17 0 BULLETIN AMERICAN METEOROLOGICAL SOCIETY Some Effects of Vertical Wind Shear on Thunderstorm Structure * HORACE R. BYERS and Louis J. BATTAN U. S. Weather Bureau, Thunderstorm Project ABSTRACT Observations of thunderclouds obtained with a 3-cm height-finding radar set are used to obtain a description of the vertical shear of thunderclouds. Several photographs are given which show the shearing of the radar clouds. A scattergram of wind shear plotted against echo shear is presented and shows that the two variables are related, with the former exceeding the latter in almost all cases. Scatter-diagrams are given which verify that strong vertical wind shear tends to restrict the growth of thunderstorms. A series of radar cross sections illustrates the displacement of the upper part of a thundercloud which is subjected to wind shear, and the growth of another cloud column from the lower part of the thundercloud. 1. INTRODUCTION 2. SOURCE OF DATA HE primary aim of the Thunderstorm Prior to the utilization of radar for meteorologi- Project has been to obtain a precise de- cal observation, it was difficult to obtain the type Tscription of the thunderstorm. In a recent of cloud observational data which would permit paper Byers and Braham have given a detailed dis- the measurement of the dimensions of most cussion of the structure and circulation of a typical thunderstorms. When isolated cumuliform clouds thunderstorm [1]. The proposed model, however, develop during the daytime, it is possible to take assumes that there is no vertical wind shear. photographs of a particular cloud; however, when there are many low clouds or the visibility is poor, When this simplifying assumption is not made, a photographic techniques using visible light become description of the thunderstorm becomes more unfeasible. This type of data permits qualitative complex. It is the purpose of this paper to demon- determinations of cloud behavior, but quantitative strate some of the effects of wind shear on the calculations are often difiicult and in many cases structure of the thunderstorm. impossible. Since the thundercloud is not a rigid body but By using a radar set with a range-height- rather consists of a volume of small particles (ex- indicator (RHI), it is possible to obtain direct cept when hail is present), having little mass and measurements of the horizontal and vertical extent subject to displacement in the wind field, it is of the thunderstorms as detected by the equip- reasonable to expect that if, for example, the wind ment [9, p. 3-6]. The characteristics of radar as flow in the upper levels is from the north while an instrument for observing clouds are such that that in the lower layers is from the south, the there is closer correspondence between the radar upper part of the cloud will move towards the cloud and the actual cloud during the developing south relative to the lower part and may eventually rather than during the dissipating stages of the blow off if it does not evaporate before this occurs. thunderstorm. This results from the fact that dur- Meteorologists have long recognized that pro- ing the former, a large quantity of water droplets nounced vertical shear often inhibits the growth and frozen particles are being carried to the higher of cumulus into thunderstorms. When the wind portions of the cloud, as well as being held aloft shear varies continuously in magnitude and is con- within the cloud, while in the latter case the ab- stant in direction from the base to the top of the sence of strong updrafts permits the precipitation thunderstorm, the axis of the cloud will have a of the large hydrometeors to the lower levels, leav- tendency to bend or tilt in the direction of the shear ing only the smaller particles in the upper part of the cloud. The progressive depletion of the larger vector [2, p. 7]. If the water content and size scatterers results in a radar echo which tends to distribution of hydrometeors are such as to permit "fade out" at the edges rather than cut off sharply. radar detection of the cloud for a relatively long Since it is improbable that the reflectivity of the period of time, the shearing of the storm can be surface layers of the clouds is sufficiently high to seen on a radar 'scope. produce a detectable echo on the scope of conven- tional "S"- and "X"-band radar sets, the radar * Published Report No. 8 of the Thunderstorm Project. Unauthenticated | Downloaded 10/10/21 04:26 PM UTC VOL. 30.. No. 5, MAY, 1949 169 Thoroughout this study the effects of rain at- tenuation were considered only in a most qualita- tive fashion. Although it is realized that in some cases the magnitude of the attenuation was large, the unavailability of power measurements and rec- ords of the settings of the radar controls made the determination of accurate corrections impossible. 