• Bluestein, H. B., 2009: The formation and early evolution of the Greensburg, Kansas, tornadic supercell on 4 May 2007. Wea. Forecasting, 24, 899920, doi:10.1175/2009WAF2222206.1.

    • Search Google Scholar
    • Export Citation
  • Bluestein, H. B., 2013: Severe Convective Storms and Tornadoes: Observations and Dynamics. Springer/Praxis, 456 pp.

  • Brock, F. V., , Crawford K. , , Elliott R. , , Cuperus G. , , Stadler S. , , Johnson H. , , and Eilts M. , 1995: The Oklahoma Mesonet: A technical overview. J. Atmos. Oceanic Technol., 12, 519, doi:10.1175/1520-0426(1995)012<0005:TOMATO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Clark, A. J., , Gallus W. A. , , Xue M. , , and Kong F. , 2010: Convection-allowing and convection-parameterizing ensemble forecasts of a mesoscale convective vortex and associated severe weather environment. Wea. Forecasting, 25, 10521081, doi:10.1175/2010WAF2222390.1.

    • Search Google Scholar
    • Export Citation
  • Coleman, T. A., , and Knupp K. R. , 2011: Radiometer and profiler analysis of the effects of a bore and solitary wave on the stability of the nocturnal boundary layer. Mon. Wea. Rev., 139, 211223, doi:10.1175/2010MWR3376.1.

    • Search Google Scholar
    • Export Citation
  • Coniglio, M. C., , Elmore K. L. , , Kain J. S. , , Weiss S. J. , , Xue M. , , and Weisman M. L. , 2010: Evaluation of WRF Model output for severe weather forecasting from the 2008 NOAA Hazardous Weather Testbed Spring Experiment. Wea. Forecasting, 25, 408427, doi:10.1175/2009WAF2222258.1.

    • Search Google Scholar
    • Export Citation
  • Crook, N. A., 1986: The effect of ambient stratification and moisture on the motion of atmospheric undular bores. J. Atmos. Sci., 43, 171181, doi:10.1175/1520-0469(1986)043<0171:TEOASA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Crook, N. A., 1988: Trapping of low-level internal gravity waves. J. Atmos. Sci., 45, 15331541, doi:10.1175/1520-0469(1988)045<1533:TOLLIG>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Crum, T. D., , and Alberty R. L. , 1993: The WSR-88D and the WSR-88D Operational Support Facility. Bull. Amer. Meteor. Soc., 74, 16691687, doi:10.1175/1520-0477(1993)074<1669:TWATWO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dabberdt, W. F., , and Schlatter T. W. , 1996: Research opportunities from emerging atmospheric observing and modeling capabilities. Bull. Amer. Meteor. Soc., 77, 305323, doi:10.1175/1520-0477(1996)077<0305:ROFEAO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Droegemeier, K. K., 1997: The numerical prediction of thunderstorms: Challenges, potential benefits, and results from real-time operational tests. WMO Bull., 46, 324336.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K., and et al. , 1995: Report of the First Prospectus Development Team of the U.S. Weather Research Program to NOAA and the NSF. Bull. Amer. Meteor. Soc., 76, 11941208.

    • Search Google Scholar
    • Export Citation
  • Fiebrich, C. A., , and Crawford K. C. , 2001: The impact of unique meteorological phenomena detected by the Oklahoma Mesonet and ARS Micronet on automated quality control. Bull. Amer. Meteor. Soc., 82, 21732187, doi:10.1175/1520-0477(2001)082<2173:TIOUMP>2.3.CO;2.

    • Search Google Scholar
    • Export Citation
  • Goff, R. C., 1976: Vertical structure of thunderstorm outflows. Mon. Wea. Rev., 104, 14291440, doi:10.1175/1520-0493(1976)104<1429:VSOTO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Haertel, P. T., , Johnson R. H. , , and Tulich S. N. , 2001: Some simple simulations of thunderstorm outflows. J. Atmos. Sci., 58, 504516, doi:10.1175/1520-0469(2001)058<0504:SSSOTO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Johns, R. H., , and Doswell C. A. III, 1992: Severe local storms forecasting. Wea. Forecasting, 7, 588612, doi:10.1175/1520-0434(1992)007<0588:SLSF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Johnson, B. C., 1983: The heat burst of 29 May 1976. Mon. Wea. Rev., 111, 17761792, doi:10.1175/1520-0493(1983)111<1776:THBOM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Johnson, R. H., , Chen S. , , and Toth J. J. , 1989: Circulations associated with a mature-to-decaying midlatitude mesoscale convective system. Part I: Surface features—Heat bursts and mesolow development. Mon. Wea. Rev., 117, 942959, doi:10.1175/1520-0493(1989)117<0942:CAWAMT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kain, J. S., and et al. , 2013: A feasibility study for probabilistic convection initiation forecasts based on explicit numerical guidance. Bull. Amer. Meteor. Soc., 94, 12131225, doi:10.1175/BAMS-D-11-00264.1.

    • Search Google Scholar
    • Export Citation
  • Koch, S. E., , Dorian P. B. , , Ferrare R. , , Melfi S. , , Skillman W. C. , , and Whiteman D. , 1991: Structure of an internal bore and dissipating gravity current as revealed by Raman lidar. Mon. Wea. Rev., 119, 857887, doi:10.1175/1520-0493(1991)119<0857:SOAIBA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Koch, S. E., , Flamant C. , , Wilson J. W. , , Gentry B. M. , , and Jamison B. D. , 2008: An atmospheric soliton observed with Doppler radar, differential absorption lidar, and a molecular Doppler lidar. J. Atmos. Oceanic Technol., 25, 12671287, doi:10.1175/2007JTECHA951.1.

    • Search Google Scholar
    • Export Citation
  • Leary, C. A., , and Houze R. A. Jr., 1979: Melting and evaporation of hydrometeors in precipitation from the anvil clouds of deep tropical convection. J. Atmos. Sci., 36, 669679, doi:10.1175/1520-0469(1979)036<0669:MAEOHI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lilly, D. K., 1990: Numerical prediction of thunderstorms—Has its time come? Quart. J. Roy. Meteor. Soc., 116, 779798, doi:10.1002/qj.49711649402.

    • Search Google Scholar
    • Export Citation
  • Marsham, J. H., , Trier S. B. , , Weckwerth T. M. , , and Wilson J. W. , 2011: Observations of elevated convection initiation leading to a surface-based squall line during 13 June _IHOP 2002. Mon. Wea. Rev., 139, 247271, doi:10.1175/2010MWR3422.1.

    • Search Google Scholar
    • Export Citation
  • Moller, A. R., , Doswell C. A. III, , Foster M. P. , , and Woodall G. R. , 1994: The operational recognition of supercell thunderstorm environments and storm structures. Wea. Forecasting, 9, 327347, doi:10.1175/1520-0434(1994)009<0327:TOROST>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • National Research Council, 1998: The Atmospheric Sciences: Entering the Twenty-First Century.National Academy Press, 364 pp.

  • OFCM, 1995: Surface weather observations and reports. Federal Meteorological Handbook 1, Office of the Federal Coordinator for Meteorological Services and Supporting Research, 99 pp.

  • Pazmany, A. L., , Mead J. B. , , Bluestein H. B. , , Snyder J. C. , , and Houser J. B. , 2013: A mobile, rapid-scanning, X-band, polarimetric (RaXPol) Doppler radar system. J. Atmos. Oceanic Technol., 30, 13981413, doi:10.1175/JTECH-D-12-00166.1.

    • Search Google Scholar
    • Export Citation
  • Rottman, J. W., , and Simpson J. E. , 1989: The formation of internal bores in the atmosphere: A laboratory model. Quart. J. Roy. Meteor. Soc., 115, 941963, doi:10.1002/qj.49711548809.

    • Search Google Scholar
    • Export Citation
  • Rotunno, R., , Klemp J. B. , , and Weisman M. L. , 1988: A theory for strong, long-lived squall lines. J. Atmos. Sci., 45, 463485, doi:10.1175/1520-0469(1988)045<0463:ATFSLL>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Schroeder, J. L., , Burgett W. S. , , Haynie K. B. , , Sonmez I. , , Skwira G. D. , , Doggett A. L. , , and Lipe J. W. , 2005: The West Texas Mesonet: A technical overview. J. Atmos. Oceanic Technol., 22, 211222, doi:10.1175/JTECH-1690.1.

    • Search Google Scholar
    • Export Citation
  • Scofield, R. A., , and Purdom J. F. W. , 1986: The use of satellite data for mesoscale analyses and forecasting applications Mesoscale Meteorology and Forecasting, P. S. Ray, Ed., Amer. Meteor. Soc., 118 – 150.

  • Scorer, R. S., 1949: Theory of waves in the lee of mountains. Quart. J. Roy. Meteor. Soc., 75, 4156, doi:10.1002/qj.49707532308.

  • Simpson, J. E., 1997: Gravity Currents in the Environment and the Laboratory.2nd ed. Cambridge Universitiy Press, 244 pp.

  • Smull, B. F., , and Houze R. A. Jr., 1987: Rear inflow in squall lines with trailing stratiform precipitation. Mon. Wea. Rev., 115, 28692889, doi:10.1175/1520-0493(1987)115<2869:RIISLW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Trapp, R. J., , Stumpf G. J. , , and Manross K. L. , 2005: A reassessment of the percentage of tornadic mesocyclones. Wea. Forecasting, 20, 680687, doi:10.1175/WAF864.1.

    • Search Google Scholar
    • Export Citation
  • Wakimoto, R. M., 1982: The life cycle of thunderstorm gust fronts as viewed with Doppler radar and rawinsonde data. Mon. Wea. Rev., 110, 10601082, doi:10.1175/1520-0493(1982)110<1060:TLCOTG>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Weber, B. L., and et al. , 1990: Preliminary evaluation of the first NOAA Demonstration Network Wind Profiler. J. Atmos. Oceanic Technol., 7, 909918, doi:10.1175/1520-0426(1990)007<0909:PEOTFN>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Weckwerth, T. M., , and Parsons D. P. , 2006: A review of convection initiation and motivation for IHOP 2002. Mon. Wea. Rev., 134, 522, doi:10.1175/MWR3067.1.

    • Search Google Scholar
    • Export Citation
  • Weisman, M. L., 1992: Convectively generated rear-inflow jets in the evolution of long-lived mesoconvective systems. J. Atmos. Sci., 49, 18261847, doi:10.1175/1520-0469(1992)049<1826:TROCGR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Weisman, M. L., , Davis C. , , Wang W. , , Manning K. W. , , and Klemp J. B. , 2008: Experiences with a 0–36-h explicit convective forecasts with the WRF-ARW Model. Wea. Forecasting, 23, 407437, doi:10.1175/2007WAF2007005.1.

    • Search Google Scholar
    • Export Citation
  • Weiss, C. C., , Bluestein H. B. , , and Pazmany A. L. , 2006: Finescale radar observations of the 22 May 2002 dryline during the International H2O Project (IHOP). Mon. Wea. Rev., 134, 273293, doi:10.1175/MWR3068.1.

    • Search Google Scholar
    • Export Citation
  • Zipser, E. J., 1977: Mesoscale and convective-scale downdrafts as distinct components of squall-line structure. Mon. Wea. Rev., 105, 15681589, doi:10.1175/1520-0493(1977)105<1568:MACDAD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Map of stations in the Oklahoma Mesonet in central, south-central, southwestern, and western Oklahoma. Data from many of these are alluded to in the text. The stations enclosed within the polygon exhibited similar behavior when a gust front passed by, as discussed later in the text.

  • View in gallery

    Specialized forecast and verification products from the SPC on 30 May 2012. (a) The “day 1” convective outlook by category issued at 1234 UTC. (b) As in (a), but for the probability of tornadoes within 25 mi (60 km) of any point. (c),(d) As in (a),(b), but for the convective outlook issued at 1630 UTC. (e) Filtered storm reports for the day. Locations of tornadoes are in red, along with high-wind [blue squares >50 knots (kt; 1 kt = 0.51 m s−1; therefore, 50 kt = 25.8 m s−1); black squares >65 kt (33.5 m s−1)] and large-hail [green triangles >1 in. (2.5 cm) in diameter; black triangles >2 in. (5.1 cm) in diameter] reports. (f) Max surface wind gusts [mi h−1; 2.2(m s−1)] in the hour prior to 0510 UTC 31 May 2012. The solid line encloses reports of wind gusts (red) in excess of 60 mi h−1 (27.3 m s−1); the dashed line encloses reports of wind gusts (orange) in excess of 40 mi h−1 (18.2 m s−1). (g) As in (f), but for wind gusts in excess of 60 mi h−1 (27.3 m s−1) at 0500–1330 UTC to show that there were sporadic high-wind reports later at night, but these were not associated with the bore to be discussed in the text.

