Research on Electrical Properties of Severe

Home , Electrical phenomena

Research on Electrical Properties W. David Rust, William L. Taylor, and of Severe Thunderstorms Donald R. MacGorman National Severe Storms Laboratory NOAA/ERL in the Great Plains Norman, Okla. 73069 and Roy T. Arnold Physics Department University of Mississippi Oxford, Miss. 38655

Abstract nosity high within the cloud of a tornadic storm, a phenome- non he calls the tornado pulse generator (Jones, 1965). Gunn In 1978 we began a coordinated effort to study the electrical behav- (1956) concludes from measurements on a large tornado that ior of large and severe thunderstorms that form over the Great passed near an eight-station field mill network that there is Plains of the central United States. Methods of approach include the little evidence of unusual electrical effects near the tornado study of characteristics of individual phenomena and storm case funnel and that the observed electrical behavior is what studies. Our goal is to understand the interrelationships between electrical phenomena and the dynamics and precipitation of storms. would be expected from an ordinary thunderstorm if turbu- Evidence that interrelationships do exist can be seen in the results to lent velocities were increased by an order of magnitude. date. In one squall-line storm we have studied, 44% of all observed Vonnegut and Moore (1957) interpret the smooth electric lightning flashes were cloud-to-ground (CG); the total flashing rate field waveforms obtained closest to the tornado by Gunn averaged 12 min-1 and coarsely followed the changes in Doppler- derived maximum updraft speed. Most of the intracloud (IC) dis- (1956) as possibly due to continuous charge flow, and they charge processes in a supercell severe storm were located predomi- suggest that cloud-to-ground (CG) lightning may be sup- nately around the region of the intense updraft of the mesocyclone pressed in the vicinity of the tornado funnel. Frier (1959) re- and near large gradients in reflectivity and horizontal velocity. ports the presence of a 4-cycle-per-minute oscillation in the Both 10 cm and 23 cm wavelength radars have been used to detect electric field that he attributes to a tornado about 100 km lightning radar echoes. The lightning echoes from the 10 cm radar generally had peak signals 10-25 dB greater than the largest precipi- away. Numerous eyewitness observations of unusual lumin- tation echo in the storm, and they usually were observed where pre- osity and electrical activity near tornadoes have been col- cipitation reflectivities were less than maximum. Comparison of lected by Vonnegut (1968). Scouten et al. (1972), Taylor lightning echoes and electric field changes shows that abrupt in- (1973), Kinzer (1974), and Brown and Hughes (1978) have creases in radar reflectivity often are associated with return strokes and K-type field changes. measured intense sferics from severe storms and conclude CG flashes that lower positive charge to earth have been observed that they are associated, not with the tornado, but with the to emanate from the wall cloud, high on the main storm tower, and parent storm. Davies-Jones and Golden (1975) made day- well out in the downwind anvil of severe storms. The percentage of light movies near eight tornadoes, and they observed and CG flashes that lower positive charge is apparently small. filmed very little electrical activity near the tornado funnels. They found no vegetation scorching (Vonnegut, 1960) along the damage paths of 21 tornadoes. Zrnic' (1976) reports that 1. Introduction no abnormal magnetograms were obtained for 10 tornado passages within 12 km of magnetometer sites. In addition to Until recently, the study of electricity from severe storms has these observations, discussions on possible direct links be- been typified by isolated measurements of sferics, surface tween electricity and tornado genesis have been presented by electric fields, or visual observations. This led the National Vonnegut (1960), Wilkins (1964), Colgate (1967, 1968), and Severe Storms Laboratory (NSSL), which is located in the Watkins et al. (1978). Great Plains, to expand its research on severe storms to in- In the central United States, large and often severe storms clude a study of their electrical behavior. typically occur during the spring and summer. These storms Results from earlier studies by others, e.g., Jones (1951) can be categorized roughly into two groups. In the first group and Vonnegut and Moore (1959), include the finding of high are those triggered by synoptic-scale events such as fronts rates of electrical activity in large storms. Furthermore, there and disturbances in the jet stream. These storms are most have been several isolated studies of electrical phenomena in frequent during the spring. In the second group are those tornadic storms. Jones (1951) discusses tornado tracking triggered by smaller mesoscale processes. These generally using sferics and observations of cyclic pulsations of lumi- occur in the summer and move more slowly than the storms in the first group. Springtime storms are emphasized here. They are typified by large hail (>3 cm in diameter is not un- 0003-0007/81/091286-08$06.00 usual), strong straight-line winds (often >30 m/s), high © 1981 American Meteorological Society cloud tops (often 15 km and above), intense updrafts, and

