(Reprinted from Preprint Volume: Eleventh Conference on Severe Local Storms, Kansas City, KS

October 2-5, 1979. Published by the American Meteorological Society, Boston, Mass.)


Using satellite imagery, the rate of thunderstorm anvil growth has been related to storm severity by Sikdar (1970) and Auvine (1973) and pulsating anvil growth has been measured with a severe storm by Fujita (1978). On the afternoon of 11 July 1978, spectacular anvil behavior was observed with a thunderstorm over eastern Kansas. An investigation into the circumstances influencing the evolution of this storm included an analysis of satellite imagery, radar data and moist static energy fields in an attempt to understand why it did not become severe.


The evolution of the thunderstorm on the afternoon of 11 July 78 began as a weak surface low pressure center moved into western Kansas (Fig. la). A stationary front extended eastward from the low and became a cold front. Moisture was being advected northward from the Gulf of Mexico as evident by high dewpoint temperatures. During the afternoon, temperatures throughout northern Oklahoma approached 40C and dewpoint temperatures neared 23C. Most of the cloudiness was confined to north of the stationary front where temperatures were moderate, near 25C .

Moist air also was being advected up over the stationary front through the 850 mb level (Fig. lb). A short-wave trough extended southeastward from a low in southern Canada to southern Nebraska. A pronounced thermal trough was entering western Kansas, moving eastward by late afternoon. The presence of these disturbances triggered the subsequent thunderstorm activity. Meanwhile, a strong ridge extended from Louisiana up into the Midwest.

At the 700 mb level, the strongest winds were located over Topeka, associated with a wide region of warm air advection (Fig. lc). diffluent zone existed between Topeka and Omaha while the short-wave trough at 850 mb was still apparent at 700 mb over western Kansas.


Analysis of the 500 mb level reveals an isolated pocket of moisture extending eastward through southern Kansas (Fig. ld). A strong ridge extended through eastern Kansas up into Canada with light winds and weak temperature gradients. During the morning, eastern Kansas was under the influence of positive vorticity advection; by afternoon however negative vorticity advection prevailed. Siebers et al. (1975) showed that vorticity advection can be related to severe weather occurrences. Thus, as the day progressed it became more unlikely that severe weather would organize. Relative humidity in excess of 70% and precipitable water near 1.70 in. existed over Topeka.

Prior to daybreak, a large thunderstorm complex in central Missouri had generated an extensive outflow. During the day, the westward progression of this air became defined by an arc of towering cumulus. With the passage of the gust front, surface pressure rises averaged 3 mb and temperature falls of 6C were common.

Satellite imagery allows the westward progression of the arc of cumulus activity to be followed hourly (Fig. 2). The estimated speed of the gust front was approximately 35 km/hr. As the day progressed, several perturbations were noticed in the line of cumulus activity; the most numerous at 2300 GMT. The cumulus began to dissipate by late afternoon as the larger convection organized; thus the leading edge of the gust front became less defined.

Below 500 mb, the Topeka sounding for 2300 GMT (Fig. 3) resembles the Fawbush and Miller (1954) Type I, characteristic of severe weather. Above this level, temperatures in the Topeka sounding were warmer than the range given for the Type I sounding. Thus, this factor was detrimental to the production of severe weather.


The thunderstorm (A) began to form along the southern edge of a cloud boundary (C) to the west of the advancing gust front (B to B'), as illustrated in Figure 4.

At 2030 GMT (Fig. 4a), the leading edge of the gust front lay across eastern Kansas and Oklahoma, depicted by an arc of towering cumulus. The region to the east of the boundary was relatively cloud free. A stream of cumulus immediately west of the gust front indicated the advection of low-level moisture northward from the Gulf of Mexico. Along the gust front, cirrus anvils appeared due to the isolated thunderstorm activity. Winds above 800 mb were westerly, therefore moving the thunderstorms over the advancing gust front; they then decayed. At this time the thunderstorm at A in eastern Kansas began to form.

By 2130 GMT (Fig. 4b), the thunderstorm developing, containing several cells. Clearing of the low-level cloud mass in eastern Kansas was evident at C resulting from surface heating around the cloud edge. Meanwhile, the advancing gust front was enhancing the moisture convergence into the region surrounding A.

Within an hour (2230 GMT), the anvil had expanded rapidly into a scallop-shape while other cloud activity began to diminish (Fig. 4c). By late afternoon (2330 GMT), the gust front had been greatly modified with bulges occurring along the leading edge. Thunderstorm A continued to develop rapidly to the west of the gust front (Fig. 4d).


Parallel, wave-like bands appeared on the anvil by 2430 G~T (Fig. 4e). Fujita (1977) has termed similar features the "herringbone pattern". They are perhaps manifestations of Kelvin-Helmholtz waves (Black, 1977) due to the windshear across the anvil. The western edge of the low-level cloud mass also revealed thunderstorm-induced waves.

The dual planimetric method described by Fujita (1978) was used to determine the anvil area at half-hour intervals; along with the anvil growth rates, these are given in Table 1. A pulsation of the anvil was observed during the initial stages of the storm's growth. A similar pattern was found for a severe storm studied in Wisconsin by Fujita (1978).


Anvil Growth

TIME Area Growth Rate

(GMT) (km2) (km2/min)

2030 233

2100 1,577 44

2130 7,417 194

2200 17,459 324

2230 26,426 298

2300 39,593 438

2330 59,065 650

2400 81,318 740

2430 104,035 756


The evolution of thunderstorm A was recorded by Kansas City radar (Fig. 5).

