by Mark West, Jeffrey A. Chapman and Ronald T. Holmes
edited by Tom Hamilton
The following story is based on a seminar presentation by Mark West, Aerostar International. The material was originally developed and posted on the Internet, Clear Air Radar and Automated Surface Observations of a Long Lived Gust Front, by Jeffrey A. Chapman and Ronald T. Holmes, National Weather Service Forecast Office, Sioux Falls, South Dakota.
On June 8, 1997 the Great Plains Balloon Race was being held just south of Sioux Falls. The evening balloon flight was expected to be rather tranquil. Instead, a surprise gust front from a two isolated thunderstorms caused some balloons to land with speeds approaching 30 mph.
The outflows from these two small storms traveled for four to five hours merging and intensifying as they reached Sioux Falls. The post event analysis by Messrs. Chapman and Holmes have added a much better understanding of events on the day in question and provide additional considerations for preflight information gathering.
The general weather synopsis for June 8, 1997 showed a high pressure ridge extending from north central Iowa to southeast North Dakota. There is a low pressure ridge in Ohio and another in Colorado, relatively light pressure gradients and relatively light winds. The winds in Sioux Falls were in the four to seven mile per hour range, clear blue skies, and a great day and weekend for ballooning. See figures one and two.
Figure three illustrates the weather balloon sounding taken on Saturday morning in south central Minnesota. This was in the area where the high pressure was located. In the lower portion of altitude the sounding shows a very normal adiabatic lapse rate, very stable, very broad air. Then all of a sudden there is a huge inversion. At this level there is an unusually large amount of moisture. This is located almost right in the center of the high pressure.
Figure four shows that there is a cloud deck at 8,000 feet and it is a relatively thin layer of moisture.
Balloonists in Sioux Falls have developed a wonderful relationship with the National Weather Service office. One of the interesting pieces of equipment they have is a Doppler Radar. There has been a two year learning curve to get to the point where the information NWS gives to balloonists from the Doppler is truly good information.
Balloonists would get up, call weather service, and ask what are the winds looking like. They would say at 3,000 feet we have winds 290 degrees at 42. You would go outside, put a pibal up, and it would just go straight up. You are faced with this horrible decision, do you go fly when the winds aloft forecast in 42 or do you pack it up on their recommendation and lose a perfectly good day to fly balloons?
We started double checking their reports before we would go out to fly. They also began figuring out what they were doing wrong reading and interpreting the Doppler information. When they are working in clear air the only thing they can see, literally, is bugs. If there are no bugs the reliability of their results is very poor.
For the Great Plains Balloon Rally the National Weather Service had two people at the balloon field to provide information. They were in contact with the personnel back in the office.
Figure five shows the Doppler Radar for about 3 p.m., June 8 in Sioux Falls. The wind that day was from approximately 080 degrees. Looking at the orange dots you will notice that they appear to have a wave pattern to them.
Think about the waves you see at the lake on a breezy day. The waves are created by the non-laminar flow of air across the surface of the lake. The wind both generates and pushes the waves perpendicular to the air flow.
The clusters of orange dots in figure five are "wind waves." The National Weather Service has named these waves "horizontal convective roll."
When there is thermal activity, it doesn't have to be strong thermal activity, and the prevailing wind is from a specific direction, the upward thermal movement, and corresponding downward movement, is sheared off by the wind. It turns this movement of air into a roll, basically becoming a wave as it rolls across the ground. Then there is the next one and the next one and the next one.
This helps to explain why on a beautiful, calm day there may be a little gust and a drop back to calm. Then a little gust and a drop back to calm. That is horizontal convective roll.
The high temperature in Sioux Falls on this particular day was about 78 degrees, not a particularly hot day.
Figure six is a depiction of the whole idea of a horizontal convective roll. In this particular diagram a sea breeze is being used to show the effect. But, you do not need a sea breeze, any relatively constant wind in a given direction will set up this particular condition. Warm areas with rising air and corresponding areas of sinking air will be sheared off by the wind and generate the roll.
Figure seven is taken at 3 p.m. from a long range radar sounding. You can see two small areas of convective activity in Minnesota. Distance from Sioux Falls to farthest convective area is approximately 175 miles.
What developed that day was the thin layer of moisture ended up with enough instability to develop into a thunderstorm. It was relatively small and not associated with any major system. When it started raining, the rain came down into a very dry stable air mass. What happens when rain falls through dry air? Virga.
The rain evaporates. But what happens when rain evaporates? Cooling, significant cooling. What happens when you cool a whole bunch of air? It sinks.
So, this little old thunderstorm put out what is called an outflow boundary. It setup a wave. The wave wound up being carried in the direction of the prevailing surface winds, which was the opposite direction that the thunderstorm was moving.
Figure eight is a surface chart that was constructed after that fact. NWS did not have this as a tool to determine what was going on that day. This data was derived from various sources and then put together as a backward looking document to determine what was really going on.
There were two small thunderstorms, figure seven, and each one of them put out an outflow boundary. Both established themselves as cold fronts. The cold fronts were moving from northeast to southwest. The distance between the two cold fronts was approximately 75 miles. Time, about 4 p.m.