3. QUALITATIVE DEMONSTRATION OF RADAR CLOUD SHEAR When the wind shear is strong along the azi- muth of an RHI cross section through a thunder- storm, the effects of the shear become very evi- dent. FIGURES 2a and 2b are typical examples of such observations. The cross section of FIGURE FIG. 1. Drawing of data available on each photograph 2a was towards an azimuth of 219° while that of of the 'scope of Radar Set AN/TPS-10. The vertical FIGURE 2b was towards 353°; they were photo- lines are 10 miles apart; the nearly horizontal lines are graphed within nine minutes of one another. No 5000 feet apart. upper wind data were available in the immediate vicinity of the thunderstorms; however, the winds cloucl will be smaller than the actual cloud in al- aloft report from Huntington, West Virginia * most all cases even during the developing stages.f (2200Z) indicated that between the levels of 10,- The data used in this study were extracted from 000 feet and 25,000 feet the vertical wind shear the photographic records of the range-height- was towards 170° at 15 mph. Although the re- indicator (RHI) 'scope of the Radar Set AN/ lease was made six to seven hours later than the TPS-10 16]. As the name implies the 'scope pre- radar observations, it is felt that the synoptic situa- sents slant range (straight line distance between tion was such as to permit the use of this wind for radar set and the target) horizontally against a qualitative comparison. It can be seen that the height vertically (FIG. 1). To facilitate the meas- bending of the storms is in the proper direction urement, an overlay was made on which were if the components of the shear along the respective drawn lines of equal horizontal distance from the azimuths are calculated. radar site. From the available RHI 'scope photographs, it was possible to obtain a number of vertical cross t See References [3], [4] and [5] for detailed discus- sions of radar-weather theory. * Approximately 110 miles southeast of the radar site. FIG. 2. Photographs of the RHI 'scope showing the tilting of radar clouds which are located at azimuths from radar site which differ by about 135 degrees. (Fig. 2a, left; Fig. 2b, right.) Unauthenticated | Downloaded 10/10/21 04:26 PM UTC 17 0 BULLETIN AMERICAN METEOROLOGICAL SOCIETY sections through the same thunderstorm. Since azimuth as well as in feet to make it directly com- the vertical "slices" were at intervals of three to parable to the vertical axis. The azimuth of 174° four degrees of azimuth and about two seconds in was selected as the reference from which to lay time, it was possible, by assembling the data ob- off the scale in feet since it was close to the center tained during one azimuthal scan of the radar an- of the storm. tenna through the storm, to construct a three- This thunderstorm occurred on August 25, 1947 dimensional model of the radar cloud. FIGURE 3 and was located about 30 miles from the radar site. presents a series of approximately east-west cross The vertical wind shear computed between the sections obtained by extracting the necessary data levels of 15 and 30 thousand feet at about the from the three-dimensional drawings. The dashed same time as the cloud observations indicated a lines are isolines of equal radar cloud thickness (3 component towards the west of 6 mph. The ef- miles) measured into the plane of the paper. From fects of the wind shear in blowing off the top of the FIGURES given it is impossible to determine the the thunderstorm are strikingly illustrated. An- extent of any portion of the radar cloud on either other interesting feature which is worthy of men- side of the central east-west vertical plane. The tion is that in this case the dissipation of the radar horizontal axis has been labeled in degrees of cloud proceeded most rapidly at the middle and FIG. 3. Series of approximately east-west cross-sections through a radar thundercloud. The azimuth scale is in degrees from north relative to the position of the radar site. The time is indicated in hours, minutes, and seconds. Unauthenticated | Downloaded 10/10/21 04:26 PM UTC VOL. 30.. No. 5, MAY, 1949 171 probably be explained by similar wind shear con- siderations [7, p. 67]. 4. MEASUREMENTS OF VERTICAL CLOUD SHEAR In order to show the relationship between radar cloud and wind shear, the vector difference be- tween the wind velocity at two selected altitudes and the rate of horizontal displacement of the radar cloud at the upper altitude relative to the lower were measured for a number of storms and corresponding values were compared.