Credit: Dallas Morning News

Meteorological Overview

The national surface map on the morning of August 2, 1985, at 7:00 A.M CDT revealed that a cold front extended from the east coast back into the southern plains. The figure (to the left and in the link) are courtesy of the National Oceanic and Atmospheric Administration (NOAA) Library. For a larger image of the surface map, click here or on the image itself. Temperatures at 7:00 A.M. CDT were already quite mild with readings in the upper 70s ahead of the southward moving cold front. Cold fronts are not too common during this time of year, due to the lack of strong temperature contrasts that drive these features. Oftentimes, the lack of any appreciable cloud cover and/or precipitation results in sufficient modification of the air mass behind the front. This in turn decreases the horizontal temperature contrast and thus effectively weakens the front. However, in the case of August 2, 1985, the widespread cloud cover and light rain behind the front (as noted by observations at Oklahoma City, Tulsa and Fort Smith [not shown]) kept afternoon temperatures fairly cool for the summertime (readings in the 80s and low 90s). The areas of scattered clouds and light precipitation likely served to reinforce this temperature contrast. The Weather Service Meteorological Observatory (WSMO) at Stephenville, TX, was responsible for radiosonde observations during this era of the NWS. The radiosonde observation was taken twice daily (1200 UTC [7:00 A.M. CDT] and 0000 UTC [7:00 P.M. CDT]). On the morning of August 2, 1985 (left panel in figure below), the radiosonde data revealed modest instability, but very weak vertical wind shear, something fairly usual for the summer. With daytime heating, the lowest several thousand feet of the atmosphere heat up and convective currents develop. These convective currents tend to "mix" the atmosphere and result in a decrease in the moisture content and an increase in the temperature near the surface. This is manifest in the increase in the theoretical cloud base height as can be seen when comparing the 1200 UTC (theoretical cloud base of 4,900 feet) and 0000 UTC (theoretical cloud base of 10,000 feet) data. Unknown to scientists at the time, the temperature and moisture profile plotted on a diagram such as a Skew-T Log P as seen in the right panel in the figure below, is actually favorable for microbursts (dependent on whether a sufficiently strong shower or thunderstorm develop).   Stephenville, TX, Radiosonde Observations 12 UTC (7:00 A.M. CDT) Weather Balloon Data from Stephenville, TX 00 UTC (7:00 P.M. CDT) Weather Balloon Data from Stephenville, TX Click one of the above thumbnails for a larger image

Visible Satellite Imagery at 5:35 P.M. CDT

Visible Satellite Imagery at 6:05 P.M. CDT

Visible Satellite Imagery at 6:35 P.M. CDT

Stephenville WSR-57 Radar imagery at 5:28 P.M. CDT

Stephenville WSR-57 Radar imagery at 5:48 P.M. CDT

Stephenville WSR-57 Radar imagery at 6:04 P.M. CDT

 

Aft section of Delta 191 post-crash landing at DFW. Credit: Dallas Morning News

Interior of aft section of Delta 191. Credit: Star-Telegram

6:17 P.M. CDT Stephenville Radar Imagery

Imagery from the Stephenville radar indicates that the weather echoes have likely decreased from VIP 3 or "strong" perhaps downward to VIP 1 ("weak") or 2 ("moderate"). While a slight weakening of the echo may have been observed, it is likely that very strong winds are continuing across the northern sides of DFW airport. Due to the non-Doppler capabilities and the distance between the Stephenville radar and the precipitation echoes, sampling the magnitude of these strong winds would have been virtually impossible. It does appear plausible that an additional core of very strong winds barreled through the north side of DFW and over the crash site of Delta Air Lines Flight 191 during this time. Moreover, the weather observer contractor transmitted another non-routine weather observation at 6:14 P.M. CDT which revealed a rapid increase in winds of up to 42 M.P.H. with wind gusts to 53 M.P.H. These winds are likely on the leading edge of a core of the microburst that downed Delta Air Lines Flight 191. The strongest core of winds appeared to have arrived at the observation site at DFW 8 minutes later (6:25 P.M) when the observer recorded an 80 M.P.H wind gust. By 6:35 P.M., visible satellite imagery reveals only a few remnant areas of showers and perhaps towering cumulus clouds.

