openCortex: ARTICLE: Weather Forecasting in Sunrayce 93 (by Jeff Zoltowski)




		May 16 4:59am PDT

HOME	Login \| New user? Register		Help
	Don't know what openCortex is? Learn about it.


		Home » Maize & Blue, Weather Forecasting and Australia (by Jeff Zoltowski) » Weather Forecasting in Sunrayce 93 (by Jeff Zoltowski)


	PARENT CONTENT Maize & Blue, Weather...

LINKED PROFILES Nazeeri, Furqan University of Michiga... Maize & Blue	ARTICLE Weather Forecasting in Sunrayce 93 (by Jeff Zoltowski) OCID cn00000405	LINKED CATEGORIES Race Logistics Racing Strategy Strategy Weather Forecasting

	LINKED CONTENT


				Link to...

		+Article \| +Photo

Weather Forecasting in Sunrayce 93 (by Jeff Zoltowski)
by Furqan Nazeeri
created 6/20/1993
submitted 1/18/2004 07:56:32 AM

This paper recounts Michigan's use of weather forecasting as a competitive weapon in Sunrayce '93. Jeff was one of 3 weather forecasters on the team and their collective ability to forecast weather was instrumental in Michigan's victory. An excellent read!

Introduction
The first solar car race in the United States took place in June of 1990. Sunrayce's eleven day route traced its way across the eastern third of the country, meandering from Orlando, Florida, to Detroit, Michigan. This first race was won by the University of Michigan, and won by a comfortable margin, at that. By the fall of 1990, even as that first team was preparing for the World Solar Challenge in Australia, a second team was forming. The challenge for the new team: defend Michigan's national title in Sunrayce '93, an eleven hundred mile race over seven days from Arlington, Texas, to Minneapolis, Minnesota.
Successful weather prediction was critical to Michigan's victory in 1990, especially when it came to dealing with stormy weather that was encountered early in the race. The advantage that Michigan had gained by being one of the few teams to have its own meteorologists did not go unnoticed by the organizers of the new team. Strategy and meteorology go hand in hand in solar car racing, and the dominance that Michigan had demonstrated in these two areas would prove just as crucial in 1993.
Early Work
I joined the University of Michigan Solar Car Team in the October of 1992, joining Joseph Bartlo, a graduate student who had been on the 1990 team, and Mark Kulie, an undergraduate who had joined the team a few months earlier. Actually, I had been approached by a team member the previous spring about participating in a survey of the race route that summer. However, the decision was made to bring Mark on the survey instead, since he was already in Ann Arbor for the summer, whereas I was home working in Sterling Heights, Michigan. When I joined, the tasks of the weather team was twofold: first, studying the data learned from the race route survey, and second, researching every weather topic we could to prepare ourselves for some of the phenomena we might encounter during the race.
The participants in the race route survey had along with them several solar cells of the type to be used on the array. The data acquired from the performance of the cells in varying amounts of sunshine, along with the data from the radiometers we had, were used is the eventual construction of several computer programs. These FORTRAN programs modeled both the attenuation of solar radiation by the atmosphere under varying sky conditions, and the response of the solar cells to this radiation. We were able to have a preliminary characterization of how the array of the new car, named Maize & Blue, would perform, long before the array was actually constructed. This would allow us to begin practicing weather forecasting under something approaching race conditions early in 1993.
Also on our agenda was continual research of every possible meteorological variable and phenomena that could impact the performance of Maize & Blue. Some of topics the three of us studied during the fall and winter of 1992':93 were low level jet, convection and convective clouds, severe storms, hail, and wind variability. Of course, we concentrated our research on variables that impact solar radiation fluxes in the atmosphere, such as aerosols, ozone, water vapor, the earth-sun geometry, and obviously clouds and cloud formation.
I was not seriously involved at this stage with research on the solar radiation topics. Instead, I was concentrating on the low level jet, and especially on fog. Fog was something that the race route survey had encountered in the southern United States, and the successful prediction of both its formation and dissipation was of concern for the rest of the team. I prepared a short, one-page paper introducing the strategists to the topic, and what the meteorologists would be looking for in regards to its behavior. Indeed, at this time period (November 1992), I personally was not all that interested in being on the race team. At that stage, I thought of my work as being almost like that of a consultant, lending a hand when Mark and Joseph had a little too much on their hands.