  • View in gallery

    The 500-hPa synoptic maps for the continental United States at (a) 1200 UTC 30 May and (b) 0000 UTC 31 May 2012. (Prepared by the SPC.) Trough axes in the southwestern portion of the United States are marked by brown-dashed segments. The arrow in (b) points to a narrow jet marked by 50–55 kt (25.8–28.4 m s−1) wind observations near the New Mexico–southwestern Texas border. (c) As in (b), but for a 12-h forecast from the operational 12-km NAM initialized at 1200 UTC 30 May and valid at 0000 UTC 31 May 2012.

  • View in gallery

    As in Figs. 3a,b, but for 250 hPa. An area of strong CVA at the left-front quadrant of a jet streak and downstream from a zonally oriented trough axis is noted over southwestern Oklahoma and northwestern Texas in (b).

  • View in gallery

    WSR-88D reflectivity constant altitude plan position indicator (CAPPI) composites for the region centered at 12 000 ft (3.7 km) over Oklahoma at (a) 0300, (b) 0600, (c) 0900, (d) 1200, (e) 1500, and (f) 1800 UTC 30 May 2012. The highest (lowest) reflectivities are shown as warm (cool) colors. The solid lines represent the 15°C isodrosotherm. In (f), the arrow points to a dissipating area of elevated convection that has disappeared by 2100 UTC. (Courtesy of the UCAR/EOL/DC-3 website.)

  • View in gallery

    Surface dewpoint temperature (°C) in the south-central United States at (a) 1500, (b) 1600, (c) 1700, and (d) 1800 UTC 30 May 2012. The solid line is the 15°C isodrosotherm. Warm (cool) colors denote high (low) dewpoints. Five areas of drying are identified: A is a dry air mass to the rear of the dryline; B is a dry air mass in Kansas, north of the returning moisture in Oklahoma; C is a dry pocket in far southwestern Arkansas, northeastern Texas, and southeastern Oklahoma; D is a dry pocket in northwestern Texas, extending just over the border into south-central/southwestern Oklahoma; and E is a dry pocket in west-central Oklahoma.

  • View in gallery

    Regional surface map centered on Oklahoma at 1843 UTC 30 May 2012. Solid contour is the 60°F (15.6°C) isodrosotherm. The dashed red lines mark the dryline in the southern Texas plains and a trough axis in northeastern Colorado and the Nebraska Panhandle. The dashed blue lines mark a cold front in western Texas and eastern New Mexico and an outflow boundary in central Kansas. The solid blue line at the border between northeastern Colorado and the Nebraska Panhandle marks a cold front. The cross is located at the center of a cyclone at the northern extent of the dryline. The arrow just east of the cyclone denotes the surface wind direction. To the right of the arrow, the surface temperatures are in the mid-80°F (~29°–30°C); to the left, temperatures are in the upper 90°F (~38°C). Therefore, there is a significant component of the temperature-gradient vector directed normal to and to the left of the surface wind. (Courtesy of UCAR/NCAR/RAL.)

  • View in gallery

    Hourly surface weather observations at the Oklahoma Mesonet sites at (a) 1330, (b) 1430, (c) 1530, (d) 1630, (e) 1730, (f) 1830, (g) 1930, and (h) 2030 UTC 30 May 2012. Winds are plotted according to convention; dewpoint temperatures are plotted such that the lowest (highest) dewpoints are shaded in cool (warm) colors. The white rectangles in the Texas Panhandle cover up a data-void region. Dry regions labeled in Fig. 6 are noted.

  • View in gallery

    Visible Geostationary Operational Environmental Satellite (GOES) imagery for the central United States at (a) 1915 and (b) 2132 UTC 30 May 2012 (from UCAR/NCAR/RAL). (c) As in (a),(b), but for a close-up view of portions of the Texas Panhandle, western Texas, southwestern and south-central Oklahoma, and northwestern Texas, at 1945 UTC (from the College of DuPage NeXt Generation Weather Laboratory). (d) Photograph of stratus cloud bands in southwestern Oklahoma, looking to the north, at ~2100 UTC (courtesy of H. Bluestein). In (a),(b), regions are identified: A shows clouds associated with the front in southwestern Nebraska, B is the trough axis in Colorado (cf. Fig. 7), C is the outflow boundary in central Kansas (cf. Figure 7), and D is the dryline in Texas (cf. Fig. 7); the dashed red line marks the axis of dry air at the surface (cf. Fig. 7). In (b), arrows point to convective storms along the front in Nebraska (A), the trough axis in northwestern Kansas (B; which earlier had been in Colorado), the outflow boundary in central Kansas (C), and the dryline in Texas (D). The arrow in (c) points to the bands of stratus seen in (d).

  • View in gallery

    Evolution of convective storms that formed along the dryline into an MCS. (a)–(d) Broader and (e)–(g) zoomed-in views of the tornadic supercell. Radar reflectivity factor [dBZ; color code shown in (a)] from the KFDR (Frederick, Oklahoma) WSR-88D at (a) 2200 and (b) 2300 UTC 30 May; and at (c) 0000, (d) 0100, and (e) 2315 UTC 31 May 2012, at 0.5°-elevation angle. Range markers are shown every 25 km in (a)–(d). The southernmost cell was a tornadic supercell in (b),(c). In (e), the arrow points to a supercell that subsequently produced a tornado. Range markers are shown every 15 km in (e). (f) Radar reflectivity factor (dBZ; color code shown) at 1.9°-elevation angle from RaXPol at 2317 UTC 30 May 2012. WEH is the weak-echo hole associated with the tornado. (g) As in (f), but for Doppler velocity (m s−1; color coded); the circle encloses the cyclonic vortex signature marking the location of the tornado. Range markers are shown every 5 km in (f),(g).

  • View in gallery

    Regional evolution of two MCSs at (a) 0305 and (b) 0501 UTC 31 May 2012, showing the dissipation of the western portion of the northern MCS. Oklahoma City is indicated by OKC. The region of dry air in western Oklahoma is noted. Composite low-level radar reflectivity factor (dBZ; color coded). (Courtesy of UCAR/NCAR/RAL.)

  • View in gallery

    Simulated low-level radar reflectivity factor [dBZ; warm (cool) colors indicate the highest (lowest) reflectivity] from the 1200 UTC operational 12-km NAM for the U.S. southern plains at (a) 2100 UTC 30 May and at (b) 0000, (c) 0300, and (d) 0600 UTC 31 May 2012. Convective storms are not explicitly simulated. Arrow in (b) points to convective storms that had formed along the dryline.

  • View in gallery

    As in Fig. 12, but for composite reflectivity (dBZ; color coded) at hourly intervals over the U.S. southern plains and environs from the 1800 UTC HRRR model at (a) 2000, (b) 2100, (c) 2200, and (d) 2300 UTC 30 May, and at (e) 0000, (f) 0100, (g) 0200, and (h) 0300 UTC 31 May 2012.

  • View in gallery

    As in Fig. 8, but for hourly surface weather observations at (a) 2130, (b) 2230, and (c) 2330 UTC 30 May, and at (d) 0030, (e) 0130, (f) 0230, (g) 0330, and (h) 0430 UTC 31 May 2012. Arrows in (b) point to locations likely affected by outflow, in which the winds are from the east, not the south, as they are at surrounding sites.

  • View in gallery

    Meteograms (traces of meteorological variables), displaying sudden drying in southwestern and western Oklahoma from the late afternoon to early evening. Data are plotted from 2000 UTC 30 May to 0700 UTC 31 May 2012 (times displayed in a.m. and p.m. along the abscissa are shown in CDT; times displayed vertically are shown in UTC): (top) wind direction (°; red) and wind speed (m s−1; blue); and (bottom) temperature (°C; red), dewpoint (°C; blue), and pressure (hPa; green) at Oklahoma Mesonet sites (a) Altus, (b) Mangum, (c) Retrop, and (d) Erick. In (a),(b), the double arrow points out a short-lived warming and drying episode; in (c),(d), the double arrow points out a drying episode, which persisted for hours. Short single arrows show the beginning of pressure rise, which marked late-evening gust-front passage.

  • View in gallery

    Radar imagery from KFDR at 2219–2222 UTC 30 May 2012, for southwestern Oklahoma, when episodes of warming and drying were beginning at Altus and Mangum [locations indicated in (a) and (c)] but before drying was observed at Retrop [locations indicated in (a) and (c)]. Radar reflectivity factor [dBZ; color scale at bottom of (a)] at (a) 0.5°-, (b) 4°-, and (c) 6.4°-elevation angles; de-aliased Doppler velocity [m s−1; color scale at bottom of (d)] at (d) 0.5°-, (e) 4°-, and (f) 6.4°-elevation angles. Arrows in (e),(f) indicate direction of Doppler velocities along and flanking the band seen in (b),(c), respectively. Range markers are shown every 20 km in (a). Dashed line in (c) indicates approximate location of leading edge of dry air at 2230 UTC, as seen in Fig. 14b. (g) The hodograph (green line) at Norman at 0000 UTC 31 May 2012 (from the Plymouth State College weather archive), below 7 km AGL. (h) Shear and horizontal vorticity vectors at 3–7 km based on the hodograph shown in (g) (top), and an idealized illustration of tilting of the environmental vorticity (vector shown by solid, curved line with arrow) between 3 and 7 km AGL onto the horizontal by gradients in vertical motion (dashed vectors) normal to the elevated band of convection (bottom).

  • View in gallery

    Radar imagery from KFDR depicting temporal evolution of dissipation of elevated convection when the air at the surface was drying and warming. Radar reflectivity factor (dBZ) at 2.4°-elevation angle [color scale at bottom of (a)]. Values at (a) 2221, (b) 2244, (c) 2303, and (d) 2322 UTC 30 May 2012. Range markers are shown every 20 km in (a). Dashed white lines encompass an area of precipitation that has dissipated, which builds in from the west and grows in area with time. {Bounded weak-echo regions (BWERs) are seen as holes in reflectivity in the northern and southern supercells [(a),(b)] and as crescent-shaped areas of reflectivity min in the southern supercell [(c),(d)]. The BWERs are evidence of strong updrafts.}

  • View in gallery

    Radar imagery from KTLX (Twin Lakes, Oklahoma) showing the area of decaying convection to the northwest, strong convection to the northeast, and multiple, parallel fine lines. Radar reflectivity [dBZ; color scale at bottom of (a)] at 0.5°-elevation angle at (a) 0458, (b) 0531, and (c) 0559 UTC 31 May 2012. (d) Doppler velocity [m s−1; color scale shown at bottom] at 0559 UTC. Locations of Ninnekah and Washington Oklahoma Mesonet sites (also Purcell Profiler) are indicated by an N and a W in (a). Arrows in (c) show new convective cells that have formed near-parallel fine lines. Range markers are shown every 15 km in (a).

  • View in gallery

    Radar imagery from KFDR showing an MCS with a cyclonic vortex to the southeast and parallel fine lines propagating southward across the radar site and apparently initiating new convective cells [cf. (c)]. Radar reflectivity factor [dBZ; color scale shown at the bottom of (a)] at 0.5°-elevation angle at (a) 0500, (c) 0601, and (d) 0702 UTC. (b) Doppler velocity (m s−1; color scale shown at the bottom) is shown at 0500 UTC. The cyclonic MCV in the MCS is noted by the arrow in (a) and the vortex signature is indicated by the solid circle in (b). The arrows in (c) and (d) point to new convective cells that have formed near-parallel fine lines. Range markers are shown every 50 km in (a).

  • View in gallery

    Meteograms at Oklahoma Mesonet sites, displaying both sudden warming and sudden drying after a gust-front passage. Data are plotted from 0330 to 0630 UTC 31 May 2012 (times displayed in a.m. and p.m. along the abscissa are shown in CDT; times displayed vertically are shown in UTC): (top) wind speed (m s−1; green), wind gusts (m s−1; red), and wind direction (°; purple); and (bottom) temperature (°C; red), dewpoint (°C; purple), and pressure reduced to sea level (hPa; green), at (a) Ninnekah, (b) Chickasha, (c) Ketchum Ranch, (d) Washington (site of Purcell profiler), (e) Norman, (f) Hobart, and (g) Weatherford. The onset of the gust front is indicated by an arrow at the time pressure began to rise suddenly; the warming and drying episodes are highlight by a double arrow.