1286 Vol. 62, No. 9, September 1981

Unauthenticated | Downloaded 10/07/21 08:25 AM UTC Bulletin American Meteorological Society 1287 mesocyclones (e.g., Lemon and Doswell, 1979), and they which have microsecond rise times and exponential decay sometimes produce tornadoes. Storms can occur as individ- time constants of 10 s for the slow antenna and 100 /JLS for the ual cells along a squall line or they can be isolated, and their fast antenna. Other parameters we measure include the at- structure can vary from the single to the multicell type. mospheric electric field at the ground, optical transients from Sometimes an isolated storm develops into a supercell lightning, strike points for CG flashes, and corona current. (Browning, 1977), which is a particularly long-lived severe Movie and television video tape recordings of clouds and storm. The varied severe storm situations make coordinated lightning also are made. study of their electrical behavior challenging. Our goal is to A van has been equipped as a mobile laboratory to meas- develop an understanding of relationships between electrical ure most of the electrical phenomena mentioned previously. phenomena and the dynamics and precipitation of storms. Mobile intercept of severe storms and tornadoes has been ac- We are examining these both in case studies of storms and for complished for several years by NSSL for photographic stud- studies of specially selected events that occur during these ies and verification of Doppler radar signatures (Davies- storms. Jonesand Kessler, 1974; Davies-Jones, 1981), but application The first season of coordinated electrical observations at of the technique to electrical measurements is new (Arnold NSSL was 1978. At first, use of the Doppler radars was dom- and Rust, 1979). The use of a mobile laboratory creates logis- inated by other experimental priorities, and only a few par- tical problems, but placing instrumentation in a relatively tial data sets suitable for our purposes were obtained. During fixed position within the severe storm, in particular beneath the 1980 season, however, radar data for use specifically in the precipitation-free cloud base (Fig. 1), provides the possi- electrical studies were acquired. Initial results from the 1978 bility of making quantitative electrical measurements in this and 1979 seasons and a few comments based on the cursory region. The region is characterized by a strong, inflowing examination of a small amount of data from 1980 are pre- low-level wind that turns upward into the updraft. A meso- sented here. cyclone, with a typical rotation diameter of 5-10 km, often develops in this region, and Doppler radar usually displays a velocity distribution resembling a Rankine combined vortex 2. Instrumentation (Donaldson, 1970). The mesocyclone is sometimes visible beneath cloud base as a lowered rotation called a wall cloud. Radar information on storms is obtained with both conven- If strong tornadoes occur, they usually form under the wall tional and Doppler radars. NSSL operates two 10 cm wave- cloud. Details of mesocyclones and related storm develop- length Doppler radars and one 10 cm conventional radar ment have been given by several investigators, e.g., Lemon (WSR-57). One Doppler (NRO) is located at NSSL and the and Doswell (1979). Using the results of the previous work other (CIM) is 42 km to the northwest. This arrangement on the correlation between visual severe storm cloud features forms a primary dual-Doppler data acquisition area that is and Doppler-derived storm windfields, we can, through shaped approximately as a figure 8, about 200 km in length documentation of the visual features of these storms, infer and 100 km in width, and with the long axis aligned from gross storm dynamics with which to correlate electrical activ- southwest to northeast (Brown et al., 1975), which is gener- ity even when storms occur outside the Doppler data acquisi- ally the direction of movement of springtime storms. Both tion area. Mobility also increases substantially the number of precipitation reflectivity and single Doppler information, severe storms that can be studied. i.e., the radial component of velocity toward or away from the radar, can be obtained to ranges greater than 300 km. Dual Doppler data usually are obtained by coordinating scans up through a selected storm or region of a storm. Each 3. Observations tilt sequence takes about 4-5 min. This scanning technique and dual Doppler data synthesis provide much information a. Squall line, 6 June 1979 on the structure and dynamics of large storms (Ray et al., 1975). Other meteorological data are obtained from atmos- We examine one group of cells (sustained, identifiable re- pheric soundings and NSSL's surface network that measures flectivity cores) within a squall line that developed during wind, pressure, temperature, and humidity. The NRO midafternoon on 6 June 1979. This group of cells constitute Doppler and a colocated 23 cm wavelength conventional the "storm" of interest. Our analysis thus far covers this radar are utilized to acquire radar echoes from lightning. The storm as it approached the laboratory from the west during NRO Doppler also has been used to study electric-field levi- 1640-1708 CST. In Fig. 2 the 60° sector (dashed lines) to the tation of precipitation. west encloses most of the storm, which is identified by the Measurements of electrical phenomena are made with hatched area within the 45 dB(Z) contour. The 60° sector is both fixed and mobile facilities. The fixed facilities include that from which impulses were received with the VHF map- Taylor's (1978) lightning discharge mapping system that al- ping system. This storm is categorized as severe because it lows 3-dimensional location of VHF radiation sources produced large hail (prior to 1640 CST) and straight-line (20-80 MHz) from lightning. Each station can record im- winds in excess of 25 m/s. There was neither a mesocyclone pulse rates of up to 1.6 X 104 s_1 and out to a range of about nor a tornado. 60 km. Typical 3-dimensional location accuracy varies be- We located 342 flashes within this storm between 1640 and tween about ±250 m at 30 km to ±1 km at 60 km. Electric 1708 CST. Of these flashes, 149 were CG, giving an intra- field changes, AE, are measured with sensors (slow and fast cloud (IC):CG ratio of 1.3:1. We mapped the CG strike lo- antennas) of the type described by Krehbiel et al. (1979), cations onto radar reflectivity contours at a height of 5 km.

