Lightning activity of thunderstorms in relation to environmental and radar characteristics

Oscar van der Velde

Abstract Introduction This is an abstract from my Master's thesis research, which I did at the KNMI (September 1999-April 2000). My supervisors were Iwan Holleman and Herman Wessels (KNMI). The main objective of this study was to give insight into the temporal and spatial aspects of lightning activity during the life cycle of diverse types of thunderstorms, and to examine the possible relationships with thunderstorm environment and radar characteristics. The two radars from the Royal Netherlands Meteorological Institute (KNMI) were used together with the SAFIR two-dimensional lightning mapping interferometer system. Synops and upper air soundings were used to examine thunderstorms and their lightning activity in their synoptic environment. The research questions were: What are the general relationships between environment, type and radar characteristics of a thunderstorm and its lightning activity? How does lightning activity evolve during a thunderstorm's life cycle and where in the storm does it occur? Can lightning activity be predicted on a basis of radar and environmental characteristics? Methods Radar and lightning data were gathered from 48 thunderstorms on 11 days in summer 1999. If possible, individual cells were followed by zooming in. In case that cells could not easily be followed in time, the whole cluster of complex has been taken. The SAFIR-determined lightning type is assumed to be more correct over a larger number of strokes. Its spatial accuracy has a mean standard deviation of about 2.6 km (Wessels, 1998). However, in some regions where the second farthest antenna is more than 160 km away from the flash, the accuracy is much less, showing a ray-effect towards an antenna. The chosen storms occurred in a good to reasonable detection area, so that lightning activity could be addressed to one cell or storm complex, at least during the greater part of their lifes. 15 minutes-data was used and strokes instead of flashes (the latter is used in most literature). The area with reflectivities greater than 40 dBZ, the highest echo tops and Vertically Integrated Liquid (VIL) were gathered. Examples of graphs with radar & lightning characteristics: Line 1 - June 2, 1999 (a small but active line with relatively many cloud-to-ground strokes) Cell 8a - June 2, 1999 (this cell produced 4 cm (1.5 inch) diameter hail and severe microburst damage around 15 UTC in the vicinity of Barchem/Lochem in the east of the Netherlands) The investigated thunderstorms occurred within a Convectively Available Potential Energy (CAPE)-range from 150-1600 J kg-1 (average about 630 J kg-1). View a graph of the investigated cells in the CAPE-shear parameter space (top graph) CAPE was computed by replacing the lowest sounding level data by those measured by ground stations in the vicinity of (but not influenced by) the thunderstorms, and then lifting the lowest 50 hPa mixed parcel. For nocturnal thunderstorms the most unstable parcel was used. In determining CAPE for an individual cell, some subjectivity cannot be ruled out. The BRN shear (BL-6 km shear vector magnitude) ranged from 2 to 16 m s-1 (average 7-8 m s-1). Conclusions The conclusions concerning the first and second question are listed below. With "relative lightning activity" is meant: the activity per km2 from the thunderstorm with a reflectivity higher than 40 dBZ (about the convective region). As found in this study: The maximum attainable peak of total lightning activity increases with the amount of instability and (less) with the amount of vertical windshear. See CAPE graphs and BRN-shear graphs, top left. A storm's total lightning activity relates closely to the area of its convective region. This gives a mean lightning production of 0.13 strokes per minute per km2 for reflectivities higher than 40 dBZ in the convective region of thunderstorms. See Area > 40 dBZ graphs, top. Total activity also increases with echo top heights, but the spread is larger (Echo top & VIL graphs). There is no correlation between the amount of Vertically Integrated Liquid and the lightning activity or relative lightning activity (Echo top & VIL graphs). The (maximum attainable) relative lightning activity only increases with instability (CAPE). For example, at 1000 J kg-1 a maximum of 