  • View in gallery

    As in Figs. 20a–d, but for (a) Minco, (b) El Reno, (c) Oklahoma City East, and (d) Oklahoma City North, while the double arrow highlights warming and moistening episodes.

  • View in gallery

    As in Figs. 20e–g, but for typical gust-front passages characterized by both temperature and dewpoint falls, at (a) Guthrie, (b) Watonga, and (c) Kingfisher (there are no double arrows, only single arrows).

  • View in gallery

    The passage of the gust front at the NOAA Purcell wind profiler, very near the Washington Mesonet site. Data plotted are from 0206 to 0900 UTC 31 May 2012 (note that time increases toward the left): (a) wind speed (m s−1) and wind direction (wind barbs) and temperature [°C; with color code at bottom; warm (cool) temperatures are depicted by warm (cool) colors] and (b) vertical velocity (m s−1; with color code at bottom ). Shown in (a) is the time after which either there are no data, or the wind has shifted, at ~750 m AGL. Instances of updrafts (vertical velocity w > 0) and downdrafts (w < 0) are pointed out in (b). The 850-hPa level in a standard atmosphere is indicated at the right, around 1.5 km AGL. (c) Sounding at OUN at 0000 UTC 31 May 2012. (Courtesy of the Plymouth State College weather archive.)

  • View in gallery

    Depiction of the gust-front passage from data plotted from the Oklahoma Mesonet at (a) 0400, (b) 0415, (c) 0430, (d) 0445, (e) 0500, and (f) 0515 UTC 31 May 2012. Warm (cool) temperatures are coded as warm (cool) colors. Winds are plotted as wind barbs. Thick solid line is approximate location of the gust front. Thin solid lines are isobars of pressure reduced to sea level (hPa). Dashed contours enclose regions of calm wind. In (e), the N and W indicate locations of the Ninnekah and Washington (Purcell wind profiler) Mesonet sites, respectively.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 60 60 9
PDF Downloads 45 45 8

An Observational Study of the Effects of Dry Air Produced in Dissipating Convective Storms on the Predictability of Severe Weather

View More View Less
  • 1 School of Meteorology, University of Oklahoma, Norman, Oklahoma
  • | 2 Radar Research and Development Division, National Severe Storms Laboratory, Norman, Oklahoma
© Get Permissions
Full access

Abstract

This paper documents features that led to major forecast errors on the 12–24-h time scale in the nature and location of severe weather in the southern plains on 30 May 2012. Evidence is presented that the forecast errors were the result of 1) dry air that originated in a region of dissipating, elevated convective storms, and which was advected in a narrow tongue into western Oklahoma, inhibiting convective initiation; 2) the development of a cyclone along the dryline in western Texas, to the east of which several supercells formed; 3) the upscale development of the supercells into a mesoscale convective system (MCS) at nightfall; and 4) the dissipation of an MCS that had formed along a cold front in southwestern Kansas and was propagating into northwestern Oklahoma, as it encountered dry, subsiding air underneath the stratiform precipitation region of the rear portion of the MCS farther south. There was a meridionally oriented swath of high winds in clear air, in between the two MCSs. This swath of high winds may have been associated with a bore triggered at night by the MCSs approaching from the north, as the MCS collapsed, producing a gust front that propagated through stable, low-level air. This case study illustrates how the predictability of severe weather in a region can be extremely sensitive to the details of where nearby convective storms form and how they evolve. It also highlights the likely importance of the accurate representation of cloud microphysics and dynamics in numerical forecast models on predictability.

Corresponding author address: Howard B. Bluestein, School of Meteorology, University of Oklahoma, Ste. 5900, 120 David L. Boren Blvd., Norman, OK 73072. E-mail: hblue@ou.edu

Abstract

This paper documents features that led to major forecast errors on the 12–24-h time scale in the nature and location of severe weather in the southern plains on 30 May 2012. Evidence is presented that the forecast errors were the result of 1) dry air that originated in a region of dissipating, elevated convective storms, and which was advected in a narrow tongue into western Oklahoma, inhibiting convective initiation; 2) the development of a cyclone along the dryline in western Texas, to the east of which several supercells formed; 3) the upscale development of the supercells into a mesoscale convective system (MCS) at nightfall; and 4) the dissipation of an MCS that had formed along a cold front in southwestern Kansas and was propagating into northwestern Oklahoma, as it encountered dry, subsiding air underneath the stratiform precipitation region of the rear portion of the MCS farther south. There was a meridionally oriented swath of high winds in clear air, in between the two MCSs. This swath of high winds may have been associated with a bore triggered at night by the MCSs approaching from the north, as the MCS collapsed, producing a gust front that propagated through stable, low-level air. This case study illustrates how the predictability of severe weather in a region can be extremely sensitive to the details of where nearby convective storms form and how they evolve. It also highlights the likely importance of the accurate representation of cloud microphysics and dynamics in numerical forecast models on predictability.

Corresponding author address: Howard B. Bluestein, School of Meteorology, University of Oklahoma, Ste. 5900, 120 David L. Boren Blvd., Norman, OK 73072. E-mail: hblue@ou.edu

1. Introduction

Forecasting severe local storms and especially the type, location, and time of severe weather events associated with them is still a challenge, despite many years of progress in numerical weather prediction and in detection technology (e.g., Lilly 1990; Johns and Doswell 1992; Droegemeier 1997; Trapp et al. 2005; Weisman et al. 2008; Clark et al. 2010; Coniglio et al. 2010; Kain et al. 2013). Major challenges include defining the large-scale forcing and environment via an operational network, representing physical processes such as boundary layer turbulence and cloud microphysics in numerical models, and observing the prestorm mesoscale environment and the subsequent storms themselves with radar and surface and upper-air instruments with adequate spatial and temporal resolution so that they can be analyzed, interpreted, and assimilated into numerical models. Owing to the highly nonlinear nature of the governing equations representing the dynamics, thermodynamics, cloud microphysics, and the moisture budget, small changes in one or more variables can lead to rather different outcomes, especially on the 12–24-h time scale. As an example of the extreme sensitivity of a severe weather forecast to the details of where storms were initiated and how they evolved, we cite, as just one example, the 4 May 2007 Greensburg, Kansas, tornado, in which the actual tornadic supercell evolved from a long series of splitting supercells and their interactions with each other (Bluestein 2009).

The focus of this particular study stems from an experience the authors had on 30 May 2012, when participating in a field program whose main objectives were to collect data in tornadic supercells with a mobile Doppler radar (e.g., Pazmany et al. 2013). Our small field program overlapped with the larger Deep Convective Clouds and Chemistry Experiment (DC-3; http://www2.acd.ucar.edu/dc3), so that more data were available than usual. The morning forecasts from National Oceanic and Atmospheric Administration/Storm Prediction Center (NOAA/SPC) called for the relatively moderate likelihood of tornadoes and high surface winds from supercells and a subsequent mesoscale convective system (MCS) propagating southeastward across Oklahoma. Instead, there were tornadoes in northwestern and west-central Texas in supercells and a nonconvective windstorm at night in central Oklahoma. We experienced the latter as we returned from having collected data in one of the tornadic supercells.

The specific objectives of this paper are 1) to determine the sequence of meteorological features that led to the observed mesoscale weather events in Oklahoma and parts of Texas and 2) to identify which most likely contributed to the forecast errors. We will focus on the synoptic- and mesoscale patterns of wind, temperature, and moisture, and how small changes in the intensity and location of a synoptic-scale disturbance may have led, via a chain of events, to changes in the surface moisture field, which in turn had profound consequences for the accuracy of severe weather forecasts. It will be shown that regions of dry air at the surface, which were produced in areas of subsidence underneath dissipating portions of MCSs, were of paramount importance.

The more general objective is to illustrate how the predictability of severe weather can be extremely sensitive to the details of when and where convective storms form and evolve. This case is rich in the variety of convective phenomena observed and is therefore particularly illustrative.

The nature of the data used is described in section 2. The forecasts from the SPC and the shorter-term fore-/nowcasts from the National Weather Service (NWS), the synoptic-scale environment, and the evolution of convection prior to the severe weather events are discussed in section 3. Section 4 contains a description of the severe weather events and the synoptic- and mesoscale influences on them. The causes and effects of the evolution of the meteorological features and the implications of these are summarized in section 5.

2. Data

a. Operational network

Most of the data used in our study came from the operational, national network of surface-observing and rawinsonde sites and from the Weather Surveillance Radar-1988 Doppler (WSR-88D) network (Crum and Alberty 1993). These data were accessible in real time online from the University Corporation for Atmospheric Research/National Center for Atmospheric Research/Research Applications Laboratory (UCAR/NCAR/RAL) and later from the SPC archive and the National Climatic Data Center (NCDC). Some surface data were obtained later online from the Oklahoma Mesonet (Fig. 1; Brock et al. 1995; Fiebrich and Crawford 2001) and the West Texas Mesonet (Schroeder et al. 2005). Doppler wind profiler data were obtained later online from the NOAA archive of the National Profiler Network (Weber et al. 1990; http://www.profiler.noaa.gov/npn/index.jsp). Satellite imagery was obtained in real time from NCAR/Research Applications Laboratory (RAL) and the College of DuPage NeXt Generation Weather Laboratory. Plymouth State College Weather Center provided some surface data later online. Some model output data and composite radar imagery were obtained later on from the DC-3 archive maintained by NCAR/Earth Observing Laboratory (EOL).

Fig. 1.
Fig. 1.

Map of stations in the Oklahoma Mesonet in central, south-central, southwestern, and western Oklahoma. Data from many of these are alluded to in the text. The stations enclosed within the polygon exhibited similar behavior when a gust front passed by, as discussed later in the text.

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

b. Special instrumentation

Mobile Doppler radar from RaXPol, a rapidly scanning, mobile, polarimetric, X band (3-cm wavelength), were used (Pazmany et al. 2013). This radar scans mechanically around a complete circle as rapidly as 180° s−1, making use of frequency hopping to collect enough independent samples to get good estimates of Doppler velocity and other variables. Its half-power beamwidth is ~1°, which allows it to resolve finescale features.

3. Operational forecasts, the synoptic-scale environment, and the evolution of convection prior to the severe weather events

a. Forecasts from the SPC

The “convective outlook” issued by the SPC shortly before 1300 UTC 30 May 2012 called for a “moderate” risk of severe thunderstorms over western Oklahoma and an adjacent portion of southwestern Kansas (Fig. 2a). This area was surrounded by a “slight” risk covering north-central Texas, the eastern portion of the Texas Panhandle, and western Kansas, among other adjacent areas. The issuance of a moderate risk is indicative of the expectation of an enhanced threat of severe weather. The tornado potential forecast was for a 10% chance of a tornado within 25 mi of any location within an area of northwestern Oklahoma and an adjacent portion of southwestern Kansas (Fig. 2b). A 10% probability also is indicative of an enhanced threat. The 5% probability contour surrounding the 10% probability contour covered the eastern portion of the Texas Panhandle and northwestern Texas, among other areas. The SPC discussion (http://www.spc.noaa.gov/products/outlook/archive/2012/day1otlk_20120530_1300.html) noted that “Considerable variability exists amongst mesoscale and convection-allowing model guidance with regard to specific details of convection evolution…” but that “storm initiation will be across western parts of Kansas or Oklahoma by mid afternoon….” Based on forecast hodographs, it was anticipated there could be “rapid supercell evolution” and the “potential for widespread damaging winds…” after “the activity grows upscale into one or more…MCSs.”

Fig. 2.
Fig. 2.

Specialized forecast and verification products from the SPC on 30 May 2012. (a) The “day 1” convective outlook by category issued at 1234 UTC. (b) As in (a), but for the probability of tornadoes within 25 mi (60 km) of any point. (c),(d) As in (a),(b), but for the convective outlook issued at 1630 UTC. (e) Filtered storm reports for the day. Locations of tornadoes are in red, along with high-wind [blue squares >50 knots (kt; 1 kt = 0.51 m s−1; therefore, 50 kt = 25.8 m s−1); black squares >65 kt (33.5 m s−1)] and large-hail [green triangles >1 in. (2.5 cm) in diameter; black triangles >2 in. (5.1 cm) in diameter] reports. (f) Max surface wind gusts [mi h−1; 2.2(m s−1)] in the hour prior to 0510 UTC 31 May 2012. The solid line encloses reports of wind gusts (red) in excess of 60 mi h−1 (27.3 m s−1); the dashed line encloses reports of wind gusts (orange) in excess of 40 mi h−1 (18.2 m s−1). (g) As in (f), but for wind gusts in excess of 60 mi h−1 (27.3 m s−1) at 0500–1330 UTC to show that there were sporadic high-wind reports later at night, but these were not associated with the bore to be discussed in the text.

