poellot, 2012: GPM Ground Validation UND Citation Cloud Microphysics MC3E. NASA Global Hydrology Resource Center Distributed Active Archive Center, accessed 5 September 2020, https://doi.org/10.5067/GPMGV/MC3E/MULTIPLE/DATA201.
Delene, D., A. Neumann, C. Kruse, R. Mitchell, G. Sever, and M. Starzec, 2015: Airborne Data Processing and Analysis Software Package. Zenodo, accessed 3 June 2021, https://doi.org/10.5281/zenodo.14053.
Delene, D., K. Hibert, M. Poellot, and N. Brackin, 2019: The North Dakota Citation Research Aircraft Measurement Platform. SAE Tech. Paper 2019-01-1990, SAE International, 5 pp., https://doi.org/10.4271/2019-01-1990.
Duda, J. D., and W. A. Gallus, 2010: Spring and summer Midwestern severe weather reports in supercells compared to other morphologies. Wea. Forecasting, 25, 190–206, https://doi.org/10.1175/2009WAF2222338.1.
Field, P. R., 2000: Bimodal ice spectra in frontal clouds. Quart. J. Roy. Meteor. Soc., 126, 379–392, https://doi.org/10.1002/qj.49712656302.
Gallus, W. A., N. A. Snook, and E. V. Johnson, 2008: Spring and summer severe weather reports over the Midwest as a function of convective mode: A preliminary study. Wea. Forecasting, 23, 101–113, https://doi.org/10.1175/2007WAF2006120.1.
Garrett, T. J., A. J. Heymsfield, M. J. McGill, B. A. Ridley, D. G. Baumgardner, T. P. Bui, and C. R. Webster, 2004: Convective generation of cirrus near the tropopause. J. Geophys. Res. Atmos., 109, D21203, https://doi.org/10.1029/2004JD004952.
Gayet, J.-F., and Coauthors, 2004: Cirrus cloud microphysical and optical properties at southern and northern midlatitudes during the INCA experiment. J. Geophys. Res. Atmos., 109, D20206, https://doi.org/10.1029/2004JD004803.
Geerts, B., and Coauthors, 2017: The 2015 Plains Elevated Convection At Night Field Project. Bull. Amer. Meteor. Soc., 98, 767–786, https://doi.org/10.1175/BAMS-D-15-00257.1.
Gerlach, J., and W. A. Petersen, 2012: GPM Ground Validation NASA S-Band Dual Polarimetric (NPOL) Doppler Radar MC3E. NASA Global Hydrology Resource Center DAAC, accessed 1 June 2021, https://doi.org/10.5067/GPMGV/MC3E/NPOL/DATA101.
Giangrande, S. E., S. Collis, J. Straka, A. Protat, C. Williams, and S. Krueger, 2013: A summary of convective-core vertical velocity properties using ARM UHF wind profilers in Oklahoma. J. Appl. Meteor. Climatol., 52, 2278–2295, https://doi.org/10.1175/JAMC-D-12-0185.1.
Glickman, T. S., 2000: Glossary of Meteorology. 2nd ed. Amer. Meteor. Soc., 855 pp.
Gossard, E. E., R. O. Strauch, and R. R. Rogers, 1990: Evolution of dropsize distributions in liquid precipitation observed by ground-based Doppler radar. J. Atmos. Oceanic Technol., 7, 815–828, https://doi.org/10.1175/1520-0426(1990)007<0815:EODDIL>2.0.CO;2.
Hallett, J., and S. C. Mossop, 1974: Production of secondary ice particles during the riming process. Nature, 249, 26–28, https://doi.org/doi:10.1038/249026a0.
Halligan, E. A., and M. D. Parker, 2004: 22 June 2003 BAMEX observations of a convective line with parallel precipitation. 22nd Conf. on Severe Local Storms, Hyannis, MA, Amer. Meteor. Soc., P3.5.
Helmus, J. J., and S. M. Collis, 2016: The Python ARM Radar Toolkit (Py-ART), a library for working with weather radar data in the Python programming language. J. Open Res. Software, 4, e25, https://doi.org/10.5334/jors.119.
Heymsfield, A. J., 1985: Summary of a workshop on processing 2-D probe data. Bull. Amer. Meteor. Soc., 66, 437–440, https://doi.org/10.1175/1520-0477-66.4.437.
Heymsfield, A. J., and J. L. Parrish, 1978: A computational technique for increasing the effective sampling volume of the PMS two-dimensional particle size spectrometer. J. Appl. Meteor., 17, 1566–1572, https://doi.org/10.1175/1520-0450(1978)017<1566:ACTFIT>2.0.CO;2.
Heymsfield, A. J., L. M. Miloshevich, C. Schmitt, A. Bansemer, C. Twohy, M. R. Poellot, A. Fridlind, and H. Gerber, 2005: Homogeneous ice nucleation in subtropical and tropical convection and its influence on cirrus anvil microphysics. J. Atmos. Sci., 62, 41–64, https://doi.org/10.1175/JAS-3360.1.