0.3-0.4 strokes per km2 > 40 dBZ min-1 (as an average over the entire convective region at peak activity) can be expected. This is one of the stronger correlations (0.71). See CAPE graphs (top right) and the one with Cell types (bottom graph). The maximum attainable percentage cloud-to-ground discharges decreases with increasing instability. There is more spread at lower values of CAPE (see CAPE graph, middle left) The same relationship is found for VIL. Vertical windshear is not significantly of influence on the percentage cloud-to-ground strokes. The same relationship is found for positive ground strokes (see BRN-shear graph, bottom). The ratio of positive to negative ground strokes does not show any relationship with CAPE, wind shear, temperature and moisture, nor the radar characteristics (graphs not shown, totally scattered) In this dataset there is too little variation in temperature at 700 hPa and K-index (used for moisture) to draw conclusions about (see K-index graphs and Temperature at 700 hPa graphs) VIL and echo tops usually peak in the 15-minute period before the total lightning activity peak (graph not shown). The time of maximum extent of the convective zone usually coincides with the total lightning peak (graph not shown). For higher CAPE values the time difference between start and peak of lightning activity is longer on average, but there is no relation with the peak-to-end time (begin/peak/end/CAPE graph). The highest relative lightning activity usually occurs around the time of the lightning peak. The majority of the individual cells has only one lightning activity peak. The mean times and activity from the cell types are summarized in the table below (the lowest concerns entire clusters instead of individual cells): Avg. time start-peak Avg. time peak-end Avg.peak- activity Maximum peakact. Avg. area >40 dBZ at peak (minutes) (minutes) (strok. min-1) (strok. min-1) (km2) Isolated cel 40 40 11 20 90 (16 pixels) Cell in cluster/line 45 55 15 30 135 (24 pixels) Large cluster or complex 70 90 >30 230 * Strongly variable In thunderstorms, reaching a precipitation intensity of 10 mm h-1 (~40 dBZ) at the ground often coincides with the beginning of lightning activity. But whether this always results in lightning activity cannot be said because no thunderless storms are investigated in this study. Lightning activity often ceases somewhat earlier (up to 45 minutes) than the 40 dBZ reflectivity. Most discharges occur within the 40 dBZ area, in the neighbourhood of the highest echo tops and VIL cores. During the life cycle of a thunderstorm the percentage of ground strokes can vary strongly. This could not be explained with the available radar parameters (no Doppler wind information and precipitation type identification). The observation that first cloud discharges peak and after that the ground discharges (e.g. Williams et al., 1989), cannot be confirmed in this study. For most cells it is found that these peaks occur simultaneously. (Possible reasons: 15 minute-data, strokes instead of flashes and perhaps the system used in this study) At low activity (mostly <3 discharges per minute, sometimes higher in a larger storm system), mostly at the end of a thunderstorm, relatively high percentages of ground strokes occur. From that, often relatively more have a positive charge transfer. Other interesting observations Although not systematically investigated here, the impression was obtained that thunderstorms with relatively many cloud discharges (= relatively few (negative) ground strokes) and a high activity often produce severe weather (large hail or damaging windgusts), while this seems less the case with thunderstorms producing relatively many ground discharges (most of them negative). Furthermore it was more anecdotically found that the thunderstorms with relatively few negative ground strokes and reports of large hailstones also have a typical structure: a strong core at the southern side of a more stratiform region, or at the southern end of a line of thunderstorms. (It occurs on June 6 1998, May 10 1999, June 2 1999, August 8 1999 and September 24 1999.) This looks a bit like the configuration of a supercell. For supercells, the angle between the orientation of the stratiform region (Forward Flank Downdraft) and the direction of movement is often smaller than those of the cases seen here. Such a configuration