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

The convective outlook issued at around 1630 UTC was similar to the early morning (1234 UTC outlook) forecast, except that the areas of moderate and slight risk were larger (Fig. 2c). The forecast issued at around 2000 UTC for the area of 10% probability of a tornado within 25 mi of any location, however, was removed (Fig. 2d), owing to the expectation, based on up-to-date observations, that the mode of convection would be an MCS, rather than supercells. The later outlook also mentioned, for the first time, the possibility of storms along the dryline, as a “secondary N–S zone of thunderstorms…just west of influential outflow from early morning convection.”

In actuality, severe weather events in Oklahoma consisted almost entirely of high-wind events, while tornadoes were observed only along the southwestern fringe of the 5% probability of tornadoes contour and far to its south (Fig. 2e). It will be shown later that many of the high-wind events in Oklahoma did not occur in areas of convective precipitation.

b. Synoptic-scale environment: Forecast and observed

At 1200 UTC there was a trough oriented from east-northeast to west-southwest in the midtroposphere (500 hPa) embedded within northerly flow over Southern California and southern Nevada (Fig. 3a), downstream from which there was a broad westerly current of 10–20 m s−1 flow over Arizona and New Mexico. There was a hint of a short-wave trough at the Arizona–New Mexico border, as evidenced by the slightly backed wind at Albuquerque, New Mexico. A Q-vector analysis computed from the observed wind and temperature fields at 500 hPa (not shown) showed convergence in New Mexico, which supports the existence of this short-wave trough. In the upper troposphere (250 hPa), however, only the trough over California and Nevada was discernible (Fig. 4a). A jet streak with winds of 40–45 m s−1 was located downstream from the trough. The diffluent jet-exit region was situated near the New Mexico–western Texas border.

Fig. 3.
Fig. 3.

The 500-hPa synoptic maps for the continental United States at (a) 1200 UTC 30 May and (b) 0000 UTC 31 May 2012. (Prepared by the SPC.) Trough axes in the southwestern portion of the United States are marked by brown-dashed segments. The arrow in (b) points to a narrow jet marked by 50–55 kt (25.8–28.4 m s−1) wind observations near the New Mexico–southwestern Texas border. (c) As in (b), but for a 12-h forecast from the operational 12-km NAM initialized at 1200 UTC 30 May and valid at 0000 UTC 31 May 2012.

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

Fig. 4.
Fig. 4.

As in Figs. 3a,b, but for 250 hPa. An area of strong CVA at the left-front quadrant of a jet streak and downstream from a zonally oriented trough axis is noted over southwestern Oklahoma and northwestern Texas in (b).

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

The 12-h operational 12-km North American Mesoscale Forecast System (NAM; http://www.ncdc.noaa.gov/data-access/model-data/model-datasets/north-american-mesoscale-forecast-system-nam) forecast for 500 hPa at 0000 UTC 31 May predicted that the main (longer wave) trough would remain over Southern California, while the short-wave trough to the east and northeast would progress eastward to near the New Mexico–western Texas border (Fig. 3c). Wind speeds of 5–15 m s−1 at 500 hPa were forecast into the Texas Panhandle. Wind speeds as high as 25 m s−1 were forecast only upstream and south of the trough axis, in northern Mexico. The actual 500-hPa wind speeds were as high as 25–27.5 m s−1 at the base of the short-wave trough and the winds to the north at Amarillo, Texas, were only ~7.5 m s−1 (Fig. 3b) and temperature advection at the surface (see Fig. 7, described in greater detail below, for the surface at 1843 UTC) was weak; we therefore infer that cyclonic vorticity advection (CVA) in the midtroposphere was greater over the Texas Panhandle than suggested by the NAM 12-h forecast, so it may be inferred that the actual synoptic-scale, quasigeostrophic upward forcing due to differential vorticity advection (temperature advection and vorticity advection at low levels were weak) was probably also greater. Vorticity advection was becoming more cyclonic with height (from ~2 × 10−9 to 5 × 10−9 s−2 at 500 hPa and from ~−1 × 10−9 to −2 × 10−9 s−2 at 700 hPa) and weak temperature advection (close to zero) was computed using observed wind and temperature data at 850, 700, and 500 hPa. A Q-vector analysis at 500 hPa indicated only weak convergence in the southern Texas Panhandle and northwestern Texas (not shown). At 0000 UTC the jet streak at 250 hPa had propagated eastward, so that the exit region was located just over the eastern portion of the Texas Panhandle and the southern Texas plains (Fig. 4b). The left-front quadrant of this jet streak, where there was strong CVA (from ~5 × 10−9 to 10 × 10−9 s−2), also was located in the general area just southwest of the southwestern tip of Oklahoma and into much of southern Oklahoma. We conclude that synoptic-scale forcing, albeit weak, was upward during the late afternoon and early evening over the eastern portion of the Texas Panhandle, but possibly stronger than had been forecast by the NAM.1

c. Evolution of convection prior to the severe weather events

During the late afternoon and early evening of 29 May 2012 an isolated supercell formed in central Oklahoma and this and neighboring cells eventually developed upscale into an MCS that propagated southward across Oklahoma during the night (Fig. 5) and dissipated the following day. By sunrise, only scattered, small, convective cells lingered over northeastern Texas while other small cells developed over Oklahoma above the cold pool left behind in the wake of the MCS. At 1800 UTC 30 May the only precipitation in the region detected by radar was found over northeastern Texas and the cells it was associated with were dissipating.

Fig. 5.
Fig. 5.

WSR-88D reflectivity constant altitude plan position indicator (CAPPI) composites for the region centered at 12 000 ft (3.7 km) over Oklahoma at (a) 0300, (b) 0600, (c) 0900, (d) 1200, (e) 1500, and (f) 1800 UTC 30 May 2012. The highest (lowest) reflectivities are shown as warm (cool) colors. The solid lines represent the 15°C isodrosotherm. In (f), the arrow points to a dissipating area of elevated convection that has disappeared by 2100 UTC. (Courtesy of the UCAR/EOL/DC-3 website.)

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

4. Description of severe weather events and the influences on them

a. Mesoscale dry tongue

At the same time that the aforementioned area of precipitation was dissipating, a narrow swath of relatively dry air at the surface appeared over northeastern Texas (Fig. 6); furthermore, in Figs. 5e and 5f it is seen that the appearance of dry air (surface dewpoints <15°C) occurred along the western and eastern flanks of and inside of the region of dissipating convection. The dewpoint temperatures of ≤14°C in the swath highlighted in Fig. 6 could not be accounted for by horizontal advection because surrounding dewpoint temperatures had generally been ≥14°C. We therefore hypothesize that the dry air was caused by subsidence from the dissipating remnants of the old MCS. This dry air may have been produced through physical processes occurring toward the rear of squall lines, similar to those first described by Zipser (1977) and later by Smull and Houze (1987) and Weisman (1992), in which horizontal pressure gradients associated with variations in latent heat release aloft and evaporative cooling below drive a rear-inflow jet carrying dry air into the stratiform precipitation region under the anvil and then downward; in our case, however, the mesoscale downdraft seems to have penetrated all the way to the surface, since the surface air was so dry. Rawinsonde observations were not available to test this hypothesis and we can only note the coincidence between the appearance of dry air and the dissipating MCS. The dry air (for clarity, in this case dewpoint temperatures ≤15.6°C) was then advected north-northwestward in a narrow band up into western Oklahoma (Figs. 7 and 8). Backward trajectories were computed for up to 12 h for portions of western and southwestern Oklahoma ending at 0000 UTC 31 May 2012, using the NOAA Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model (https://ready.arl.noaa.gov/HYSPLIT.php) to demonstrate that the air in western and southwestern Oklahoma originated in north-central and northwestern Texas at 1200 UTC. In Fig. 7 it is seen that the dry air that had appeared between 1500 and 1800 UTC (Figs. 5e,f) was subject to moderate south-southeasterly winds, some in excess of 15 m s−1. In Fig. 9a the clouds associated with the dissipating convection at 1915 UTC, and a band of clear air to the west and southwest that curves northward into southwestern Oklahoma, are seen. It is possible there could have also been subsidence just outside the area of dissipating elevated convection. The convection must have been elevated because the surface temperature behind the outflow boundary associated with the earlier convection (~26°C) was well below the convective temperature (~37°C based on the sounding at 1200 UTC in Norman, Oklahoma), and the convection was not along a surface boundary (Figs. 5 and 7). The main point is that the drying most likely must have been due to subsidence of dry air aloft and that dry air was also being advected in a narrow band, the width of just ~50–100 km, through western Oklahoma. Other factors that may have played a role, however, are variations in the depth of the moist boundary layer (perhaps due to subsidence itself) and variations in evapotranspiration due to changes in vegetation type and previous rainfall.

Fig. 6.
Fig. 6.

Surface dewpoint temperature (°C) in the south-central United States at (a) 1500, (b) 1600, (c) 1700, and (d) 1800 UTC 30 May 2012. The solid line is the 15°C isodrosotherm. Warm (cool) colors denote high (low) dewpoints. Five areas of drying are identified: A is a dry air mass to the rear of the dryline; B is a dry air mass in Kansas, north of the returning moisture in Oklahoma; C is a dry pocket in far southwestern Arkansas, northeastern Texas, and southeastern Oklahoma; D is a dry pocket in northwestern Texas, extending just over the border into south-central/southwestern Oklahoma; and E is a dry pocket in west-central Oklahoma.

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

Fig. 7.
Fig. 7.

Regional surface map centered on Oklahoma at 1843 UTC 30 May 2012. Solid contour is the 60°F (15.6°C) isodrosotherm. The dashed red lines mark the dryline in the southern Texas plains and a trough axis in northeastern Colorado and the Nebraska Panhandle. The dashed blue lines mark a cold front in western Texas and eastern New Mexico and an outflow boundary in central Kansas. The solid blue line at the border between northeastern Colorado and the Nebraska Panhandle marks a cold front. The cross is located at the center of a cyclone at the northern extent of the dryline. The arrow just east of the cyclone denotes the surface wind direction. To the right of the arrow, the surface temperatures are in the mid-80°F (~29°–30°C); to the left, temperatures are in the upper 90°F (~38°C). Therefore, there is a significant component of the temperature-gradient vector directed normal to and to the left of the surface wind. (Courtesy of UCAR/NCAR/RAL.)

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

Fig. 8.
Fig. 8.

Hourly surface weather observations at the Oklahoma Mesonet sites at (a) 1330, (b) 1430, (c) 1530, (d) 1630, (e) 1730, (f) 1830, (g) 1930, and (h) 2030 UTC 30 May 2012. Winds are plotted according to convention; dewpoint temperatures are plotted such that the lowest (highest) dewpoints are shaded in cool (warm) colors. The white rectangles in the Texas Panhandle cover up a data-void region. Dry regions labeled in Fig. 6 are noted.

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

Fig. 9.
Fig. 9.

Visible Geostationary Operational Environmental Satellite (GOES) imagery for the central United States at (a) 1915 and (b) 2132 UTC 30 May 2012 (from UCAR/NCAR/RAL). (c) As in (a),(b), but for a close-up view of portions of the Texas Panhandle, western Texas, southwestern and south-central Oklahoma, and northwestern Texas, at 1945 UTC (from the College of DuPage NeXt Generation Weather Laboratory). (d) Photograph of stratus cloud bands in southwestern Oklahoma, looking to the north, at ~2100 UTC (courtesy of H. Bluestein). In (a),(b), regions are identified: A shows clouds associated with the front in southwestern Nebraska, B is the trough axis in Colorado (cf. Fig. 7), C is the outflow boundary in central Kansas (cf. Figure 7), and D is the dryline in Texas (cf. Fig. 7); the dashed red line marks the axis of dry air at the surface (cf. Fig. 7). In (b), arrows point to convective storms along the front in Nebraska (A), the trough axis in northwestern Kansas (B; which earlier had been in Colorado), the outflow boundary in central Kansas (C), and the dryline in Texas (D). The arrow in (c) points to the bands of stratus seen in (d).

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

b. Storm initiation and evolution

There were four areas of potential convection initiation in and upstream from the southern plains at 1900 UTC (Figs. 7 and 9a): a front in extreme northeastern Colorado that was collocated with a diffuse band of clouds (A in Fig. 9a), a trough in eastern Colorado collocated with a narrow band of shallow convective clouds (B in Fig. 9a), an outflow boundary in central Kansas associated with developing convective storms (C in Fig. 9a), and a north–south-oriented band of shallow convective clouds in the Texas Panhandle and southern Texas plains near the dryline (D in Fig. 9a), the northern edge of which was marked by a cyclone. To the north of the cyclone there was a trough in easterly flow and to the west-southwest of the cyclone there was a weak cold front.