But by January 1993, the situation within the team had changed. In particular, Mark Kulie had become concerned that his class and work situations were suffering because of the time he had been putting into the Solar Car Team. I began to assume a more prominent role on the team, helping with the formation of the solar radiation forecasting computer programs, as well as looking at the equipment the weather team would need for the race. I had a change of motivation as well, as I began to actively pursue a spot on the actual race team, the group of twenty or so students who would actually race the car in June. I tried to impress on team leaders the benefits of having as many meteorologists along as possible, noting that consensus forecasts among knowledgeable meteorologists tend to be better than those of ANY of the individuals. My advice was soon heeded; the
race team was selected in February, and all three of us - Joseph, Mark, and myself - were selected to be race team members.
Preparing the Weather Team
On March 1st, the three of us 'officially' began to practice solar radiation forecasting under near-race conditions. We had unofficially been practicing forecasting for at least month, but we saw little practical benefit of those practices, since the mechanisms involved in weather would be so very different come June. But with the approach of spring, Joseph and I (Mark's job commitments did not allow him free time until the evening) would arrive at the Space Research Building early on Monday, Wednesday, and Friday evenings. Each of us would spend approximately an hour in producing a solar radiation forecast, hour by hour, for that day; we would also generate general weather and solar radiation forecasts for the next two days. I felt that I lagged behind Joseph in experience regarding this kind of forecasting, so by the end of month, I was coming in to do a forecast Monday through Friday. It soon became apparent, however, that even though our forecasts were made with no input from the other meteorologists, our forecasts were almost always very similar. But it also became obvious who was the pessimist and who was the optimist. It came as no surprise to me that when we differed, I was almost always the one with the lowest solar radiation forecast. I guess I had been around Michigan winters too long.
These practice forecast helped the three of us gain confidence in each other's abilities and weather forecasters. Unfortunately, virtually all of our work was independent of what the rest of the team was doing. hl fact, there were I lot of members of the race team that I had never even met. That changed at the unveiling of Maize & Blue at the Power Center at the end of March. The car was shown to the public for the first team, and the race team was formally announced during a short ceremony. Afterwards, members of the press and public were allowed on stage to talk to team members and to get a closer look at the car. I spent most of that afternoon, however, trying to talk to as many of my teammates as possible. I wanted them to be able to trust me and my abilities as a weather forecaster. For this same reason, I decided to attend the qualifier, which was just over a week later.
The Eastern Qualifier for Sunrayce '93 took place in Indianapolis, Indiana, from Thursday, April 5th, through Saturday, April 10th. Starting on the Monday before, I was asked by our head strategist, Eric Slimko, to attempt to forecast the weather for the Saturday, the day of the actual qualifying run. It was obvious even that far ahead of time that a frontal system would be moving through the area that weekend, and I was unsure of the timing for a while. By Thursday, though, I had decided that Friday - the day of scrutineering, when the car would be checked to make sure it passed all the race regulations - would be the rainy day. Saturday looked very sunny, which would be good for the qualifying run, which consisted of 50 miles - 20 laps - around Indianapolis Motor Speedway. The forecast ended up being right on target. Friday was a cold, windy, rainy, and thoroughly miserable day. (I had to stand outside most of that day, and even my raincoat couldn't protect me from the ear infection I came down with the following week.) Saturday was just the opposite; sunny and very warm, with the temperature getting above seventy degrees. Unfortunately for us, Maize & Blue had a wiring problem in its array that was not discovered until just days before the qualifier. Because of this, the array was not functioning during the qualifier, and with the sun shining all day long, we were at a significant disadvantage. Maize & Blue had to complete the fifty miles solely on battery power. Nonetheless, we still averaged about forty miles per hour during our run. This placed us, after the results from the Western Qualifier in Phoenix, Arizona, the following weekend, in tenth position. California State University at Los Angeles (Cal State-LA) had captured the pole.