appears to favor strong updrafts (needed to keep large hail up in the cloud), but may not be favorable for the development of rotation and tornadoes. A meteorologist should be aware of the high probability of large hail if such storms appear on the radar. But perhaps not all large-hail storms show such a structure. View a gif-animation of the June 6, 1998 storms. This animation is made of 2-radar-combined PseudoCAPPI 800m reflectivity images, 1 Mb. First you can see the morning complexes. They have relatively more CG discharges than the afternoon complexes. Note the southern tip of the asymmetrical Mesoscale Convective System (the first appearing MCS in the afternoon), with a very high stroke rate (mostly intracloud), and at its most intense moment producing more positive than negative strokes and 10 cm (4 inch) hail around Zeewolde, province of Flevoland. Also note that for image clarity the lightning timespan is only 3 minutes! You can also see that in S-Belgium, N & E-Netherlands and Germany the lightning detection and positioning is worse than in C-Netherlands and Belgium. Contrary to most observations of positive ground flashes in the stratiform region of MCS's, in this case the SAFIR system detects almost none of them (neither in a 15 minutes-lightning period). Local time is UTC + 2 hours. Also some suspect mini-supercells were discovered (September 24 1999), but not studied in full detail. With a CAPE of about 500 J kg-1 and SREH of 249 m2 s-2, and near an upper level cold pool above Great Britain, some storms developed. At first sight they looked like ordinary clustered multicell storms on the radar, but when looking at vertical cross sections they exhibited strong echo overhang at the southeastern side of the cores. The cores were small but very strong: >55, sometimes even >63 dBZ at PseudoCAPPI height (800 meters). Hail damage was reported and there were some visual observations of tuba's and collar clouds below a smooth cloud base. Doppler data was not available yet. The lightning activity for these storms were quite the same: a high activity for such a small cell (30 strokes min-1), very few cloud-to-ground strokes (about 5% CG, sometimes more positive than negative strokes). View a gif-animation of these storms, followed by an active line, which dies with relatively many positive strokes (PseudoCAPPI 800m reflectivity images, 540 Kb) Suggestions for further research: Study more thunderstorms occurring in higher CAPE and shear environments. Use of (maximum) lightning densities instead of lightning activity normalized with convective region size. Compare local lightning (type) densities to local values of VIL, maximum reflectivity, echo tops and Doppler vertical velocity. In the Netherlands, CAPE's of 1500 J kg-1 (determined via the method described above) occur only few times a year. CAPE's above 2500 J kg-1 are very rare. With a broader range of CAPE a better defined relationship between CAPE and relative lightning activity could be found. It is also better to investigate the maximum total lightning density (i.e. strokes or flashes per km2 per minute) and when use smaller time windows (which is for a radar volume scan a limiting factor). Use of Doppler and multiparameter (polarisation diversity) radar, and vertical reflectivity structure, preferrably with a high time resolution. Vertical velocities are thought to have a great influence on flash type (e.g. Lang et al., 2000). Multiparameter radar can detect precipitation type, which is important in charge generation and transportation processes. Investigate systematically the lightning activity before, during and after large hail occurrence. Accurate reports of hail size, time and location are needed. A problem will be that a lack of reports does not have to mean that no hail occurred. Perhaps the use of multiparameter radar can help on this point. For operational use: automatic cell tracking algorithm and radar & lightning data presentation in graphs versus time. In this way a meteorologist can get a quick overview of which cells are intensifying and which cells already are over their peak. Vertically Integrated Liquid proved to be a good parameter to track individual cells manually in this study. List of references used in this study: Atkins, N.T., en R.M. Wakimoto, 1991: Wet microburst activity over the Southeastern United States: Implications for