The Aftermath and Lessons Learned

Improved Forecast Methodologies

Prior to the Delta Air Lines Flight 191 incident at DFW International Airport the field phase of a research project, the Joint Airport Weather Studies (JAWS) aimed at studying microbursts had just completed. The JAWS project, however, targeted mainly dry microbursts across the high plains of Colorado. A new project, the MIcroburst and Severe Thunderstorm (MIST) experiment, would look at environments supportive of wet microbursts. This project occurred during the summer of 1986 and examined convection across the southeast United States. Research from both JAWS and MIST revealed that certain thermodynamic environments lend themselves to two primary classes of microbursts: wet and dry. These classes are a product of processes that drive or enhance the downdraft of the thunderstorm, the environment in which the downdraft formed and the amount of precipitation produced at the surface. Dry microbursts develop as precipitation particles either melt or evaporate below the very high cloud base. Melting or the subliming of ice particles or the evaporation of liquid particles can help to drive a very strong downdraft. As the name implies, little to no precipitation occurs with dry microbursts. Conversely, wet microbursts can form when very dry air at mid levels is mixed with the moist air along and inside the cloud. This results in evaporation necessary to create a strong downdraft. Precipitation amounts of at least one-tenth of an inch or greater typically accompany wet microbursts. In addition,  a third class of known as "hybrid" microbursts has often been used to characterize microbursts that exhibit characteristics of both wet and dry microbursts. The figure below summarizes key findings from the data collected via various research projects (e.g., JAWS and MIST).

Conceptual model of dry (left) and wet (right) microburst. Figure adapted from Wakimoto and Atkins (1991)

From the figure above, it was theorized that the factors that drive or enhance the thunderstorm downdraft are related to the melting or evaporation of precipitation particles(hail/graupel or rain) below the cloud base (dry microburst) or entrainment of dry air into the convective cloud, creating cooler denser air that sinks rapidly through the cloud. Thermodynamic diagrams (Skew-T Log P) were determined to be very useful in determining whether or not conditions are favorable for microbursts. In addition, Wakimoto and Atkins (1991) examined various thermodynamic diagrams (Skew-T Log P) and plots of Equivalent Potential Temperature or theta-E to diagnose environments favorable for microbursts. The figures below are plots that essentially summarize the data gathered from radiosondes during the MIST project. The project determined that environments for microburst were often characterized as those having very weak deep layer wind shear and weak to modest instability. In addition, this difference in "theta-E" is thought to be important for the development of a strong downdraft. This difference can be maximized when low theta-E air exists at mid levels, while large amounts of high theta-E air is located in lower levels of the atmosphere. As seen in the composite profiles, morning soundings on "microburst days" usually exhibit a fair amount of low level stability as noted by the bulge in the temperature (solid black line in panel A) and theta-E (solid black line in panel B) towards the right side of the figure. Composite soundings from MIST revealed that afternoon "microburst days" were characterized by a low level temperature and moisture profile that exhibited an upside down V or "inverted V" and a sufficient difference in mid- and low-level theta-E values as seen in panels C and D, respectively. During the MIST project, it was determined that a 20K difference in theta-E was a good discriminator between microburst and non-microburst days (assuming conditions are favorable for convection).

A/B show typical Skew-T & Theta-E vs. pressure profiles on the A.M. of a "microburst day" C/D show typical Skew-T & Theta-E vs. pressure profiles on the P.M. of a "microburst day".

12 UTC (7:00 A.M. CDT) data from Stephenville, TX

00 UTC (7:00 P.M. CDT) data from Stephenville, TX

Theta-E vs. Pressure profile for 12 UTC and 00 UTC data from Stephenville, TX.

Click one of the above thumbnails for a larger image

Improved Detection

The Terminal Doppler Weather Radar (TDWR)

The Terminal Doppler Weather Radar or TDWR, is a type of radar system that has been deployed at 45 locations across the United States and Puerto Rico. This system was developed at Lincoln Laboratories, a division of the Massachusetts Institute of Technology, back in the late 1990s. These radars were strategically placed at airports (see map) that were particularly susceptible to microbursts and other low-level atmospheric phenomenon that may be hazardous to aircraft. The TDWR are maintained by the FAA, but are used heavily by NWS and DoD personnel as well. The TDWR operate in 1 of 2 modes - monitor or hazardous mode. While in monitor mode, the TDWR typically takes around 5 to 6 minutes to complete a sampling cycle. The monitor mode is somewhat analogous to the WSR-88D's "clear air mode", in which the radar is placed in a state in which rapid low level updates are not necessary. When the TDWR detects precipitation echoes of a certain strength or detects wind shear associated with a developing microburst, the hazardous mode is activated. While in this mode, the TDWR can provide one minute updates to FAA staff, NWS/DoD meteorologists and pilots. According to the Lincoln Laboratory, no commercial airline accidents caused by wind shear have occurred at airports in which a TDWR is present and operating in the U.S. In addition to providing improved detection of weather hazardous to aircraft and airports, the TDWR offers an additional tool for NWS meteorologist to utilize to issue severe thunderstorm and tornado warnings! The figures and movies below compare and contrast the evolution of a microburst that occurred over Dallas, TX, as viewed from both the TDWR at Dallas Love Field (TDAL) and the WSR-88D at the Fort Worth Spinks Airport (FWS) on the afternoon of June 9, 2015. The right thumbnail (radial velocity) is particularly useful for meteorologists as it displays a "classic" downburst signature.