Storms began along the dryline at ~1945 UTC (Fig. 9c). East of the dryline there were long quasi-parallel bands of stratus (Figs. 9c,d), evidence of which was detected by the WSR-88D at Frederick, Oklahoma (KFDR; Fig. 10a); these bands are thought to be indicative of a stable boundary layer [Scofield and Purdom (1986), p. 131, their Fig. 7.9a]. This assertion is supported in part by the cooler air observed in southwestern Oklahoma and northwestern Texas rather than farther to the west just east of the dryline (Fig. 7). A clear zone or one marked by shallow convective clouds was seen in between the deep, convective storms forming along the dryline and the bands of stratus in the stable boundary layer across southwestern Oklahoma and northwestern Texas. By 2132 UTC deep convection, as evidenced by extensive anvils, was in progress along the front in southwestern Nebraska, along the trough in northwestern Kansas, along the outflow boundary in central Kansas, and along the dryline in the Texas Panhandle and the southern Texas plains (Fig. 9b). The areas of developing storms in central and western Kansas were anticipated by the SPC, but the area of storms along the dryline was not forecast until the early afternoon.

Fig. 10.
Fig. 10.

Evolution of convective storms that formed along the dryline into an MCS. (a)–(d) Broader and (e)–(g) zoomed-in views of the tornadic supercell. Radar reflectivity factor [dBZ; color code shown in (a)] from the KFDR (Frederick, Oklahoma) WSR-88D at (a) 2200 and (b) 2300 UTC 30 May; and at (c) 0000, (d) 0100, and (e) 2315 UTC 31 May 2012, at 0.5°-elevation angle. Range markers are shown every 25 km in (a)–(d). The southernmost cell was a tornadic supercell in (b),(c). In (e), the arrow points to a supercell that subsequently produced a tornado. Range markers are shown every 15 km in (e). (f) Radar reflectivity factor (dBZ; color code shown) at 1.9°-elevation angle from RaXPol at 2317 UTC 30 May 2012. WEH is the weak-echo hole associated with the tornado. (g) As in (f), but for Doppler velocity (m s−1; color coded); the circle encloses the cyclonic vortex signature marking the location of the tornado. Range markers are shown every 5 km in (f),(g).

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

The broken line of storms, oriented in the north–south direction, which developed along the dryline, evolved into isolated supercells at its southern end and propagated southeastward into northwestern Texas; over the next 3 h, the supercells and surrounding cells developed into an MCS (Fig. 10). While two high-precipitation supercells (Moller et al. 1994) each produced tornadoes, the southernmost supercell produced the larger and longer-lived mesocyclone (Figs. 10b and 10e–g). Weaker, convective cells formed underneath the anvils of the cells that had developed along the dryline (Figs. 10a,b). These convective cells were probably elevated because the surface temperature well to the east of the dryline (~25°–27°C) was well below the convective temperature (~37°C based on the Norman sounding at 1200 UTC) and there were no obvious surface convergence zones; visually, the convection looked to be relatively high based. Another MCS developed in Kansas and propagated southward, but the western end of this MCS decayed between 0300 and 0500 UTC, as it progressed through northwestern Oklahoma (Fig. 11).

Fig. 11.
Fig. 11.

Regional evolution of two MCSs at (a) 0305 and (b) 0501 UTC 31 May 2012, showing the dissipation of the western portion of the northern MCS. Oklahoma City is indicated by OKC. The region of dry air in western Oklahoma is noted. Composite low-level radar reflectivity factor (dBZ; color coded). (Courtesy of UCAR/NCAR/RAL.)

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

The operational 12-km NAM, initialized at 1200 UTC, which makes use of parameterized convection, correctly forecast convection initiation along the dryline (arrow in Fig. 12b). The High Resolution Rapid Refresh (HRRR) model (http://rapidrefresh.noaa.gov/), which explicitly simulates convection, on the other hand, while also forecasting convection initiation along the dryline, did not forecast the MCS that subsequently developed in northwestern Texas, but did forecast an MCS propagating southward from Kansas into western, central, and eastern Oklahoma (Fig. 13); the western edge of this MCS, however, did not decay as it propagated through western Oklahoma as was observed. So, neither the NAM nor the HRRR accurately forecast the formation/behavior of the northern MCS (which formed in Kansas) and the HRRR did not accurately forecast the southern MCS (which formed in northwestern Texas).

Fig. 12.
Fig. 12.

Simulated low-level radar reflectivity factor [dBZ; warm (cool) colors indicate the highest (lowest) reflectivity] from the 1200 UTC operational 12-km NAM for the U.S. southern plains at (a) 2100 UTC 30 May and at (b) 0000, (c) 0300, and (d) 0600 UTC 31 May 2012. Convective storms are not explicitly simulated. Arrow in (b) points to convective storms that had formed along the dryline.

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

Fig. 13.
Fig. 13.

As in Fig. 12, but for composite reflectivity (dBZ; color coded) at hourly intervals over the U.S. southern plains and environs from the 1800 UTC HRRR model at (a) 2000, (b) 2100, (c) 2200, and (d) 2300 UTC 30 May, and at (e) 0000, (f) 0100, (g) 0200, and (h) 0300 UTC 31 May 2012.

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

c. Mesoscale dry area from elevated convection

By late afternoon, the earlier, narrow corridor of dry air (≤14°C dewpoints) in western Oklahoma at the surface had disappeared, while another mesoscale region of even drier air at the surface (~6°–10°C dewpoints) appeared in far southwestern Oklahoma (Fig. 14) at 2230 UTC. (Though not depicted in Fig. 14, such low dewpoints were not observed nearby in the West Texas Mesonet; not plotted). Like the earlier area of dry air, horizontal advection could not have accounted for its appearance because there was no dry air nearby and upstream.

Fig. 14.
Fig. 14.

As in Fig. 8, but for hourly surface weather observations at (a) 2130, (b) 2230, and (c) 2330 UTC 30 May, and at (d) 0030, (e) 0130, (f) 0230, (g) 0330, and (h) 0430 UTC 31 May 2012. Arrows in (b) point to locations likely affected by outflow, in which the winds are from the east, not the south, as they are at surrounding sites.

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

The Mesonet sites at Altus, Mangum, Retrop, and Erick in southwestern and western Oklahoma experienced the onset of surface drying just prior to 2230, at 2230, just after 2300, and just before 2330 UTC, respectively (Fig. 15). At Altus, the drying was sudden and accompanied by a jump in temperature and rapid change in wind direction from southeasterly to northerly and then back to east-southeasterly. There was no obvious drop in pressure as is usually noted during heat burst events [e.g., Johnson (1983); Johnson et al. (1989); and Bluestein (2013), p. 130, his Fig. 3.27]. At Mangum, the traces of temperature, dewpoint, and wind were similar to those at Altus, but there was only a small shift in wind direction. At Retrop, there was drying, but no concomitant well-defined change in temperature. In this case, however, there was a drop in pressure consistent with wake lows at the rear of MCSs (Johnson et al. 1989; Bluestein 2013), but no change in wind direction. Finally, at Erick, the traces were similar to those at Retrop. The event seemed to be captured during its onset at Altus and Mangum and during its more mature stage at Retrop and Erick. Johnson et al. (1989) hypothesized that the cold pool from evaporated precipitation becomes transformed into a wake low when stratiform precipitation from an MCS dissipates. The drying event may therefore have been due to local sinking motion. However, it is also noted that there was little if any change in temperature during the mature stage of the event, when one would have expected warming, perhaps because warming from sinking motion was being countered by evaporative cooling.

Fig. 15.
Fig. 15.

Meteograms (traces of meteorological variables), displaying sudden drying in southwestern and western Oklahoma from the late afternoon to early evening. Data are plotted from 2000 UTC 30 May to 0700 UTC 31 May 2012 (times displayed in a.m. and p.m. along the abscissa are shown in CDT; times displayed vertically are shown in UTC): (top) wind direction (°; red) and wind speed (m s−1; blue); and (bottom) temperature (°C; red), dewpoint (°C; blue), and pressure (hPa; green) at Oklahoma Mesonet sites (a) Altus, (b) Mangum, (c) Retrop, and (d) Erick. In (a),(b), the double arrow points out a short-lived warming and drying episode; in (c),(d), the double arrow points out a drying episode, which persisted for hours. Short single arrows show the beginning of pressure rise, which marked late-evening gust-front passage.

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

Evidence that stratiform precipitation was present comes not only from Figs. 10a and 10b, but also from observations of the precipitation rate. The area of drying was coincident with the measurement of small amounts of precipitation at Oklahoma Mesonet sites of only ~1–8 mm over a 2-h period (from 2200 to 0000 UTC), that is, ~0.5–4 mm h−1; this precipitation rate is characteristic of stratiform precipitation from an anvil, rather than from deep, active convection (e.g., Leary and Houze 1979) [rainfall rates of <2.5 mm h−1 are considered “light rain”; rates of 2.5–10 mm h−1 are considered “moderate” rain (OFCM 1995)]. The first appearance of drying in extreme southwestern Oklahoma was also at 2230 UTC. It is therefore hypothesized that the area of dry air at the surface was caused by subsidence from decaying, elevated convective cells (Figs. 16a–c and 17).

Fig. 16.
Fig. 16.

Radar imagery from KFDR at 2219–2222 UTC 30 May 2012, for southwestern Oklahoma, when episodes of warming and drying were beginning at Altus and Mangum [locations indicated in (a) and (c)] but before drying was observed at Retrop [locations indicated in (a) and (c)]. Radar reflectivity factor [dBZ; color scale at bottom of (a)] at (a) 0.5°-, (b) 4°-, and (c) 6.4°-elevation angles; de-aliased Doppler velocity [m s−1; color scale at bottom of (d)] at (d) 0.5°-, (e) 4°-, and (f) 6.4°-elevation angles. Arrows in (e),(f) indicate direction of Doppler velocities along and flanking the band seen in (b),(c), respectively. Range markers are shown every 20 km in (a). Dashed line in (c) indicates approximate location of leading edge of dry air at 2230 UTC, as seen in Fig. 14b. (g) The hodograph (green line) at Norman at 0000 UTC 31 May 2012 (from the Plymouth State College weather archive), below 7 km AGL. (h) Shear and horizontal vorticity vectors at 3–7 km based on the hodograph shown in (g) (top), and an idealized illustration of tilting of the environmental vorticity (vector shown by solid, curved line with arrow) between 3 and 7 km AGL onto the horizontal by gradients in vertical motion (dashed vectors) normal to the elevated band of convection (bottom).

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

Fig. 17.
Fig. 17.

Radar imagery from KFDR depicting temporal evolution of dissipation of elevated convection when the air at the surface was drying and warming. Radar reflectivity factor (dBZ) at 2.4°-elevation angle [color scale at bottom of (a)]. Values at (a) 2221, (b) 2244, (c) 2303, and (d) 2322 UTC 30 May 2012. Range markers are shown every 20 km in (a). Dashed white lines encompass an area of precipitation that has dissipated, which builds in from the west and grows in area with time. {Bounded weak-echo regions (BWERs) are seen as holes in reflectivity in the northern and southern supercells [(a),(b)] and as crescent-shaped areas of reflectivity min in the southern supercell [(c),(d)]. The BWERs are evidence of strong updrafts.}

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

During the beginning of the period of drying (on and after 2230 UTC; Figs. 14a–c) a radar signature (Figs. 16e,f; valid at ~2220 UTC) was observed just southwest of the Retrop, Mangum, and Altus Mesonet sites (cf. Figs. 1, 14a–c, and 16) that the authors do not believe has been described previously in the literature: a band of elevated convection was coincident with a very narrow zone (~5 km) of receding Doppler velocities, flanked by wider zones of approaching Doppler velocity, the strongest being on the southern side of the highest radar reflectivity factor (Figs. 16e,f). This radar signature was detected at midlevels, up to at least 6–7 km AGL. The close proximity (in time) between the signature and the appearance of drying at the surface could possibly be significant. Along the southern edge of the signature there is strong cyclonic shear vorticity, at least with respect to the Doppler velocity component; there is anticyclonic shear on the northern side, but it is weaker, owing to the weaker approaching Doppler velocities. This cyclonic–anticyclonic couplet, elongated in the west-northwest–east-southeast direction, is qualitatively consistent with tilting of environmental vorticity, which from the 0000 UTC 31 May 2012 sounding at Norman (nearest in time and space to the event) is directed toward the northeast (the vertical shear is directed approximately toward the southeast above ~3 km AGL; Figs. 16g,h) by an updraft aloft (above 3 km AGL) within the band of radar echo associated with the elevated convection or [cf., e.g., Fig. 4.29, p. 211, in Bluestein (2013) and Fig. 16h] with tilting by downdrafts straddling the precipitation-laden region aloft. This inference is highly speculative, however, because we cannot measure vertical velocity and we are neglecting any contribution to vorticity from the component of wind normal to the radar beam. That there is no vortex signature in the low, 0.5°-elevation angle scan (Fig. 16d) is consistent with tilting being more significant aloft where vertical velocities are expected to be stronger. How the production of a banded zone of a couplet of cyclonic–anticyclonic vorticity might be related to surface drying, however, is not clear. It may be that the drying resulted from the downward advection of dry air in the southwestern of the two inferred flanking downdrafts. The same tilting mechanism shown in Fig. 16h could have occurred in the absence of a downdraft on the northeastern side of the band.