Unauthenticated | Downloaded 10/07/21 08:25 AM UTC 1288 Vol. 62, No. 9, September 1981

FIG. 1. Schematic representation of a typical tornadic storm as viewed from the southeast. The sketch is not to scale; the horizontal dimension is compressed (after original diagram by C. Doswell and B. Dirham). The favored position of the mobile laboratory is shown by the symbol beneath and left of the wall cloud.

The median value of reflectivity at 5 km above the strike dual-Doppler data acquisition periods. The lightning activity points was 40 dB(Z). About 75% of all flashes struck beneath is characterized generally by a slight decrease in CG flashes reflectivities of >20 dB(Z). The strike point does not, how- and an increase in IC activity during the last interval ana- ever, indicate the location of the intracloud position of these lyzed. There is an interesting correlation of increases in up- flashes because of significant horizontal propagation (e.g., draft, reflectivity, and flashing rates in the 16-20 min inter- Teer and Few, 1974; Taylor, 1978; and Krehbiel et al., 1979). We examined the electric field change records for waveforms indicative of continuing current (CC) (Kitagawa et al., 1962) and found them in only 9% of the CGs. This is a lower limit, since if we were unable to discern clearly a waveform indicative of a CG with CC we did not include it in the count. The actual occurrence of CC appears to have been substantially less than the 46% of CGs containing CC found for storms in New Mexico by Kitagawa et al. (1962). Flashing rates in the observed storm, while not extreme, were high, with averages of 5 min-1 for CGs and 12 min-1 for all flashes. From the dual-Doppler radar data, we have derived esti- mates of all three wind velocity components. Shown in Fig. 3 are the Doppler analysis at one height and the reconstructed VHF source locations for one flash that was typical this storm. The source locations tended to occur in regions of weak updraft (<10 m/s), often adjacent to regions of downdraft, and mostly in the northwest half of the storm. After storm motion was subtracted from the horizontal wind, horizontal streamlines at the level of the lightning generally flowed from the reflectivity cores and strong updrafts toward the lightning. Few VHF sources were within cores of high reflectivity (>45-50 dB(Z)). We also have examined various radar-determined parameters to correlate with lightning. Maximum reflectivity and maximum inferred updraft speed are shown in the lower por- FIG. 2. Outline of the squall line and storm of interest (hatched) tion of Fig. 4, with the number of lightning flashes plotted in 6 June 1979 (79157) at 1638 CST. Inner, heavy contours >45 dB(Z) the upper portion of the figure. The plots are partitioned into reflectivity. Dashed lines denote the 60° sector of the VHF mapping 4 min time intervals that coincide approximately with the system; range rings are labeled in km.