forecasting. Wea. Forecast., 6, 470-482. Bluestein, H.B., en D.R. MacGorman, 1998: Evolution of cloud-to-ground lightning characteristics and storm structure in the Spearman, Texas, tornadic supercells of 31 May 1990. Mon. Wea. Rev., 126, 1451-1467. Branick, M.L., en C.A. Doswell III, 1992: An observation of the relationship between supercell structure and lightning ground-strike polarity. Wea. Forecast., 7, 143-149. Changnon, S.A., 1992: Temporal and spatial relations between hail and lightning. J. Appl. Meteor., 31, 587-604. Engholm, C.D., E.R. Williams, en R.M. Dole, 1990: Meteorological and electrical conditions associated with positive cloud-to-ground lightning. Mon. Wea. Rev., 118, 470-487. Fuquay, D.M., 1982: Positive cloud-to-ground lightning in summer thunderstorms. J. Geophys. Res., 87, 7131-7140. Goodman, S.J., D. Buechler, S. Hodanish, D. Sharp, E. Williams, B. Boldi, A. Matlin, en M. Weber, 1999: Total lightning activity associated with tornadic storms. Proc. 11th Conf. Atm. Electricity, 515-518. Houze, R.A., Jr., B.F. Smull, en P. Dodge, 1990: Mesoscale organisation of springtime thunderstorms in Oklahoma. Mon. Wea. Rev., 118, 613-654. Kane, R.J., 1991: Correlating lightning to severe local storms in the northeastern United States. Wea. Forecast., 6, 3-12. Keighton, S.J., H.B. Bluestein, en D.R. MacGorman, 1991: The evolution of a severe mesoscale convective system: cloud-to-ground lightning location and storm structure. Mon. Wea. Rev.,119, 1533-1556 Lang, T.J., S.A. Rutledge, J.E. Dye, M. Venticinque, P. Laroche, E. Defer, 2000: Anomalously low negative cloud-to-ground lightning flash rates in intense convective storms observed during STERAO-A. Mon. Wea. Rev., 128, 160-173. MacGorman, D.R., D.W. Burgess, V. Mazur, W.D. Rust, W.L. Taylor, en B.C. Johnson, 1989: Lightning rates relative to tornadic storm evolution on 22 May 1981. J. Atmos. Sci., 46, 221-250 MacGorman, D.R. en D.W. Burgess, 1994: Positive cloud-to-ground lightning in tornadic storms and hailstorms. Mon. Wea. Rev., 122, 1671-1697. MacGorman, D.R. en W.D. Rust, 1998: The Electrical Nature of Storms. Oxford University Press, New York, 422 pp. Markowski, P.M. en J.M. Straka, 2000: Some observations of rotating updrafts in a low-buoyancy, highly sheared environment. Mon. Wea. Rev., 128, 449-461. Nielsen, K.E., R.A. Maddox, en S.V. Vasiloff, 1994: The evolution of cloud-to-ground lightning within a portion of the 10-11 June 1985 squall line. Mon. Wea. Rev., 122, 1809-1817. Perez, A.H., L.J. Wicker, en R.E. Orville, 1997: Characteristics of cloud-to-ground lightning associated with violent tornadoes. Wea. Forecast., 12, 428-437. Rasmussen, E.N., en D.O. Blanchard, 1998: A baseline climatology of sounding-derived supercell and tornado forecast parameters. Wea. Forecast., 13, 1148-1164. Richard, P., 1992: Application of atmospheric discharges localisation to thunderstorm nowcasting. Proc. WMO Tech. Conf. Instruments and Methods of Observation, World Meteor. Org. Rutledge, S.A., C. Lu, en D.R. MacGorman, 1990: Positive cloud-to-ground lightning in mesoscale convective systems. J. Atmos. Sci., 47, 2085-2100. Rutledge, S.A., en W.A. Petersen, 1994: Vertical radar reflectivity structure and cloud-to-ground lightning in the stratiform region of MCS's: Further evidence for in-situ charging in the stratiform region. Mon. Wea. Rev., 122, 1760-1776. Schmid, W., H.H. Schiesser, M. Furger, en M. Jenni, 2000: The origin of severe winds in a tornadic bow-echo storm over Northern Switzerland. Mon. Wea. Rev., 128, 192-207. Stolzenburg, M., 1994: Observations of high ground flash densities of positive lightning in summertime thunderstorms. Mon. Wea. Rev., 122, 1740-1750. Watson, A.I., R.L. Holle en R.E. L?pez, 1995: Lightning from two national detection networks related to vertically integrated liquid and echo-top information from WSR-88D radar. Wea. Forecast., 10, 592-605. Weisman, M.R., en J.B. Klemp, 1982: The dependence of numerically simulated convective storms on vertical wind shear and buoyancy. Mon. Wea. Rev., 110, 504-520. Wessels, H.R.A., 1998: Evaluation of a radio interferometry lightning positioning system. KNMI Scientific Report 98-04. Williams, E.R., M.E. Weber, en R.E. Orville, 1989: The relationship between lightning type and convective state of thunderclouds. J. Geophys. Res., 94, 13,213-13,220

Back to Oscar's weather photography site

Last edited: March 8, 2001.