Heymsfield, A. J., A. Bansemer, M. R. Poellot, and N. Wood, 2015: Observations of ice microphysics through the melting layer. J. Atmos. Sci., 72, 2902–2928, https://doi.org/10.1175/JAS-D-14-0363.1.
Houze, R. A., Jr., S. A. Rutledge, M. I. Biggerstaff, and B. F. Smull, 1989: Interpretation of Doppler weather-radar displays in midlatitude mesoscale convective systems. Bull. Amer. Meteor. Soc., 70, 608–619, https://doi.org/10.1175/1520-0477(1989)070<0608:IODWRD>2.0.CO;2.
Jensen, E. J., and Coauthors, 2005: Ice supersaturations exceeding 100% at the cold tropical tropopause: Implications for cirrus formation and dehydration. Atmos. Chem. Phys., 5, 851–862, https://doi.org/10.5194/acp-5-851-2005.
Jensen, M. P., and T. Toto, 2014: Humidity-corrected sounding observations (SONDEADJUST)—April–June 2011, 36.695°N, 97.485°W; 37.7°N, 98.75°W; 37.674°N, 95.488°W; 36.071°N, 99.204°W; 35.678°N, 95.586°W; 34.985°N, 97.522°W: Southern Great Plains Sites C1, S1, S2, S3, S4, S5. Atmospheric Radiation Measurement (ARM) Climate Research Facility Data Archive, accessed 25 August 2020, https://doi.org/10.5439/1164928.
Jensen, M. P., and Coauthors, 2015: The Midlatitude Continental Convective Clouds Experiment (MC3E) sounding network: Operations, processing and analysis. Atmos. Meas. Tech., 8, 421–434, https://doi.org/10.5194/amt-8-421-2015.
Jensen, M. P., and Coauthors, 2016: The Midlatitude Continental Convective Clouds Experiment (MC3E). Bull. Amer. Meteor. Soc., 97, 1667–1686, https://doi.org/10.1175/BAMS-D-14-00228.1.
Korolev, A., and P. R. Field, 2015: Assessment of the performance of the inter-arrival time algorithm to identify ice shattering artifacts in cloud particle probe measurements. Atmos. Meas. Tech., 8, 761–777, https://doi.org/10.5194/amt-8-761-2015.
Korolev, A., J. W. Strapp, G. A. Isaac, and A. N. Nevzorov, 1998: The Nevzorov airborne hot-wire LWC–TWC probe: Principle of operation and performance characteristics. J. Atmos. Oceanic Technol., 15, 1495–1510, https://doi.org/10.1175/1520-0426(1998)015<1495:TNAHWL>2.0.CO;2.
Korolev, A., E. F. Emery, J. W. Strapp, S. G. Cober, G. A. Isaac, M. Wasey, and D. Marcotte, 2011: Small ice particles in tropospheric clouds: Fact or artifact? Airborne icing instrumentation evaluation experiment. Bull. Amer. Meteor. Soc., 92, 967–973, https://doi.org/10.1175/2010BAMS3141.1.
Korolev, A., J. W. Strapp, G. A. Isaac, and E. Emery, 2013: Improved airborne hot-wire measurements of ice water content in clouds. J. Atmos. Oceanic Technol., 30, 2121–2131, https://doi.org/10.1175/JTECH-D-13-00007.1.
Lenschow, H., Ed., 1986: Aircraft measurements in the boundary layer. Probing Atmospheric Boundary Layer, Amer. Meteor. Soc., 39–55.
Lin, Y.-L., R. D. Farley, and H. D. Orville, 1983: Bulk parameterization of the snow field in a cloud model. J. Climate Appl. Meteor., 22, 1065–1092, https://doi.org/10.1175/1520-0450(1983)022<1065:BPOTSF>2.0.CO;2.
Liu, C., and M. W. Moncrieff, 2017: Shear-parallel mesoscale convective systems in a moist low-inhibition mei-yu front environment. J. Atmos. Sci., 74, 4213–4228, https://doi.org/10.1175/JAS-D-17-0121.1.
Mace, G. G., A. J. Heymsfield, and M. R. Poellot, 2002: On retrieving the microphysical properties of cirrus clouds using the moments of the millimeter-wavelength Doppler spectrum. J. Geophys. Res. Atmos., 107, 4815, https://doi.org/10.1029/2001JD001308.
Markowski, P., and Y. Richardson, 2010: Mesoscale Meteorology in Midlatitudes. Wiley-Blackwell, 407 pp.
McFarquhar, G. M., M. S. Timlin, R. M. Rauber, B. F. Jewett, J. A. Grim, and D. P. Jorgensen, 2007: Vertical variability of cloud hydrometeors in the stratiform region of mesoscale convective systems and bow echoes. Mon. Wea. Rev., 135, 3405–3428, https://doi.org/10.1175/MWR3444.1.
McPherson, R. A., and Coauthors, 2007: Statewide monitoring of the mesoscale environment: A technical update on the Oklahoma Mesonet. J. Atmos. Oceanic Technol., 24, 301–321, https://doi.org/10.1175/JTECH1976.1.