Unauthenticated | Downloaded 10/07/21 08:25 AM UTC Bulletin American Meteorological Society 1289

FIG. 3. VHF source locations for a flash at 1640:18 CST superim- posed on radar data for an Oklahoma storm on 6 June 1979. Each calculated VHF source location in this flash is plotted as + . Radar data are shown at a height of 6.5 km from a series of radar scans between 1644 and 1648 CST. Contour lines indicate radar reflectivity in dB(Z). Arrows indicate streamlines in the horizontal wind after storm motion is substracted. Dotted shading indicates areas of downdraft (mostly <5 m/s). Striped shading indicates areas of updraft >10 m/s.

FIG . 4. Number of flashes, maximum radar reflectivity, and maximum Doppler-derived updraft speeds during a 28 min period of a storm on 6 June 1979. Radar data are from a height of 6.5 km. Time val. Note also that the maximum updraft speed and the total zero is at 1640 CST. flashing rate change in the same direction. A determination of whether this is characteristic of large storms awaits addi- tional analysis.

radar to study lightning is twofold: 1) to study the physical b. Severe storm, 20 June 1978 processes of lightning and 2) to locate lightning activity Only single Doppler data up to midstorm levels (7 km) are within severe storms throughout extended periods of the life- available from a severe thunderstorm on 20 June 1978. This time of a storm. Shown is the concept of acquiring lightning storm contained hail, produced a mesocyclone with three re- echo data with the Doppler radar at NSSL schematically in ported funnel clouds (but no observed tornado), and was Fig. 6. Figure 7 shows an example of a lightning radar echo characterized as a supercell. Discharge processes within the with the simultaneous AE waveform from the slow antenna. cloud were located predominantly around the mesocyclone The radar video signal shown is from a selected range inter- and near the large gradients in reflectivity and horizontal ve- val, AR, at a range of about 86 km and an altitude of 4 km. locity. Figure 5 is a plan view of the storm's reflectivity struc- The simultaneous slow antenna AE waveform shows a field ture at a height of 4 km. Superposed on the reflectivity are all change lasting about 1.4 s that apparently was caused by an the VHF impulses that could be located from those observed IC flash that developed into a CG flash with at least eight re- between 1906 and 1911 CST, and a star showing the center of turn strokes. The range of the CG strike point for this flash the mesocyclone. The impulses are from seven flashes (four was placed independently by the CG location system at were CGs) and have a mean height of 6.5 km. 123 km from the radar site. The radial length of the flash within the radar antenna beam was about 90 km. We have reported previously a preliminary analysis of 19 c. Radar observations of lightning lightning radar echoes observed during a summertime severe Observations of radar echoes from lightning have been re- storm (Szymanski and Rust, 1979). Those lightning radar ported by several investigators, e.g., Ligda (1950, 1956); echoes were detected at ranges from 30 to 180 km and at al- Miles (1952, 1953); Hewitt (1953, 1957); Goncher et al. titudes of up to 10 km. They generally had peak signals (1977); and Szymanski et al. (1980). Our purpose in using 10-25 dB greater than the largest precipitation echo in the

Unauthenticated | Downloaded 10/07/21 08:25 AM UTC 1290 Vol. 62, No. 9, September 1981