At 2030 GMT (Fig. 5a) a disorganized precipitation area (A) lay southwest of Kansas City with the most intense cell Just south of Manhattan, moving slowly eastward. Among the numerous cells embedded within the echo at 2130 GMT (Fig. 5b), the leading cell was the most intense. At this time the storm veered to the southeast, to the right of the moan environmental wind.

Isolated level 5 regions appeared by 2230 GMT (Fig. 5c) as the cells merged. Meanwhile, a separate cell (D) developed along the gust front boundary west of Joplin. This storm decayed rapidly Just after reaching 56,000 ft. Thunderstorm A attained its greatest intensity at 2330 GMT (Fig. 5d). Its movement was to the south, almost perpendicular to the mean environmental flow. During the next hour it weakened substantially as solar heating diminished (Fig. 5e).

The total rainfall from the storm is shown in Figure 6. The path of the storm began Just south of Manhattan at 2030 GMT and coincided closely with the highest isohyets. The maximum rainfall recorded was 75 mm. The isolated isohyet maximum to the southwest of the storm track may be attributed to a second short-lived storm evident at 2130 GMT on the satellite imagery.


The occurrence of severe weather has been linked to the pattern of surface moist static energy (Darkow, 1975). From hourly surface data, temperature (T) and mixing ratio (w) can be obtained, allowing the computation of the moist static energy.

θ (K) = T (K) + 9.8 z (km) + 2.5w (g/kg)

where z is the station altitude.

Generally, lower values may be expected with thunderstorm outflow areas and furthermore, contrasts will develop with variations in solar heating.

The hourly evolution of moist static energy is shown in Fig. 7. At 2000 ~MT (Fig. 7a), the moist static energy field displayed a narrow gradient demarcating the gust front boundary. The greatest values occurred across the cloud-free region in southern Kansas, whereas smallest values were within the cooler and drier air beneath the cirrus anvil in southeastern Missouri. By 2100 GHT (Fig. 7b), the gradient of moist static energy had steepened across the gust front boundary due to the differential heating.

By 2200 GMT (Fig. 7c), the effect of the outflow from thunderstorm A had become apparent. Values of moist static energy began to decrease near Emporia while higher values con-tinged to feed into the southern portion of the storm. By 2300 GMT (Fig. 7d), the gradient of moist static energy continued to steepen as the gust front moved through eastern Kansas. During the next hour an isolated minimum coincided with the precipitation area, produced from the thunderstorm north of Emporia (Fig. 7e).



Over a four-hour period on the afternoon of 11 July 1978, the areal extent of a massive anvil topping a storm in eastern Kansas was monitored using GOES-imagery. The anvil appeared scallop-shaped with a rippling wavy surface. Successive measurements from satellite imagery yielded a maximum growth rate exceeding 700 km / min (270 mi2/min) during the final stages of development. During the life of the storm a distinct pulsation was noted in the anvil growth rate.


The isolated thunderstorm developed near the intersection of two boundaries. A gust front had been produced that morning by a large thunderstorm complex in central Missouri. This boundary had a well-defined arc shape with towering cumulus activity along the leading edge, followed by relatively clear skies. The arc cloud had several perturbations which became more numerous with time as it moved westward at 35 km/hr. In addition, a cloud boundary was located just north of a stationary front extending from a weak surface low-pressure center in western Kansas.


Hourly surface reports indicated temperatures greater than 38C with dewpoints exceeding 21C preceding the gust front. Within the gust front region there was a transition zone averaging 150 km wide with pressure rises of 3 mb and temperature decreases of 6C. High values of surface moist static energy were found feeding into the southern portion of the storm.


Radar observations indicated that during the initial stages of development, storm cells had merged while moving eastward; subsequently the storm veered toward the southeast and rapidly intensified. An influx of low-level moisture from the south in the lowest 100 mb had destabilized the atmosphere enough to produce the thunderstorm activity. However, the upper-air patterns - in particular the evolution to negative vorticity advection, warm air advection and light winds - proved detrimental to the generation of severe weather. Thus, neither hail nor high winds were reported with this storm.


This research was supported in part by the U.S. Nuclear Regulatory Commission under Contract NRC-04-77-016.


Auvine, B., and C. Anderson, 1972: The use of cumulonimbus anvil growth data for inferences about the circulations in the thunderstorms and severe local storms. Tellus, 26, 1001-1015.

Black, P., 1975: Cover photograph. Bull. Amer. Meteor. Soc., 56, 1239.

Darkow, G., and R. Livingston, 1975: The evolution of the surface static energy fields on 3 April 1974. 9th Conf. on Severe Local Storms, 264-269.

Fawbush, E. J., and R. C. Miller, 1954: The types of air masses in which North American tornadoes form. Bull. Amer. Meteor. Soc., 35, 154-165.

Fujita, T. T., 1978: Manual of downburst identification for Project Nimrod. SMRP #156, 56 pp.

_____., and M. Hjelmfelt, 1977: Meso-analysis of record Chicago rainstorm using radar, satellite and rainguage data. 10th Conf. on Severe Local Storms, 65-72.

Siebers, J., F. Hidalgo, S. Tegtmeier, and M. Young, 1975: Guide for using GOES/SMS imagery in severe weather forecasting. United States Air Forcer AWS Tech. Memo., 56 pp.

Sikdar, D., V. Suomi, and C. Anderson, 1970: Convective transport of mass and energy in severe storms over the United States -an estimate from a geostationary altitude. Tellus, 22, 521-532.