SR/SSD 97-55

12-15-97

Technical Attachment

The Heavy Rain and Flood Event of June 21-22, 1997

in South Central Texas and the Texas Hill Country

Robert Blaha and Bruce Thoren

NWS Austin/San Antonio, Texas

Introduction

During the early morning hours of June 21, 1997 through the evening of June 22, 1997, a subtropical storm system produced excessive rain over parts of south central and central Texas, particularly over the Hill Country. The heaviest rain fell over Bandera and Kendall Counties, where storm total amounts ranged from 15 to near 20 in, with the highest total of 19.72 in in eastern Bandera County. A few estimates exceeded 20 in, where gauges overflowed, and exact measurements were unknown. Another concentrated area of 12 to 18 in of rain fell over the northern Hill Country in portions of Mason, Kimble and Gillespie Counties. The result of the excessive rain was record flooding along stretches of the Llano River in Llano County. Major river flooding also occurred on the Frio, Nueces, Sabinal, Guadalupe, Lavaca, and Navidad rivers.

Heavy rain is not uncommon for central or south central Texas. It is part of the general climate of the area, where rainfall amounts vary considerably from east to west and year to year. Some notable heavy rains the area has experienced include an unofficial report of 48 in over a period of about 52 hr. in early August, 1978, resulting from the remnants of tropical storm Amelia and an upper-level disturbance. This rain fell near the town of Bluff, located northwest of San Antonio in the Texas Hill Country. A short-term heavy rainfall rate of 21.87 in in 2 hr, 45 min occurred in Medina County, just north of D'Hanis, in late May, 1935. On average, south Texas can expect on 20 in rainfall event and several 10 to 20 in events per year (Grice and Maddox 1982).

Upper-Level Features

At 0000 UTC on Friday, June 20, 1997, a relatively cold pocket of air was located at 500 mb over southern New Mexico, with an open wave at 300 mb. To the east, high pressure at 850 mb was centered over the eastern Gulf of Mexico. By 0000 UTC on June 21, the cold pocket at 500 mb was associated with a closed low and extended into the upper troposphere. The center of circulation was over west center Texas. Downstream, a ridge at 500 mb extended from the western Gulf of Mexico into the middle Mississippi Valley, while another mid and upper-level low was located across the Mississippi Valley strengthened. As a result, the upper low over west central Texas drifted southeast and became quasi-stationary over the Hill Country (Fig.1). A strong low-level jet (40-50 kt at 850 mb) developed over the eastern half of Texas during the evening of June 21 and the morning of June 22 (Fig. 2). This pattern allowed a deep tropical flow of warm, moist air over central and south central Texas. The upper low moved slowly northward and began to fill on June 22, as an upper trough moved into the Great Basin.

Surface Features\Radar Summary

During the afternoon and early evening of June 20, scattered thunderstorms developed along a surface trough/dryline in west and southwest Texas. By late evening the thunderstorms had developed into a line and continued moving toward the Edwards Plateau. Strong gusty winds and locally heavy rains accompanied the storms. The line continued an eastward push during the early morning hours of June 21, while southerly onshore winds along the Texas coast pumped rich, tropical moisture (dewpoints in the mid and upper 70s) into south central Texas. By 1200 UTC June 21, the line of storms had pushed through the Hill Country and continued through portions of south central Texas.. Light to moderate rain with embedded thunderstorms formed behind the line, and satellite pictures indicated that it was a part of a Mesoscale Convective System (MCS).

During late morning into the afternoon on June 21, thunderstorms intensified over interior sections of the Coastal Plains on an outflow boundary left behind from the MCS. This complex of storms produced flash flooding during the afternoon, with parts of Dewitt County (northwest of Victoria) receiving up to 11 in of rain. Over the Hill Country and south central Texas, thunderstorms continued to develop as a result of upper-level forcing and a developing mesohigh that was centered between San Antonio and Austin.

Throughout the day Saturday, the mesohigh and attendant outflow boundaries strengthened. Spiral bands developed around the low aloft and a band of drier air aloft was brought in from the west. The drier air aloft likely enhanced the convective instability and helped reinforce the storms and their associated outflow boundaries. The storms were very pronounced on the 0.5 deg reflectively WSR-88D Elevation Scan during the late afternoon. By early afternoon, thunderstorms had produced 10 in rains in Bandera and Medina Counties.

During the early evening hours, areas of rain and thunderstorms near the coastal bend and over Lavaca, DeWitt and Karnes Counties in south central Texas dissipated. After these storms dissipated, widely scattered showers and thunderstorms became more numerous over the Hill Country on the night of June 21 and morning of June 22. With the help of the upper-level features interacting with low-level features, torrential rains became well established on the cool side of the meso boundaries over the Hill Country (Figs. 3 and 4). The Balcones Escarpment, which runs from Uvalde County to Travis County, also aided in lift. Eight to 12 in of rain fell over parts of Bandera, Blanco, Gillespie, Kendall, Kerr, Medina, Real and Uvalde Counties during the early morning, quickly causing life-threatening flash flooding and river flooding in those counties.