FIG. 5. Reflectivity (dB(Z))at a height of 4 km on 20 June 1978 at 1902 CST. The star marks the mesocyclone center, and the dots are the plan location of all VHF impulses from all heights that could be mapped between 1906 and 1911 CST.

storm. Lightning echoes usually were observed where precipitation reflectivities were less than maximum, in agreement with lightning locations observed by Proctor (1974) and MacGorman (1978) using VHF and acoustic mapping tech- niques, respectively. The location of flashes reported by Szymanski and Rust may be biased, however, since very intense precipitation echoes at 10 cm wavelength can obscure the echoes from lightning channels. To reduce substantially the radar return from precipitation, present experiments at NSSL also are utilizing a longer wavelength (23 cm) L-band radar with circular polarization. Other aspects of our present study include continued lightning echo measurements with the Doppler radar in storms overhead to correlate lightning occurrence with changes in precipitation reflectivity and velocity and with the L-band radar for long-range location of lightning within severe storms. As seen in Fig. 7, there can be abrupt increases in the am- FIG. 6. Sketch of data acquisition scheme for lightning radar plitude of the lightning radar echo. Lightning echoes from echoes and associated electric field changes (AF). The radar antenna this and other range intervals show shifts in the times of the typically is held at a fixed position while acquiring the data. The dot- reflectivity peaks relative to the return strokes. Many of the ted region represents a volume of space in which time variations of abrupt changes in reflectivity due to lightning are coincident reflectivity may be studied in detail. Typically, this volume can be with return stroke field changes, but some are not. In the moved in range in 150 m increments.

Unauthenticated | Downloaded 10/07/21 08:25 AM UTC Bulletin American Meteorological Society 1291

FIG. 7. Top: Time history of a lightning radar echo from a range interval, AR, at 86 km and 4 km altitude, 3 May 1979 (79i23) at 0002:52 CST. Bottom: Slow antenna waveform recorded simultane- ously with radar data. The relatively slow field change at the begin- ning of the flash (at 0.3 s) suggests an IC flash. The more abrupt FIG. 8. Sketch of observed locations and polarities of CG flashes changes, denoted/?, are typical of cloud-to-ground return strokes as from severe thunderstorms. The spiral denotes the region of intense negative charge is effectively brought to earth. This flash was inde- updraft and rotation. Only negative CGs have been observed in the pendently identified as a CG with at least eight strokes by a lightning precipitation core. The + CGs seem to constitute only a very small strike location system. percentage of the total flashes to ground.

lightning echoes that we have observed, there often appear to served occasional +CGs during the dissipation stage of be reflectivity changes associated with phenomena that can storms. be identified from AE records, e.g., return strokes, K-type We have observed +CGs during both the severe and final changes, and events during continuing currents. stages of severe storms. +CGs with documentation in addition to AE records number less than 100 prior to 1980, but we have obtained numerous other waveforms suggestive of d. Positive CGs in severe storms -bCGs. During a 7 min interval of the 6 June 1979 storm (See Although several investigators have detected positive CG Section 3a), six + CGs were documented. During the active flashes (+CG) in mountainous terrain and to tall structures,- stage of a severe storm on 20 May 1980, several flashes ema- e.g., Berger (1977),+CGs to level terrain were thought to be nating from well out in the downwind anvil were positive, rare. On the other hand, Takeuti et al. (1978) have observed a while those from the downwind anvil, but close to the main predominance of + CGs in the highly sheared winter storms tower, were of both polarities (Fig. 8). From a limited in the Hokuriku coastal region of Japan. Others have ob- number of visually confirmed CFs, we think that CG light-

FIG . 9. Oversimplified depiction of apparent differences between very small (airmass, nonsevere) storms and supercell, severe storms. Other differences are likely.