The region of dry air eventually was advected northward through portions of western Oklahoma, but not the warmer air. The dry air and its advection were not forecast by the HRRR. The 2221 UTC Mesoscale Convective Discussion (MCD 0991) issued by the SPC (http://www.spc.noaa.gov/products/md/2012/md0991.html) noted that “curiously…the RAP based objective analysis has stubbornly kept a very strong axis of instability over the eastern Texas Panhandle into western Oklahoma this evening…despite hours of cool and dry surface observations from earlier outflow. Modifying the CDS (Childress, Texas) sounding for current conditions yields perhaps 100 J kg−1 surface-based CAPE…while the RAP model suggests there should be 2500 J kg−1…The highest threat area overall is from central Oklahoma into northeastern Oklahoma where more vigorous cells are keeping up with the gust front.”

d. A nonconvective windstorm and a bore and solitary wave train

The MCD continued, “For areas from western Oklahoma into the Texas Panhandle…there is little additional activity being generated behind the advancing outflow…which will result in a gradual weakening of these winds if new storms do not develop” (italics added by the authors).

In fact, although the MCS approaching from the north did weaken in northwestern Oklahoma as a result of the dry surface air, the winds ahead of the MCS strengthened rather than weakened (Fig. 2f), as had been forecast. A swath of surface winds of 30–37 m s−1 was observed northwest, west, and southwest of Oklahoma City, aligned in the north–south direction; this swath was surrounded by a broader swath of winds of 20–26 m s−1, aligned in the northeast–southwest direction. The region of the high winds in western Oklahoma did not occur in convective storms (cf. Figs. 2f and 11b), contrary to expectations. The NWS in Norman, ironically, issued a severe thunderstorm warning at 2248 UTC for portions of 11 counties in central Oklahoma, even though there were no storms in much of the area that experienced the high winds west of Oklahoma City (Fig. 11b). It must have seemed strange to the public, as it was to us, to hear a severe thunderstorm warning when there was no thunder or lightning nearby and no heavy precipitation.

The broad area of high winds was in fact associated with the passage of a gust front produced by the northern MCS. To the south and southwest of the highest radar reflectivity, the leading edge of the gust front was characterized not only by a fine line, but also by parallel, curved bands of higher reflectivity (Figs. 18a,b), which bear a strong resemblance to those seen in bores with trailing solitary wave trains [e.g., Koch et al. (2008), their Fig. 4, p. 1274]. The fine lines, perhaps associated with a bore and solitary wave train, continued to propagate to the south, approaching the southern MCS, which at 0500 UTC had a mesoscale convective vortex (MCV) on its northeastern-most end, as evidenced by a cyclonic-shear signature and counterclockwise spiral in reflectivity (Figs. 19a,b). As the fine lines neared the southern MCS around 0600 UTC, new precipitation cells formed coincident with the fine lines (Figs. 19c,d). It thus appears that the nocturnal bore and solitary wave train initiated precipitation (e.g., Marsham et al. 2011). The new cells appeared in between the fine lines, suggesting that there were alternating rising and sinking bands associated with the fine lines (Figs. 18c,d and 19c,d).

Fig. 18.
Fig. 18.

Radar imagery from KTLX (Twin Lakes, Oklahoma) showing the area of decaying convection to the northwest, strong convection to the northeast, and multiple, parallel fine lines. Radar reflectivity [dBZ; color scale at bottom of (a)] at 0.5°-elevation angle at (a) 0458, (b) 0531, and (c) 0559 UTC 31 May 2012. (d) Doppler velocity [m s−1; color scale shown at bottom] at 0559 UTC. Locations of Ninnekah and Washington Oklahoma Mesonet sites (also Purcell Profiler) are indicated by an N and a W in (a). Arrows in (c) show new convective cells that have formed near-parallel fine lines. Range markers are shown every 15 km in (a).

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

Fig. 19.
Fig. 19.

Radar imagery from KFDR showing an MCS with a cyclonic vortex to the southeast and parallel fine lines propagating southward across the radar site and apparently initiating new convective cells [cf. (c)]. Radar reflectivity factor [dBZ; color scale shown at the bottom of (a)] at 0.5°-elevation angle at (a) 0500, (c) 0601, and (d) 0702 UTC. (b) Doppler velocity (m s−1; color scale shown at the bottom) is shown at 0500 UTC. The cyclonic MCV in the MCS is noted by the arrow in (a) and the vortex signature is indicated by the solid circle in (b). The arrows in (c) and (d) point to new convective cells that have formed near-parallel fine lines. Range markers are shown every 50 km in (a).

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

The highest surface (actually on an instrumented tower, but no higher than 10 m AGL) wind speed reported by a station in the Oklahoma Mesonet (or at NWS sites) was 37 m s−1 at Ninnekah at 0500 UTC (see Fig. 1 for a map of the Oklahoma Mesonet sites in southwestern Oklahoma). Just 2 min earlier, at 0458 UTC, Ninnekah was located just to the rear of the leading fine line, but ahead of subsequent fine lines (Fig. 18a). The winds prior to ~0450 UTC were <5 m s−1 (Fig. 20a). At the same time that the wind speed dramatically increased, the pressure also rose rapidly ~4.5 hPa over a ~20-min period and then leveled off and decreased slowly. The traces of wind and pressure are consistent therefore with the leading fine line marking the passage of a gust front (e.g., Goff 1976; Wakimoto 1982). However, following the passage of the leading fine line, the surface temperature increased several degrees Celsius while the dewpoint decreased as it often does to the rear of a gust front. This atypical behavior of rapid warming and drying followed by more traditional gust-front behavior was also observed at the following other nearby Mesonet sites: Chickasha, Ketchum Ranch, Washington, and Norman (Figs. 20b–e). At all these other sites, the surface wind was also weak (<5 m s−1) prior to the abrupt increase in wind speed. The increase in temperature and drop in dewpoint also lagged the increase in wind speed that marked the gust-front passage. Similar behavior was also noted at Hobart, to the west, and, to a lesser extent, at Weatherford, also to the west (Figs. 20f,g). This behavior (i.e., increase in temperature and drop in dewpoint), however, was not observed at the surrounding sites of Paul’s Valley to the south, Byars or Shawnee to the east (not shown), nor at any of the sites to the north, such as Oklahoma City East (Fig. 21c), Oklahoma City West (not shown), Oklahoma City North (Fig. 21d), or Guthrie (Fig. 22a). It was therefore confined to a swath to the west of the leading edge of the most intense convection (cf. Figs. 1 and 18a–c). The Mesonet traces at Minco, to the north, were similar to the aforementioned, except that the dewpoint did not drop. Instead, the dewpoint at Minco rose slightly, as it did at El Reno and Oklahoma City East and Oklahoma City North (Fig. 21).

Fig. 20.
Fig. 20.

Meteograms at Oklahoma Mesonet sites, displaying both sudden warming and sudden drying after a gust-front passage. Data are plotted from 0330 to 0630 UTC 31 May 2012 (times displayed in a.m. and p.m. along the abscissa are shown in CDT; times displayed vertically are shown in UTC): (top) wind speed (m s−1; green), wind gusts (m s−1; red), and wind direction (°; purple); and (bottom) temperature (°C; red), dewpoint (°C; purple), and pressure reduced to sea level (hPa; green), at (a) Ninnekah, (b) Chickasha, (c) Ketchum Ranch, (d) Washington (site of Purcell profiler), (e) Norman, (f) Hobart, and (g) Weatherford. The onset of the gust front is indicated by an arrow at the time pressure began to rise suddenly; the warming and drying episodes are highlight by a double arrow.

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

Fig. 21.
Fig. 21.

As in Figs. 20a–d, but for (a) Minco, (b) El Reno, (c) Oklahoma City East, and (d) Oklahoma City North, while the double arrow highlights warming and moistening episodes.

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

Fig. 22.
Fig. 22.

As in Figs. 20e–g, but for typical gust-front passages characterized by both temperature and dewpoint falls, at (a) Guthrie, (b) Watonga, and (c) Kingfisher (there are no double arrows, only single arrows).

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

The second and third highest wind speeds recorded at Mesonet sites were at Minco and El Reno (Figs. 1, 2g, and 21a,b) between 0400 and 0500 UTC. These strong wind gusts occurred, like the gust recorded at Ninnekah, <10 km to the rear of the leading radar fine line. While the temperature at both Minco and El Reno rose as it did at Ninnekah, the dewpoint did not fall at either site, as it did at Ninnekah. It is also noted that the highest wind gust at nearby Chickasha (Fig. 20b) was much less than that at Ninnekah, which is relatively nearby (Fig. 1). From Fig. 2g, we see that the swath of wind gusts ≥27 m s−1 was very narrow. Between 0400 and 0500 UTC the strongest radar echoes were located nearly 100 km or so northeast of Oklahoma City and the radar reflectivity factor near the Mesonet sites where the three highest wind speeds were recorded (Figs. 20a and 21a,b) was mainly below 15–20 dBZ (Figs. 11b and 18a).

Mesonet sites to the north and far northwest (Oklahoma City North, Guthrie, Watonga, and Kingfisher; Figs. 21d and 22) experienced more typical gust-front passages: the wind shifted from southeast/south/southwest to north and increased in speed, while the pressure and wind speed both increased, and the temperature and dewpoint fell. In southwestern and far western Oklahoma, at Altus, Retrop, Mangum, and Erick, the gust-front passage (~0500 UTC; see arrows) was also more typical (Fig. 15).

Some insight can be obtained from data from the Purcell Doppler wind profiler, which is collocated with the Washington Mesonet site (Figs. 23a,b) and located to the east of where the highest surface winds were experienced (at Ninnekah; Fig. 1). The surface gust front passed by the Purcell wind profiler site and the Washington Mesonet site at 0500 UTC. Prior to a wind shift just below 1 km AGL at ~0530–0540 UTC, there had been warming at ~1.3 km AGL, followed by cooling. The warming occurred at the same time that there was rising motion at ~1–1.8 km AGL (0520 UTC) and was preceded by sinking motion around 1.5–2.5 km AGL when there was little if any change in temperature (0510 UTC). It thus appears as if there was a small-scale horizontal roll circulation a few kilometers above and ahead of both the wind shift and the gust front, and which was out of phase with the vertical motion. This roll circulation may have been associated with a bore, since it was found near the surface gust front and above the surface, and was not accompanied by a decrease in surface temperature. Typically, there is no surface temperature response during the passage of a bore2 (e.g., Koch et al. 2008; Coleman and Knupp 2011). It is noted that this roll circulation is of the opposite sense of that described by Wakimoto (1982, his Fig. 29), which is up at the leading edge and down to the rear. From the Norman sounding at 0000 UTC (Fig. 23c), it is seen that there is a stable layer from 850 to 775 hPa, and that if air at the top of this stable layer were brought dry adiabatically down to the surface, the air would warm up to ~33°C, which is much warmer than the actual surface temperature was at Ninnekah (25°C) when it experienced simultaneous warming and strong wind gusts (Fig. 20a). However, with mixing and a downward extension of the stable layer at night, mixing is a possible mechanism for the observed warming (Koch et al. 1991; Simpson 1997; Coleman and Knupp 2011).

Fig. 23.
Fig. 23.