  Dallas Downburst on 06/09/2015 (Click thumbnail for larger view)

Reflectivity for Dallas Downburst as seen from TDAL (as seen in yellow dot) at 2100 UTC (4:00 P.M. CDT) on 06/09/2015

Velocity for Dallas Downburst as seen from TDAL (as seen in yellow dot) at 2100 UTC (4:00 P.M. CDT) on 06/09/2015

 

  GIF Animations (animations may take a few moments to load) of Dallas Microburst on 06/09/2015 as viewed from  TDAL between 2100-2130 UTC
(1 Minute Radar Data)

Click here for a GIF animation of radar reflectivity from TDAL on 06/09/2015

Click here for a GIF animation of radial velocity from TDAL on06/09/2015

 

In the velocity figure (from TDAL at 2100 UTC or 4:00 P.M. CDT), annotated TDAL radial velocity, note that the winds within the orange circle are spreading apart or diverging outward (as shown by the arrows) with some winds flowing towards the top left portion of the figure (towards the northwest) and some winds flowing towards the bottom right portion of the figure (towards the southeast). This signature is one that often tells meteorologist that a downburst is in progress. With the one minute data, one can clearly observe the complete evolution of the microburst from the TDWR. The thumbnails below show an animation of the part of the lifecycle of this microburst near Dallas Love Field Airport as seen from TDAL. Contrast that with radar imagery from the annotated KFWS radial velocity, the divergent wind pattern is not clearly as defined due to both the longer distance between the KFWS radar and the thunderstorm, but also the larger gaps in time between volume scans. The Dallas microburst on 06/09/2015 produced a 58 knot (67 M.P.H.) wind gust just after 4:15 P.M. CDT. For a map of some of the associated damage, click here Dallas Microburst Storm Reports. This storm report map was created with assistance from the Iowa Environmental Mesonet.

 

GIF Animations (animations may take a few moments to load) of Dallas Microburst on 06/09/2015 as viewed from KFWS between 2100-2130 UTC
(5 Minute Radar Data)

 

Click here for a GIF animation of radar reflectivity from KFWS on 06/09/2015

Click here for a GIF animation of radial velocity from KFWS on 06/09/2015

 

In addition to improvements in ground based remote sensing of these hazards, the FAA mandated that all commercial airliners have the capability to detect low level wind shear. Airline companies had approximately 5 years to comply with this new regulation.

 

The Integrated Terminal Weather System (ITWS)

The Integrated Terminal Weather System or ITWS is a tool that integrates a variety of meteorological instrumentation in order to support airport and aircraft operations. The ITWS combines information from surface weather observation stations, the TDWR, the Next Generation Radar (NEXRAD), aircraft observations, lightning networks, low level wind shear alert systems and Air Surveillance Radars (ASRs). Many of the algorithms that the ITWS uses were developed by engineers and scientists at Lincoln Laboratories. Full implementation of the ITWS occurred in 2001 and today services many airports across the U.S. (including DFW International and Dallas Love Field Airports). By utilizing information from each of these systems, the ITWS is able to extrapolate conditions between 30 and 60 minutes into the future, offering critical information to decision makers at airports. The ITWS has an algorithm that is dedicated to the detection of low level wind shear associated with microbursts. No tool as powerful as the ITWS Microburst Detection algorithm existed prior to the late 1990s. In addition to low level wind shear associated with microbursts, the ITWS has the capability to diagnose and extrapolate the position of gust fronts or wind shift lines that can be used to plan various runway configurations. ITWS also provides information on lightning within a certain range of the airport as well as information critical for determining the intensity and direction of non-thunderstorm low level winds. Much of this information is viewed by FAA personnel at Air Route Traffic Control Centers (ARTCCs); however, this information is often packaged and can be sent to pilots directly for their own situational awareness and decision making. In addition, the ITWS is displayed at many Center Weather Service Units (CWSUs),which are co-located with ARTCCs, and meteorologists there can often provide additional service via their expert analysis of the meteorological data.

 

Memorial

The following inscription can be found on a memorial site located at Founder's Plaza located on the northwest side of DFW International Airport. It reads:   On August 2, 1985, Delta Air Lines Flight 191, a Lockheed L-1011, crashed on final approach to the Dallas-Fort Worth International Airport, approximately 2 miles due east of this site. A total of 135 persons died in the accident, including 126 of the 152 passengers on board, 8 of 11 crew members and one motorist on the ground. The National Transpiration Board investigation deemed the meteorological phenomenon known as wind shear pushed the aircraft to the ground well short of the runway. Investigators also cited the aircrafts lack of ability to detect wind shear as a contributing cause. In 1988, the Federal Aviation Administration ordered all commercial aircraft to have on-board wind shear detection systems installed by 1993. Many airports across the United States, including DFW, also installed ground-based wind shear detection equipment. Due to the improved ability to detect wind shear, the number of aircraft accidents caused by wind shear has dropped significantly since 1995.