Unauthenticated | Downloaded 10/07/21 08:25 AM UTC 1292 Vol. 62, No. 9, September 1981 ning emanating from high on the main storm tower under the H. Golde, Academic Press, London, pp. 119-190. upwind (back-sheared) anvil and from the mesocyclone re- Brown, R. A., and H. G. Hughes, 1978: Directional VHF sferics gion of severe storms can be positive (Arnold and Rust, from the Union City, Oklahoma, tornadic storm. 7. Geophys. Res., 1979). The percentage of-fCGs during a severe storm is ap- 83, 3571-3574. , D. W. Burgess, J. K. Carter, L. R. Lemon, and D. Sirmans, parently small. It is intriguing that positive CGs can occur 1975: NSSL dual Doppler radar measurements in tornadic prior to, as well as in, the dissipating stage of the storm and storms: A preview. Bull. Am. Meteorol. Soc., 56, 524-526. thus may be directly related to a particular phase of severe Browning, K. A., 1977: Hail: The structure and mechanisms of hail- storm development. storms. In A Review ofHail Science and Hail Suppression, edited by G. B. Foote and C. A. Knight, Meteorol. Monogr. No. 38, AMS, Boston, pp. 1-39. Colgate, S. A., 1967: Tornadoes: Mechanism and control. Science, 4. Concluding remarks 157, 1431-1434. , 1968: Electrical heating of a tornado vortex. 7. Geophys. Res., Our attempt to study electrical activity in its relationship to 73, 6121. other parameters within severe storms began in 1978. We be- Davies-Jones, R. P., 1981: Tornado interception with mobile teams. lieve that both individual, selected events and case studies of In Thunderstorms: A Social, Scientific, and Technological Docu- storms should be examined to relate electrical processes to mentary, Vol. 3, edited by E. Kessler, U.S. Government Printing the evolution of large storms in the central United States. Office, Washington, D.C., in press. , and J. H. Golden, 1975: On the relation of electrical activity to Based on first results of our study and on findings of other tornadoes. 7 Geophys. Res., 80, 1614-1616. investigators, we present in Fig. 9 a depiction of apparent dif- , and E. Kessler, 1974: Tornadoes. In Weather and Climate Mod- ferences in characteristics of thunderstorms with extremely ification, edited by W. N. Hess, John Wiley and Sons, New York, different scales, ranging from the very small (airmass, non- pp. 552-595. severe) to the supercell severe storm. Undoubtedly, the figure Donaldson, R. J., 1970: Vortex signature recognition by Doppler is a simplification, and there is a wide variety of combina- radar. 7 Appl. Meteorol., 9, 661-670. tions of storm sizes and characteristics between those shown. Frier, G. D., 1959: The earth's electric field during a tornado. 7 Me- It appears that there are differences in electrical phenomena teorol.., 16, 333-334. beyond a mere scaling of the activity in a small storm to that Goncher, A. F., S. M. Galperin, and V. N. Egorov, 1977: Possibili- in a large severe storm, e.g., the occurrence of + CGs in the ties of increasing the accuracy of the lightning azimuth determination with radars. (In Russian). Issues Leningrad Poly tech. Inst., mature stage of a severe storm. Other parameters may also be 64, 129-134. different and are being investigated. While we are continu- Gunn, R., 1956: Electric field intensity at the ground under active ally sensitive to the prospect of possible unique electrical thunderstorms and tornadoes. 