As sunrise approached, the area of heavy rain shifted east and south into Bexar, Comal and Hays Counties (Fig. 5). By Sunday afternoon, June 22, as daytime heating increased, tropical showers and thunderstorms replaced the concentrated area of thunderstorms. During early evening a band of thunderstorms developed over Mason, Kimble and Gillespie Counties, as the peripheral tropical showers dissipated. Unlike the previous night, this band of intense rain moved slowly northward, as the upper low did, and affected the extreme northern parts of the Hill Country. The Llano River near Llano responded immediately with a record stage, near 39 ft, and major flooding.

Conclusion/Summary

Since heavy rain events are so common across south central Texas it is operationally imperative to recognize pattern that are associated with them. The June, 1997 rain event consisted of an open wave that stalled in weak westerlies aloft, then developed into a closed low. This closed low became nearly stationary over the western Texas Hill Country on June 21 and part of June 22. On the night of June 22 and early morning of June 23, the low was lifted north by an approaching trough from the west. Rain amounts varied from near 9 to near 20 in over the Hill Country to between 5 and 11 in over south central Texas. The least amount of rain fell in the Del Rio area west of the low. As rain developed, it formed its own mesohigh north of spiral bands around the low. Each meso boundary was overrun with a low level jet, developing more rain. In addition, at night, heavy rains tended to concentrate at the center of the low. This pattern has been associated with core rains that are commonly observed with warm core tropical systems (Ward 1981).

The Meso Eta and Eta models forecast a heavy rain event further east and southeast of the area. Although still generally underforecasting rain amounts, the Aviation model handled the rain forecasts better, showing where the rain would fall. The 12 hr forecast from the 1200 UTC Aviation model run of June 21 revealed where the rain would fall, however, it missed the 11 in rain over DeWitt County. The 24 hr forecast also located the area where the rain would fall, however, it underestimated the amounts. As in many other cases, it is difficult for the models to depict mesoscale features that uniquely evolve during the weather event.

Mesoanalysis and trends become very important in determining the severity of the rain event threat. Figure 6 shows the storm total rain for this event from selected sites across the Hill Country and south central Texas. The models showed the upper air features quite well. Using 500 mb forecasts and pattern recognition, plus mesoscale trends, forecasters can gain superior insights as to where and how heavy rain will fall. This method of using the model as a tool and combining its output with logical decisions, goes a long way in attempting to make a good heavy rain forecast.

A long-range clue to the possibility of heavy rain events to come in the spring/early summer of 1997 was evident during the late winter and early spring over south Texas, as west winds did not bring the usual 80 and 90 deg days. Instead the area experienced rather cool 60s and 70s. The trend of cooler air over the southwest U.S. moving east toward Texas was in place. This scenario combined with dynamic troughs and pointed to active spring weather patterns. Such observations sometimes establish a trend that can foretell the coming weather events. In addition, the transition period between spring and summer is always carefully watched as the strong westerlies retreat to the north and the prevailing easterlies south of the subtropical high begin to evolve. The case on June 21-22, 1997 showed an upper-level open wave moving through the weak westerlies, then stalling to become a closed low over the western Texas Hill Country. In the spring to summer transition, such features in the weak westerlies aloft, along with other heavy rain patterns not discussed here, have made May and June (along with September), some of the wettest months for south central Texas since the late 1800's.

Acknowledgments:

The authors would like to thank several people for their help and input in putting this paper together. John Patton, the service hydrologist, provided valuable input and rainfall data. Jimmy Don Ward and Larry Peabody provided valuable input in reviewing this paper and providing alternative adjustments that were incorporated into the text. In addition, the use of the local area network (LAN) at the NWSFO Austin/San Antonio was a great resource. Thanks go to Mark Oliver and his staff for setting up and maintaining this network of valuable software for use in making professional reports. Carl Morgan gave assistance in making the graphic for total storm rainfall. Al Dreumont (MIC) and Esther Atkins (DMIC) reviewed the paper and provided support and ideas in making professional presentations and reports.

References:

Grice, G.K., and R.A. Maddox: Synoptic Aspects of Heavy Rain Events in South Texas Associated With the Westerlies. NOAA Technical Memorandum, NWSSR-106, Fort Worth, Texas.

Ward, J.D., 1981: Spatial and Temporal Heavy Rainfall Patterns Overland Associated With Weakening Tropical Cyclones. AMS Preprint, Fourth Conference on Hydrology, 174-180.