The passage of the gust front at the NOAA Purcell wind profiler, very near the Washington Mesonet site. Data plotted are from 0206 to 0900 UTC 31 May 2012 (note that time increases toward the left): (a) wind speed (m s−1) and wind direction (wind barbs) and temperature [°C; with color code at bottom; warm (cool) temperatures are depicted by warm (cool) colors] and (b) vertical velocity (m s−1; with color code at bottom ). Shown in (a) is the time after which either there are no data, or the wind has shifted, at ~750 m AGL. Instances of updrafts (vertical velocity w > 0) and downdrafts (w < 0) are pointed out in (b). The 850-hPa level in a standard atmosphere is indicated at the right, around 1.5 km AGL. (c) Sounding at OUN at 0000 UTC 31 May 2012. (Courtesy of the Plymouth State College weather archive.)

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

From analyses of the surface temperatures at Mesonet sites between 0400 and 0515 UTC, we see that the gust front was marked by much cooler air to its rear and much warmer air mainly in the eastern half of the domain until 0445 UTC; there was warmer air ahead of the gust front and to the west of Ninnekah at 0445 UTC, just prior to the high winds (Fig. 24d). In fact, the swath of high winds is well correlated with the bulging out of relatively cool air ahead of the gust front also after 0445 UTC. By 0500 and 0515 UTC, the highest temperature gradient at the surface was located to the rear of the leading edge of the gust front as marked by the wind shift. We therefore conjecture that the collocation of cooler air at the surface ahead of the gust front and the highest surface wind gusts behind the gust front are circumstantial evidence that the lowest region of air near the ground was relatively stable with respect to air elsewhere, and that the generation of high winds may have been related to a gravity wave/bore phenomenon associated with the impinging of the gust front on the stable air (Rottman and Simpson 1989; Marsham et al. 2011). From a radar perspective, however, no difference in the character of the fine lines to the rear of the gust front can be seen along the length of the gust front. In other words, one could not tell from the radar imagery alone that strong winds were occurring locally along the fine lines because there were no localized bulges in radar reflectivity factor.

Fig. 24.
Fig. 24.

Depiction of the gust-front passage from data plotted from the Oklahoma Mesonet at (a) 0400, (b) 0415, (c) 0430, (d) 0445, (e) 0500, and (f) 0515 UTC 31 May 2012. Warm (cool) temperatures are coded as warm (cool) colors. Winds are plotted as wind barbs. Thick solid line is approximate location of the gust front. Thin solid lines are isobars of pressure reduced to sea level (hPa). Dashed contours enclose regions of calm wind. In (e), the N and W indicate locations of the Ninnekah and Washington (Purcell wind profiler) Mesonet sites, respectively.

Citation: Weather and Forecasting 30, 1; 10.1175/WAF-D-14-00065.1

Since the surface winds were relatively light at many locations ahead of the gust front (Figs. 24a–d, areas within dashed lines), it is likely that mixing was inhibited so that radiational cooling led to relatively cool surface temperatures, and that after the gust had passed, vertical mixing actually led to warmer surface temperatures briefly, before the brunt of the cold air behind the gust front was felt. Why the air cooled off more along a narrow swath ahead of the gust front may have been related to topography, but we can offer no good evidence for this hypothesis. Air at the surface dried out as it warmed behind the gust front, in most places, but to the north the air moistened as it warmed, behind the gust front. The air to the north moistened rather than dried perhaps because there was stronger subsidence or more vigorous vertical mixing to the south.

In summary, the gust front near the MCS in northern Oklahoma behaved in a typical fashion, while much of the gust front farther to the south and west, where there was evidence of borelike and solitary wave train–like fine lines, south of where convection was dissipating, had similar single episodes of warming and drying to the rear of the leading edge of the gust front/first fine line. The local warming and drying may have been due to the sinking branch of a bore to the rear of the gust front, and the warmer and drier air was then mixed down to the surface. This speculation, however, cannot be backed up with hard evidence from observational data.

5. Summary and discussion

This case study illustrates how two separate areas of mesoscale drying at the surface likely had a large influence on the formation and evolution of convective storms. Weckwerth and Parsons (2006), in summarizing the motivation for the IHOP, noted studies in which it was argued that the predictability of convective precipitation is limited by our inability to map out water vapor on small scales (e.g., Emanuel et al. 1995; Dabberdt and Schlatter 1996; National Research Council 1998).

In the case study presented here, two separate instances of mesoscale drying were found. One was a narrow band in western Oklahoma, which originated from apparently subsiding air in decaying elevated convection over northeastern Texas, prior to the initiation of convection later in the day; this area of decaying elevated convection had its origin in an MCS the previous night in Oklahoma and may have inhibited the early development of convection in western and northwestern Oklahoma. During the late afternoon of 30 May 2012 and early evening, a second mesoscale area of drying was found underneath anvil precipitation and elevated convection across northwestern Texas and far southwestern Oklahoma, linked to earlier supercells that had developed in western Texas ahead of the dryline. This area of relatively dry air was advected northward into the path of an MCS propagating southward into western Oklahoma, likely causing it to weaken owing to a decrease in CAPE. The weakening MCS collapsed, generating a southward-propagating gust front, which triggered a bore and a solitary wave train as it hit a stable, nocturnal boundary layer across the southern half of Oklahoma. To the east, the air was not dry and bores were not observed. Since drying was apparently so important to the predictability of subsequent events, and it seems to have come from subsidence underneath decaying convection, numerical prediction models must be able to faithfully simulate the decaying convection in addition to the location and character of the preceding convection. To do so, accurate microphysics and spatial resolution will be necessary. We recommend that numerical simulation experiments be carried out with a cloud-resolving model to see if the observed drying can be simulated. Suppose that mesoscale variations of moisture are important for the predictability of convective storm formation and evolution. Suppose also that in many instances the variations of moisture are due to convective processes, not to horizontal advection. It is then essential that for improving forecasts we must be able to predict the areas of drying accurately; otherwise, predictability will be limited by our ability to observe the drying and not extend back to times prior to the drying. Better initial conditions for numerical weather prediction model runs alone may not be sufficient for improving predictability.

If the mesoscale area of drying located in and upstream from southwestern Oklahoma that occurred underneath the anvil of elevated convection east and northeast of supercells, which developed near the dryline in western Texas, were indeed of primary importance in influencing the late-evening high-wind event, then the formation of the supercells and the upscale growth of them into an MCS were of great importance since they were responsible for events that resulted in the drying produced beneath the anvils. The supercells formed in a region upstream from which there was a band of boundary layer moisture advecting northwestward toward the dryline. This band was associated with a mesoscale cyclone along the dryline, which was probably forced by low-level convergence underneath a current of rising motion, quasigeostrophically induced by vorticity advection becoming more cyclonic with height (Figs. 3b, 4b, and 7). If this were the case, then the location and strength of the trough approaching from the west were of great importance also. Any upward forcing in the presence of low static stability, as found in the deep mixed layer west of the dryline, would have an enhanced response of upward motion. The cyclone, however, straddled a zone separating a deep mixed layer (low static stability) from a shallow mixed layer, if any. The response to upward forcing should therefore have been muted east of the dryline.

Another possible source of vorticity for the mesoscale cyclone was baroclinically generated horizontal vorticity (along the northwest–southeast-oriented zone of temperature contrast between that of the cooler air behind the outflow boundary and the warmer air to the southwest of the outflow boundary, but east of the dryline) that was tilted upward as air parcels encountered the upward branch of the east–west-oriented dryline circulation. From Fig. 7 we can see that the rate of baroclinic generation of vorticity along the boundary, which we can align along the southeasterly flow, was ~(0.1 m s−2)/(105 m) ~ (10−6 s−1) s−1, where buoyancy B ~ gΔθ/θ (where g is the acceleration of gravity and θ is the potential temperature) ~(10 m s−2)(3 K/300 K) ~ (0.1 m s−2), and the change in buoyancy occurred across a zone ~100 km wide. If an air parcel traveled 100 km along the baroclinic zone at a speed of ~10 m s−1, then it traveled a distance of 100 km in ~104 s, so that horizontal vorticity of ~10−2 s−1 would have been generated. If the air parcel was tilted upward as it approached the dryline, then the rate of generation of vertical vorticity (product of vertical shear normal to the boundary and the gradient of vertical velocity across the dryline) ~(10−2 s−1)(1 m s−1)/(104 m) = (10−6 s−1) s−1. The air parcel resided within the dryline zone, which was ~10 km wide (e.g., Weiss et al. 2006), as it moved along at ~10 m s−1, for (104 m)/(10 m s−1) = 103 s and therefore vertical vorticity of ~10−3 s−1 would have been generated. This vorticity, however, would have been generated on a scale of ~10 km, not 100 km. Since the observed scale of the cyclone was ~150–300 km, we conclude that baroclinic generation could not be the mechanism by which the mesoscale cyclone formed and that synoptic-scale processes associated with the upper-level trough were probably dominant.

Because the development and evolution of MCSs that lead to subsidence drying under the anvil is important, another forecasting problem is to be able to predict when supercells develop upscale into MCSs and when they do not. Important factors include the amount of convective inhibition (CIN) and vertical shear in the upstream environment. When a supercell’s cold pool becomes too extensive, for example, when there is copious precipitation and the boundary layer is relatively dry, the supercell may collapse; according to Rotunno–Klemp–Weisman (RKW) theory (Rotunno et al. 1988), additional low-level vertical shear is needed to sustain convection that develops a stronger cold pool. If the supercell moves into an environment of less deep-layer shear, then the supercell may evolve into a multicell. If CIN becomes too large or if baroclinic generation of horizontal vorticity at the leading edge of the cold pool is no longer balanced by the import of environmental horizontal vorticity, the convective system will collapse. Even if we can predict when MCSs will evolve from supercells, we need to understand the conditions under which the anvil region of MCSs generates subsidence drying: not all anvils generate regions of drying near the ground.

A swath of high winds at the surface along a gust front was hypothesized to be a gravity wave-/bore-related phenomenon and occurred where there was a locally more stable low-level air mass ahead of the gust front. More cases need to be examined to see if this finding for just one case can be generalized. If so, then looking in real time at the surface-temperature distribution along and ahead of gust fronts may hold some predictive value. Crook (1986, 1988) demonstrated how when the Scorer parameter (Scorer 1949) does not decrease with height above the stable layer (or decreases relatively slowly), a bore should not last long and may be muted because gravity wave energy is radiated away in the upward direction; however, when the Scorer parameter decreases with height (relatively rapidly), or when there is an inversion “at a certain height above the lower stable layer,” upward radiation of energy is inhibited, and in the latter case, wave energy is reflected back downward, so that a bore should be more long lived and robust. From the 0000 UTC 31 May 2012 sounding at the Norman radar site (OUN), which we assume to be qualitatively representative of the environment, at least above low levels, it is seen that there are stable (nearly isothermal) layers centered ~1.4 km (825 hPa) and 2.6 km (725 hPa) AGL, surmounted by a layer having a lapse rate just under dry adiabatic up to ~7.1 km (425 hPa) (Fig. 23c). Since the Scorer parameter (~N/u) is proportional to the Brunt–Väisälä frequency N and inversely proportional to the zonal wind speed u, it decreases with height z because N decreases with height and u either remains approximately constant or increases with height [the (1/u)(d2u/dz2) term is negligible unless there is a narrow low-level jet that we cannot define very well]. Furthermore, there is a shallow stable layer aloft ~7.1 km AGL. Thus, the necessary conditions for relatively long-lived, high-amplitude bores and solitary wave trains were likely present.

Haertel et al. (2001) numerically showed that by varying the stratification of the atmosphere, the vertical structure of sudden cooling, and the depth and intensity of the cold pool, a density current and/or a gravity wave occur. It is suggested that a simple numerical experiment be conducted to find out how the time-dependent (transient, i.e., gusts in wind speed) wind field responds to different intensities and depths of the stable layer ahead of gust fronts and to different vertical stratifications above the stable layer.

One may say that in the case described in this paper, a “butterfly in Texas flapped its wings and tornadoes were not produced in Oklahoma.” Was the original “butterfly” the trough approaching from the west, which was not forecast accurately enough? Did the trough lead to the formation of a mesoscale cyclone along the dryline and subsequent supercells, which grew upscale into an MCS, the anvil region of which produced subsidence drying, which led to the weakening of the MCS approaching from the north, etc.? Or was the butterfly the decaying MCS over northeastern Texas, from which subsidence drying led to the inhibition of convection from forming in Oklahoma? Or were there two butterflies? We hope that observational analyses of cases like this stimulate thinking about how to conduct future numerical experiments aimed at improving the predictability of severe weather.