7. Meteorol., 13, 269-273. characteristics of severe storms, we feel that our goal of Hewitt, F. J., 1953: The study of lightning streamers with 50 cm correlating electrical features with the dynamics and precipi- radar. Proc. Phys. Soc. London, B., 66, 895-897. tation of storms is the most important function of the present , 1957: Radar echoes from inter-stroke processes in lightning. program. Proc. Phys. Soc. London, B., 79, 961-979. Jones, H. L., 1951: A sferics method of tornado identification and tracking. Bull. Am. Meteorol. Soc., 32, 380-385. Acknowledgments. G. Moore provided forecasting and guided the , 1965: The tornado pulse generator. Weatherwise, 18, 78-80. mobile laboratory for storm intercept. S. Goodman provided VHF Kinzer, G. D., 1974: Cloud-to-ground lightning versus radar reflec- mapping and Doppler information for the 20 June storm. We thank tivity in Oklahoma thunderstorms. 7. Atmos. Sci., 31, 787-799. D. Burgess for his suggestions. J. Weaver spent many hours engaged Kitagawa, N., M. Brook, and E. J. Workman, 1962: Continuing cur- in nowcasting for storm intercept. Electrical measurements were rents in cloud-to-ground lightning discharges. 7. Geophys. Res., made with the help of K. Cameron, L. Cunningham, S. Horsburgh, 67, 637-647. M. MacNiven, and C. Watson. M. Maier provided the CG strike lo- Krehbiel, P. R., M. Brook, and R. McCrory, 1979: An analysis of cation data. There are many others at NSSL, too numerous to men- charge structure of lightning discharges to ground. 7. Geophys. tion, whose continual help keeps the storm electricity program func- Res., 84, 2432-2456. tioning. We thank M. Brook of New Mexico Tech for the loan of slow and fast antennas to make our initial measurements. Lemon, L. R., and C. A. Doswell III, 1979: Severe thunderstorm This study was supported in part by the Severe Storms and Local evolution and mesocyclone structure as related to tornadogenesis. Weather Research Office of the National Aeronautics and Space Mon. Wea. Rev., 107, 1184-1197. Administration under Government Order No. H-39299B and the At- Ligda, M. H., 1950: Lightning detection by radar. Bull. Am. Mete- mospheric Sciences Program of the Office of Naval Research under orol. Soc., 31, 279-283. Government Order Nos. N00014-79-F-0023, N00014-80-F0039, and , 1956: The radar observation of lightning. 7. Atmos. Terr. Phys., N00014-76-C-0169. D. R. MacGorman participated in this study as 3, 329-346. a postdoctoral Research Associate of the National Research Council. MacGorman, D. R., 1978: Lightning location in a storm with strong wind shear. Ph.D. dissertation, Rice University, Houston, Tex. Miles, V. G., 1952: Radar echoes from lightning. Nature, 170, 365-366. References , 1953: Radar echoes associated with lightning. 7. Atmos. Terr. Phys., 3, 258-262. Arnold, R. T., and W. D. Rust, 1979: Initial attempt to make electri- Proctor, D. E., 1974: VHF radio pictures of lightning. Spec. Rept. cal measurements on tornadic storms by surface intercept. Pre- TEL 120, Council for Scientific and Industrial Research, Johan- prints, 11th Conference on Severe Local Storms (Kansas City), nesburg, S. Africa. AMS, Boston, pp. 320-325. Ray, P. S., R. J. Doviak, G. B. Walker, D. Sirmans, J. Carter, and Berger, K., 1977: The earth flash. In Lightning, Vol. 1, edited by R. B. Bumgarner, 1975: Dual-Doppler observation of a tornadic