Other than qualifying, the month of April was fairly uneventful for the meteorologists. Mark and I had finals to worry about, and so neither of us had much time to do more than continue our practice forecasts and our research. This left May to be a very busy month, especially since I had previous class commitment that kept me away from Ann Arbor during the first week of the month. When I got back, I had to get busy acquiring the data sources and instruments we would need. The team owned two
radiometers, so we had what we needed in that area, but we were woefully deficient in more basic meteorological instrumentation. I acquired a barometer, a wet-dry bulb thermometer, and an anemometer from various companies during the month.
The most important thing we needed, however, was a commercial weather data service. Other members of the team were responsible for the computers and other electronic equipment, but it was my job to establish an account at a service. After considerable debate, the three meteorologists decided on WSI Corporation. Their Weather for Windows software promised to provide all the data we needed, including satellite images, radar composites, surface observations, and DIP AX maps. The software also looked to be easy to use and configure, which was crucial, since we did not have a lot of time to learn complicated commands. We had the software and an account by the middle of the month. After some initial difficulties with modem incompatibilities, the WSI software proved to be just about everything we could have hoped. What deficiencies there were would only become obvious after the race.
Race Setup and the "Spy"
By this time, we had also decided on the caravan setup for the race. The weather vehicle, called the "Scout" vehicle, was a Mercury Villager. Due to space restrictions from our extensive computer equipment especially a very large printer - only three people, including a driver, would fit inside. This would not be a problem, however, since it had been decided that I would not be in this vehicle.
Our information from WSI had to come in through a telephone transmission. If you have ever talked over a cellular phone, you know that even under good cellular coverage, the signal is still much noisier than a call over a line. When trying to receive data over a phone line, this noise needs to be eliminated, if possible, since one bad bit can ruin an entire line of data, if not the entire picture being transmitted. The modems we had available were extremely effective at filtering out most of noise. However, it was decided to have one of the meteorologists at a stationary location - namely, a hotel room - so that the data could be transmitted over a phone line, or at worst, through a cellular phone that was stationary. The obvious place for this meteorologist would be at each day's finish line, since he could then also keep track of conditions at the after-racing charging location. It was then logical for this meteorologist to be concerned with long-range weather forecasting, and have the two in the Scout vehicle be concerned with the short-range, i.e. the next two or three hours. The long-range weather forecaster would seem to need a reliable data source more anyway, since theoretically the short -range forecasters would only have to look at the sky around them to do at least a respectable job.
Thus was born what came to be known as the 'Spy' position. This became my job for the race. I must admit I was somewhat apprehensive about being the Spy - I rather selfishly wanted to be with the caravan, instead of waiting for it at each day's finish line - but I soon came to appreciate that it was easily the best spot for me. I considered weather forecasting, not solar radiation forecasting, to be my strongest area, and the long-range forecaster would not have to be as concerned about solar radiation as the Scout crew would. Additionally, long-range forecasts that I made in Mayas a test impressed many people, not the least of which was myself. By the end of May, I had fully embraced the concept of the Spy position.
There is a rather interesting spotlight to the creation of the Spy position. By having a Spy, the team required another vehicle. Only an hour or so after Maize & Blue crossed the finish line for one day's stage, I had to head out to the next day's finish line so that I could be ready bright and very early the next morning. So the team requested a vehicle suitable for the job of transporting me and my equipment from Ford Motor Company, which provided all but one of the rest of our vehicles (including the previously mentioned Scout vehicle ). We weren't expecting all that much; an Escort maybe, or perhaps a Taurus, or if were really lucky, some sort of Probe. Imagine my surprise, then, when I found out that I would be driving a Lincoln Mark VIII - a $40,000 automobile (!) - from Texas to Minnesota. I have no idea who decided to give a vehicle like that to a group of college students, but believe me the entire team got the chance to put the car through its paces and then some. We would, in fact, get that chance during the weeks we would spend testing in Texas. And with the caravan set up and all of our equipment procured, it was time indeed time to head south.