Acknowledgments

This study was supported by NSF Grants AGS-0934307, AGS-1262048, and AGS-1237404. This work was also supported by a National Research Council postdoctoral research associateship awarded to the second author. Early versions of this work were presented under the title “A butterfly flaps its wings in Texas and tornadoes are not produced in Oklahoma.” Conversations with Adam Clark (NOAA/NSSL), Mike Coniglio (NOAA/NSSL), and John Brown (NOAA/ESRL) were very useful at the beginning of this study. Rita Roberts (NCAR) provided some helpful comments at a seminar presentation and Rich Rotunno (NCAR), George Bryan (NCAR), and Morris Weisman (NCAR) provided some references. Conversations with Lance Bosart (University at Albany) were also very helpful. Doug Speheger (NWS, Norman) provided warning information. Jana Houser (Ohio University) provided assistance with RaXPol data collection. Conversations with Kevin Haghi (OU) about bores were very helpful. Shawn Riley (OU) provided assistance with data acquisition and display software.

REFERENCES

  • Bluestein, H. B., 2009: The formation and early evolution of the Greensburg, Kansas, tornadic supercell on 4 May 2007. Wea. Forecasting, 24, 899920, doi:10.1175/2009WAF2222206.1.

    • Search Google Scholar
    • Export Citation
  • Bluestein, H. B., 2013: Severe Convective Storms and Tornadoes: Observations and Dynamics. Springer/Praxis, 456 pp.

  • Brock, F. V., , Crawford K. , , Elliott R. , , Cuperus G. , , Stadler S. , , Johnson H. , , and Eilts M. , 1995: The Oklahoma Mesonet: A technical overview. J. Atmos. Oceanic Technol., 12, 519, doi:10.1175/1520-0426(1995)012<0005:TOMATO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Clark, A. J., , Gallus W. A. , , Xue M. , , and Kong F. , 2010: Convection-allowing and convection-parameterizing ensemble forecasts of a mesoscale convective vortex and associated severe weather environment. Wea. Forecasting, 25, 10521081, doi:10.1175/2010WAF2222390.1.

    • Search Google Scholar
    • Export Citation
  • Coleman, T. A., , and Knupp K. R. , 2011: Radiometer and profiler analysis of the effects of a bore and solitary wave on the stability of the nocturnal boundary layer. Mon. Wea. Rev., 139, 211223, doi:10.1175/2010MWR3376.1.

    • Search Google Scholar
    • Export Citation
  • Coniglio, M. C., , Elmore K. L. , , Kain J. S. , , Weiss S. J. , , Xue M. , , and Weisman M. L. , 2010: Evaluation of WRF Model output for severe weather forecasting from the 2008 NOAA Hazardous Weather Testbed Spring Experiment. Wea. Forecasting, 25, 408427, doi:10.1175/2009WAF2222258.1.

    • Search Google Scholar
    • Export Citation
  • Crook, N. A., 1986: The effect of ambient stratification and moisture on the motion of atmospheric undular bores. J. Atmos. Sci., 43, 171181, doi:10.1175/1520-0469(1986)043<0171:TEOASA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Crook, N. A., 1988: Trapping of low-level internal gravity waves. J. Atmos. Sci., 45, 15331541, doi:10.1175/1520-0469(1988)045<1533:TOLLIG>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Crum, T. D., , and Alberty R. L. , 1993: The WSR-88D and the WSR-88D Operational Support Facility. Bull. Amer. Meteor. Soc., 74, 16691687, doi:10.1175/1520-0477(1993)074<1669:TWATWO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dabberdt, W. F., , and Schlatter T. W. , 1996: Research opportunities from emerging atmospheric observing and modeling capabilities. Bull. Amer. Meteor. Soc., 77, 305323, doi:10.1175/1520-0477(1996)077<0305:ROFEAO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Droegemeier, K. K., 1997: The numerical prediction of thunderstorms: Challenges, potential benefits, and results from real-time operational tests. WMO Bull., 46, 324336.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K., and et al. , 1995: Report of the First Prospectus Development Team of the U.S. Weather Research Program to NOAA and the NSF. Bull. Amer. Meteor. Soc., 76, 11941208.

    • Search Google Scholar
    • Export Citation
  • Fiebrich, C. A., , and Crawford K. C. , 2001: The impact of unique meteorological phenomena detected by the Oklahoma Mesonet and ARS Micronet on automated quality control. Bull. Amer. Meteor. Soc., 82, 21732187, doi:10.1175/1520-0477(2001)082<2173:TIOUMP>2.3.CO;2.

    • Search Google Scholar
    • Export Citation
  • Goff, R. C., 1976: Vertical structure of thunderstorm outflows. Mon. Wea. Rev., 104, 14291440, doi:10.1175/1520-0493(1976)104<1429:VSOTO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Haertel, P. T., , Johnson R. H. , , and Tulich S. N. , 2001: Some simple simulations of thunderstorm outflows. J. Atmos. Sci., 58, 504516, doi:10.1175/1520-0469(2001)058<0504:SSSOTO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Johns, R. H., , and Doswell C. A. III, 1992: Severe local storms forecasting. Wea. Forecasting, 7, 588612, doi:10.1175/1520-0434(1992)007<0588:SLSF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Johnson, B. C., 1983: The heat burst of 29 May 1976. Mon. Wea. Rev., 111, 17761792, doi:10.1175/1520-0493(1983)111<1776:THBOM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Johnson, R. H., , Chen S. , , and Toth J. J. , 1989: Circulations associated with a mature-to-decaying midlatitude mesoscale convective system. Part I: Surface features—Heat bursts and mesolow development. Mon. Wea. Rev., 117, 942959, doi:10.1175/1520-0493(1989)117<0942:CAWAMT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kain, J. S., and et al. , 2013: A feasibility study for probabilistic convection initiation forecasts based on explicit numerical guidance. Bull. Amer. Meteor. Soc., 94, 12131225, doi:10.1175/BAMS-D-11-00264.1.

    • Search Google Scholar
    • Export Citation
  • Koch, S. E., , Dorian P. B. , , Ferrare R. , , Melfi S. , , Skillman W. C. , , and Whiteman D. , 1991: Structure of an internal bore and dissipating gravity current as revealed by Raman lidar. Mon. Wea. Rev., 119, 857887, doi:10.1175/1520-0493(1991)119<0857:SOAIBA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Koch, S. E., , Flamant C. , , Wilson J. W. , , Gentry B. M. , , and Jamison B. D. , 2008: An atmospheric soliton observed with Doppler radar, differential absorption lidar, and a molecular Doppler lidar. J. Atmos. Oceanic Technol., 25, 12671287, doi:10.1175/2007JTECHA951.1.

    • Search Google Scholar
    • Export Citation
  • Leary, C. A., , and Houze R. A. Jr., 1979: Melting and evaporation of hydrometeors in precipitation from the anvil clouds of deep tropical convection. J. Atmos. Sci., 36, 669679, doi:10.1175/1520-0469(1979)036<0669:MAEOHI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lilly, D. K., 1990: Numerical prediction of thunderstorms—Has its time come? Quart. J. Roy. Meteor. Soc., 116, 779798, doi:10.1002/qj.49711649402.

    • Search Google Scholar
    • Export Citation
  • Marsham, J. H., , Trier S. B. , , Weckwerth T. M. , , and Wilson J. W. , 2011: Observations of elevated convection initiation leading to a surface-based squall line during 13 June _IHOP 2002. Mon. Wea. Rev., 139, 247271, doi:10.1175/2010MWR3422.1.

    • Search Google Scholar
    • Export Citation
  • Moller, A. R., , Doswell C. A. III, , Foster M. P. , , and Woodall G. R. , 1994: The operational recognition of supercell thunderstorm environments and storm structures. Wea. Forecasting, 9, 327347, doi:10.1175/1520-0434(1994)009<0327:TOROST>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • National Research Council, 1998: The Atmospheric Sciences: Entering the Twenty-First Century.National Academy Press, 364 pp.

  • OFCM, 1995: Surface weather observations and reports. Federal Meteorological Handbook 1, Office of the Federal Coordinator for Meteorological Services and Supporting Research, 99 pp.

  • Pazmany, A. L., , Mead J. B. , , Bluestein H. B. , , Snyder J. C. , , and Houser J. B. , 2013: A mobile, rapid-scanning, X-band, polarimetric (RaXPol) Doppler radar system. J. Atmos. Oceanic Technol., 30, 13981413, doi:10.1175/JTECH-D-12-00166.1.

    • Search Google Scholar
    • Export Citation
  • Rottman, J. W., , and Simpson J. E. , 1989: The formation of internal bores in the atmosphere: A laboratory model. Quart. J. Roy. Meteor. Soc., 115, 941963, doi:10.1002/qj.49711548809.

    • Search Google Scholar
    • Export Citation
  • Rotunno, R., , Klemp J. B. , , and Weisman M. L. , 1988: A theory for strong, long-lived squall lines. J. Atmos. Sci., 45, 463485, doi:10.1175/1520-0469(1988)045<0463:ATFSLL>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Schroeder, J. L., , Burgett W. S. , , Haynie K. B. , , Sonmez I. , , Skwira G. D. , , Doggett A. L. , , and Lipe J. W. , 2005: The West Texas Mesonet: A technical overview. J. Atmos. Oceanic Technol., 22, 211222, doi:10.1175/JTECH-1690.1.

    • Search Google Scholar
    • Export Citation
  • Scofield, R. A., , and Purdom J. F. W. , 1986: The use of satellite data for mesoscale analyses and forecasting applications Mesoscale Meteorology and Forecasting, P. S. Ray, Ed., Amer. Meteor. Soc., 118 – 150.

  • Scorer, R. S., 1949: Theory of waves in the lee of mountains. Quart. J. Roy. Meteor. Soc., 75, 4156, doi:10.1002/qj.49707532308.

  • Simpson, J. E., 1997: Gravity Currents in the Environment and the Laboratory.2nd ed. Cambridge Universitiy Press, 244 pp.

  • Smull, B. F., , and Houze R. A. Jr., 1987: Rear inflow in squall lines with trailing stratiform precipitation. Mon. Wea. Rev., 115, 28692889, doi:10.1175/1520-0493(1987)115<2869:RIISLW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Trapp, R. J., , Stumpf G. J. , , and Manross K. L. , 2005: A reassessment of the percentage of tornadic mesocyclones. Wea. Forecasting, 20, 680687, doi:10.1175/WAF864.1.

    • Search Google Scholar
    • Export Citation
  • Wakimoto, R. M., 1982: The life cycle of thunderstorm gust fronts as viewed with Doppler radar and rawinsonde data. Mon. Wea. Rev., 110, 10601082, doi:10.1175/1520-0493(1982)110<1060:TLCOTG>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Weber, B. L., and et al. , 1990: Preliminary evaluation of the first NOAA Demonstration Network Wind Profiler. J. Atmos. Oceanic Technol., 7, 909918, doi:10.1175/1520-0426(1990)007<0909:PEOTFN>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Weckwerth, T. M., , and Parsons D. P. , 2006: A review of convection initiation and motivation for IHOP 2002. Mon. Wea. Rev., 134, 522, doi:10.1175/MWR3067.1.

    • Search Google Scholar
    • Export Citation
  • Weisman, M. L., 1992: Convectively generated rear-inflow jets in the evolution of long-lived mesoconvective systems. J. Atmos. Sci., 49, 18261847, doi:10.1175/1520-0469(1992)049<1826:TROCGR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Weisman, M. L., , Davis C. , , Wang W. , , Manning K. W. , , and Klemp J. B. , 2008: Experiences with a 0–36-h explicit convective forecasts with the WRF-ARW Model. Wea. Forecasting, 23, 407437, doi:10.1175/2007WAF2007005.1.

    • Search Google Scholar
    • Export Citation
  • Weiss, C. C., , Bluestein H. B. , , and Pazmany A. L. , 2006: Finescale radar observations of the 22 May 2002 dryline during the International H2O Project (IHOP). Mon. Wea. Rev., 134, 273293, doi:10.1175/MWR3068.1.

    • Search Google Scholar
    • Export Citation
  • Zipser, E. J., 1977: Mesoscale and convective-scale downdrafts as distinct components of squall-line structure. Mon. Wea. Rev., 105, 15681589, doi:10.1175/1520-0493(1977)105<1568:MACDAD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
1

Calculations were made from the 12-h NAM forecast valid at 0000 UTC 31 May 2012 and the NAM analysis at 0000 UTC, but the forcing estimated from the divergence of Q vectors was noisy, even when filtered.

2

A number of cases have been found in the International H2O Project (IHOP) (Weckwerth and Parsons 2006) data archive in which the temperature at the surface increased when a bore passed, but none, to the best of our knowledge, has been documented in the literature (K. Haghi, OU, 2014, personal communication).

Save