Unauthenticated | Downloaded 10/07/21 08:25 AM UTC Bulletin American Meteorological Society 1293

storm. J. Appl. Meteorol., 14, 1512-1530. discharge processes. J. Geophys. Res., 83, 3575-3583. Scouten, D. C., D. T. Stephenson, and W. G. Biggs, 1972: A sferic Teer, T. L., and A. A. Few, 1974: Horizontal lightning. J. Geophys. rate azimuth-profile of the 1955 Blackwell, Oklahoma, tornado. Res., 79, 3436-3441. J. Atmos. Sci., 29, 929-936. Vonnegut, B., 1960: Electrical theory of tornadoes. 7. Geophys. Res., Szymanski, E. W., and W. D. Rust, 1979: Preliminary observations 65, 203-212. of lightning radar echoes and simultaneous electric field changes. , 1968: Inside the tornado. Natur. Hist., 77, 26-33. Geophys. Res. Lett., 6, 527-530. , and C. B. Moore, 1957: Electrical activity associated with the

, S. J. Szymanski, C. R. Holmes, and C. B. Moore, 1980: An Blackwell-Udall tornado. Jt Meteorol., 14, 284-285. observation of a precipitation echo intensification associated with ,and , 1959: Giant electrical storms. In Recent Advances in lightning. J. Geophys. Res., 85, 1951-1953. Atmospheric Electricity, edited by L. G. Smith, Pergamon Press, Takeuti, T., M. Nakano, M. Brook, D. J. Raymond, and P. Kreh- London, pp. 399-410. biel, 1978: The anomalous winter thunderstorms of the Hokuriku Watkins, D. C., J. D. Cobine, and B. Vonnegut, 1978: Electric dis- Coast. J. Geophys. Res., 83, 2385-2394. charges inside tornadoes. Science, 199, 171-174. Taylor, W. L,, 1973: Electromagnetic radiation from severe storms Wilkins, E. M., 1964: The role of electrical phenomena associated in Oklahoma during April 29-30, 1970. J. Geophys. Res., 78, with tornadoes. J. Geophys. Res., 69, 2435-2447. 8761-8771. Zrnic', D. S., 1976: Magnetometer data acquired during nearby tor- , 1978: A VHF technique for space-time mapping of lightning nado occurrences. J. Geophys. Res., 81, 5410-5412. •

announcements1

USGS Index to Information on U.S. Water Symposium on Hydrological Processes of Forested Areas The most complete compilation of information on surface water and ground water data collected at more than 100 000 sites across the A symposium on hydrological processes of forested areas will be United States and bordering areas of Canada and Mexico is now held at the University of New Brunswick in Fredericton, N.B., on available in a new edition of the index to the catalog of information 14-15 June 1982. This is the first symposium which is explicitly on water data, released by the U.S. Geological Survey (USGS), devoted to this topic in the series of meetings, sponsored by the Department of the Interior. Associate Committee on Hydrology (ACH) of the National Based on information provided by hundreds of participating fed- Research Council of Canada. eral, state, and local agencies across the country, the index presents Abstracts in either English or French for papers for any of the selected information from the catalog, a comprehensive computer- following topics related to the hydrology of forested areas are ized file of information about water-data acquisition activities in the invited: Forests and Water Quality; Physical Processes in Forested United States, its territories and possessions, as well as activities in Basins; Snow and Snowmelt; Integrated Basin Response; and parts of Canada and Mexico. Forests and Floods. The proposed paper title, an extended abstract The catalog does not contain the actual data but provides infor- for review (200-300 words), and a short abstract for the program mation on where and by whom data are being collected, the type or (50-100 words) are to be submitted by 15 January 1982 to: Prof. R. types of data acquired, and how these data can be obtained. B. B. Dickison, Chairman; Canadian Hydrology Symposium: 82; As in previous editions, the information presented in the new c/o Department of Forest Resources; University of New Brunswick; seventh edition is published in 21 regional volumes, representing the Bag Service Number 44555; Fredericton, N.B., Canada E3B 6C2 18 largest drainage basins of the United States, plus Alaska, Hawaii, (tel: 506-453-4501). and Puerto Rico. These 21 regions, designated by the U.S. Water The final program will be based on the submitted abstracts and on Resources Council, include such river basins as the Missouri, Missis- a few invited papers. The detailed program will be mailed to prospec- sippi, and Rio Grande. tive attendees in late February, 1982. Completed papers must be Each regional volume is divided into four informational sections: submitted by 1 June 1982, must not exceed 20 pages, and may be in streamflow and stage; quality of surface water; quality of ground either English or French. If a paper is selected for the symposium water; and areal investigations and miscellaneous activities. program, it is expected that the author (or a substitute) will make a Copies of any of the 21 separate volumes of the Index to the presentation of the paper at the symposium. Catalog of Information on Water Data are available from: Office of Contributions are also solicited for poster sessions. Titles and de- Water Data Coordination, U.S. Geological Survey, 417 National scriptions should be submitted by 15 April 1982. Details concerning Center, Reston, Va., 22092. poster sessions may be obtained from the Symposium Chairman at the address above.

'Notice of registration deadlines for meetings, workshops, and seminars, deadlines for submittal of abstracts or papers to be presented at meetings, and deadlines for grants, proposals, awards, nominations, and fellowships must be received at least three months prior to deadline dates.—News Ed. (continued on page 1307)

Unauthenticated | Downloaded 10/07/21 08:25 AM UTC