Winds associated with these cold fronts were not very high at this point in time. They were definitely waves of energy that were coming through. Sioux Falls is about 75 miles from the closest front in figure eight. The closest front was moving toward Sioux Falls at about 15 miles per hour. The trailing front was moving approximately 35 to 40 miles per hour. At this point in time there was clear air and no weather coming with it.
Figure nine is a topographical depiction of this part of the country. This particular area is known as the Buffalo Ridge formation. Over this area there is a slight increase in elevation from the northeast where the ground is approximately 900 feet above sea level to the southwest where Sioux Falls is 1400 feet above sea level.
The fronts are moving into a rise. The opinion is that the elevation gain helped to magnify the frontal movement.
The National Weather Service briefer called back to the office at 5:55 p.m. to update his information and the pilot briefing started at 6:00 p.m.
Figure ten shows what the Doppler Radar looked like at 6:08 p.m. The line of orange dots on the right looks just like all the other horizontal convective rolls. Guess what? It was the first gust front moving into the Doppler picture. It was eight minutes after the briefing began and balloons were taking off about 45 minutes after the briefing. The National Weather looked at this and didn't think it was anything more than a little horizontal convective roll. It just happen to extend out to the left and right a little farther than the others.
There are two different types of pictures that the Doppler Radar provides, inflow and outflow. Inflow is not something that NWS looks at a lot, normally they are looking at outflow. Figure eleven is an inflow look at the same view as the preceding figure. From this you can see that there truly is a difference between the gust front and the rest of the horizontal convective rolls.
Figure twelve shows the next constructed chart that was put together. Now you can see that instead of the 75 mile separation between the two fronts the second one is starting to overrun the first one. Back behind the fronts the winds are more than 20 miles per hour. In front they are less than ten miles per hour. This particular chart represents 7:00 p.m. The hare balloon and most of the hounds were off the ground at that point in time.
Figure thirteen is the next picture from the Doppler Radar taken at 7:06 p.m. You can now see that the first gust front has become much more pronounced and that the second gust front is just moving into the picture. Because of the type of product, these images do not show the velocity of the wind. And this is not something that NWS was looking for. The winds on the backside of the front were about 26 knots.
Figure fourteen is taken when the second gust front hit. At this point in time, all but four of the 43 balloons were on the ground. With the exception of the hare, all the rest of the balloons landed in 12-15 knot winds.
During the briefing the winds were expected to be 6-7 knots, 3-4 knots at sunset, and by midnight, calm. That is what was expected to happen. Normally it does happen that way. Sometimes, however, the winds stiffen a little at sunset and stay that way.
Mark West's crew had just gotten packed up. The door of the trailer was still open and the second gust front hit. The door was forced out of a hand, hit Mark, and knocked him down. The thought, "what the heck was that?" There was still one balloon in the air coming straight at the crew. It flew over at about 300 feet, low enough to see the anxiety in the eyes of everyone in the basket. How to reach up and pull them down and get them out of the air?
Again, blue sky, clear air, and no associated weather.
On the backside of the second front winds speeds of 35-40 knots.
Fortunately there were no reported injuries. One balloon did report landing in a "fresh" pasture. The basket dragged for 300 yards before coming to a stop.
The two gust fronts continued to move across South Dakota and were about half way across the state, before it finally died out. South Dakota is about 400 miles wide. The front traveled in excess of 350 miles. Most of the rest of the travel was after sunset.
A pilot, looking at all the information that was available, really didn't have any clues that said this should be a bad deal. The weather services guys didn't look at anything that said, "Oh my God, we had better look out for something here because, there is something bad about to happen."
Whenever a great big huge thunderstorm collapses, it will put out a gust front. And, most of the time the gust front that it puts out is in the direction that the thunderstorm is moving. The gust front goes shooting out in front of the thunderstorms and leads the storm's path. In this case, because the downflow was coming into an air mass that didn't have the same path as the storm, it sheared off and flowed with the prevailing wind. We don't have cold fronts in South Dakota that come from the northeast.
The next morning winds were still at 20 knots and it was cold back out to the balloon field. It had turned into a chilly night.
What would we do differently? It is a question that we have all asked ourselves. Now when we call for a summertime afternoon weather briefing we ask if there is any kind of convective activity. If there is, start asking questions about it and whether something could be coming this way.
In Sioux Falls we have worked very closely with our National Weather Service office to develop a good working relationship and understanding of useful weather information. As a result of this experience the weather people now take a closer look at the inflow picture.
This situation was very unusual in that thunderstorms generally don't build in the middle of a high pressure center, unless there is lots of moisture in the air and lots of day time heating. We have a greater respect for the downdraft that a Virga can generate as a result of this incident.
Prior to this, we thought 100 miles away was safe enough from a thunderstorm. In fact, if the system was 50 to 75 miles and moving away we thought that was okay. We are a lot less certain of that now. Anytime the briefers say that there is thunderstorm activity, we start asking more extensive questions. Like what are other reporting stations seeing? It's a good idea to call weather service and receive a final report, after getting to the launch field and before starting the fan.
The report Clear Air Radar and Automated Surface Observations of a Long Lived Gust Front can be found on the Internet at http://www.crh.noaa.gov/fsd/gust.htm. The Internet story includes two Doppler Radar movie loops which show the gust fronts approaching and passing through Sioux Falls.