Mitchell, D. L., S. K. Chai, Y. Liu, A. J. Heymsfield, and Y. Dong, 1996: Modeling cirrus clouds. Part I: Treatment of bimodal size spectra and case study analysis. J. Atmos. Sci., 53, 2952–2966, https://doi.org/10.1175/1520-0469(1996)053<2952:MCCPIT>2.0.CO;2.
Moller, A. R., C. A. Doswell, M. P. Foster, and G. R. Woodall, 1994: The operational recognition of supercell thunderstorm environments and storm structures. Wea. Forecasting, 9, 327–347, https://doi.org/10.1175/1520-0434(1994)009<0327:TOROST>2.0.CO;2.
Moore, J. T., and J. P. Gagan, 2000: Examination of a training cold season heavy rain event over the Ohio River valley. 20th Conf. on Severe Local Storms, Orlando, FL, Amer. Meteor. Soc., 465–468.
NCEI, 2020: Radar data. Accessed 5 September 2020, https://www.ncdc.noaa.gov/data-access/radar-data.
Neumann, A., 2020: Software for a kinematic and microphysical analysis of a parallel stratiform MCS. Zenodo, accessed 19 October 2020, https://doi.org/10.5281/zenodo.4107069.
Neumann, A., M. Poellot, D. Delene, M. Askelson, K. North, and M. Oue, 2020: A multi-platform, in-situ kinematic and microphysical analysis of a parallel stratiform mesoscale convective system. University of North Dakota Scholarly Commons, accessed 21 October 2020, https://doi.org/10.31356/data018.
North, K. W., M. Oue, P. Kollias, S. E. Giangrande, S. M. Collis, and C. K. Potvin, 2017: Vertical air motion retrievals in deep convective clouds using the ARM scanning radar network in Oklahoma during MC3E. Atmos. Meas. Tech., 10, 2785–2806, https://doi.org/10.5194/amt-10-2785-2017.
Oue, M., K. Inagaki, T. Shinoda, T. Ohigashi, T. Kouketsu, M. Kato, K. Tsuboki, and H. Uyeda, 2014: Polarimetric Doppler radar analysis of organization of a stationary rainband with changing orientations in July 2010. J. Meteor. Soc. Japan, 92, 457–481, https://doi.org/10.2151/jmsj.2014-503.
Oue, M., K. North, and A. Neumann, 2020: Multi-Doppler radar wind retrieval data for deep convective clouds observed in the Southern Great Plains on May 11, 2011. SoMAS Research Data 9, accessed 1 December 2020, https://commons.library.stonybrook.edu/somasdata/9.
Parker, M. D., 2007a: Simulated convective lines with parallel stratiform precipitation. Part I: An archetype for convection in along-line shear. J. Atmos. Sci., 64, 267–288, https://doi.org/10.1175/JAS3853.1.
Parker, M. D., 2007b: Simulated convective lines with parallel stratiform precipitation. Part II: Governing dynamics and associated sensitivities. J. Atmos. Sci., 64, 289–313, https://doi.org/10.1175/JAS3854.1.
Parker, M. D., and R. H. Johnson, 2000: Organizational modes of midlatitude mesoscale convective systems. Mon. Wea. Rev., 128, 3413–3436, https://doi.org/10.1175/1520-0493(2001)129<3413:OMOMMC>2.0.CO;2.
Parker, M. D., S. A. Rutledge, and R. H. Johnson, 2001: Cloud-to-ground lightning in linear mesoscale convective systems. Mon. Wea. Rev., 129, 1232–1242, https://doi.org/10.1175/1520-0493(2001)129<1232:CTGLIL>2.0.CO;2.
Press, W. H., S. A. Teukolsky, W. T. Vetterling, and B. P. Flannery, 1992: Numerical Recipes in FORTRAN: The Art of Scientific Computing. Cambridge University Press, 963 pp.
Radhakrishna, B., and T. N. Rao, 2009: Multipeak raindrop size distribution observed by UHF/VHF wind profilers during the passage of a mesoscale convective system. Mon. Wea. Rev., 137, 976–990, https://doi.org/10.1175/2008MWR2607.1.
Roebber, P. J., and J. Eise, 2001: The 21 June 1997 flood: Storm-scale simulations and implications for operational forecasting. Wea. Forecasting, 16, 197–218, https://doi.org/10.1175/1520-0434(2001)016<0197:TJFSSS>2.0.CO;2.
Rogash, J. A., and R. D. Smith, 2000: Multiscale overview of a violent tornado outbreak with attendant flash flooding. Wea. Forecasting, 15, 416–431, https://doi.org/10.1175/1520-0434(2000)015<0416:MOOAVT>2.0.CO;2.
Rotunno, R., J. B. Klemp, and M. L. Weisman, 1988: A theory for strong, long-lived squall lines. J. Atmos. Sci., 45, 463–485, https://doi.org/10.1175/1520-0469(1988)045<0463:ATFSLL>2.0.CO;2.