San Angelo, Texas
During the winter months, a small group of team members had come to San Angelo, Texas, to test the chassis of Maize & Blue when it was completed. The Goodyear Tire Company had made one of its test tracks at its proving grounds in San Angelo, Texas, available to us when weather closed Michigan International Speedway for the winter. Goodyear was also kind enough to make the same track - a five mile circular oval that they did not use anymore - available to us for the first two weeks of June. By June 3rd, just about the entire team had arrived in San Angelo. One exception was Mark, as he was required by his job to stay in Ann Arbor another couple of weeks before he could join us. We were there until the 13th, testing Maize & Blue and all of our equipment, and getting everything into race condition, including ourselves.
Those first two weeks of June were, in my opinion, the hardest on the team members, both physically and psychologically. The heat was oppressive - our first day there the high temperature was 105 degrees, and never cooled down much from that - and the working conditions were brutal. Eighteen hour gays were the norm, and twenty was not all uncommon. This was not the way to recover from the twenty-four hour drive it took to get to San Angelo from Ann Arbor. And things kept going wrong with Maize & Blue. If the motor controller magnets were not malfunctioning due to the heat, then the tires were blowing out, and if not that, then some of the electronics in the car were failing.
The weather team was in no way immune from equipment failures. Our laptop computers did not handle the stress we put on them very well, as one had a built-in mouse fail, and another stopped working completely, necessitating an emergency call back to Michigan to get a replacement. More serious than that, though, was that we were unable to test our communication abilities. The cellular coverage in the San Angelo was extremely low quality and very inconsistent. We had wanted to test whether or not we could send data files between two moving vehicles. Namely, we wanted to test between Scout and Chase, where the strategists were. Otherwise, any forecast changes would have to be done by voice, which, with a long forecast file, could take a substantial period of time. Such testing would have to wait until we got to Arlington.
But we still accomplished quite a bit driving those two weeks. We put the our computer models of solar radiation fluxes through many tests, and they passed with flying colors. With Eric Slimko, head strategist, and Steve Lukachko, our cellular communications expert, we were able to iron out the protocol we would use during the race. And, using the anemometer I had acquired, we tested how the car responded to head, tail, and cross winds. These results confirmed what aerodynamic models of the car had told us. Most importantly, this was the first time that I had worked closely with some of team members, and the harshness of the conditions made for the formation of some solid bonds between team members.
Our time in San Angelo was not without its interesting episodes, either, especially meteorologically speaking. The morning of June 9th brought with it some extremely unusual looking clouds. Cylindrical clouds, with their long axis pointing east-west, moved from north to south over us. These clouds were very obviously rotating about that long axis, giving the impression that they were literally rolling across the sky. The roll clouds were low enough that when underneath them, an observer got the distinct and very bizarre sensation that the cloud was falling on him. Our best guess was that those clouds were orothographic ally induced by the numerous hills and buttes in our area, but this was really only an educated' I have never seen anything like them before or since.
The same could be said for what happened to us that very same evening. Two days earlier, I had Maize & Blue put away in its trailer due to approaching severe thunderstorms. The garage in which the car was normally located did not seem very strong structurally, and even worse, it had windows on the west side of building. As the storms approached from the WSW, I became alarmed about the possibility of hail breaking through the windows and wreaking havoc on the solar array of Maize & Blue. Well, in the situation two days earlier, the storms fizzled out just as they came overhead, giving us a brief drizzle but nothing more. On June 9th, though, stormy weather returned. Joseph and I were in a small shack next to the garage eating dinner, and so neither of us looked outside for about half an hour. What had been a fine afternoon with only some cumulus development became during that half hour something akin to a solar eclipse. It was too late to move Maize & Blue to safety, because the blackness of the sky shortly gave way to a downpour. What followed was the most ferocious thunderstorm I have ever experienced. The wind was gusting to strengths that I can only speculate at, since we had move all of our instruments indoors. The hail accompanying the storm lasted about fifteen minutes. It never got above about goofball-sized, but was of such intensity and duration that at one point, hail covered the ground to a depth of about four inches. An especially large hailstone managed to crack, but thankfully not break, one of the windows in the garage. Other than that, the worst damage was a roof that leaked everywhere, and water and ice everywhere, inside and outside. The tents that we had set up outside, although buried under ice, were miraculously not severely damaged. The only damage to team vehicles was dented bodywork. From the satellite pictures we looked at later that evening, it was obvious that we had been hit by a developing musicale convective complex (MCC), that would go on to wreak havoc on central Texas that night. We were very, very lucky that things did not turn out worse. From that point on, Joseph and I never ate at the same time.
Anyway; those two weeks passed slowly, but they did pass. On the afternoon of the 13th, it was on to Arlington, if only temporarily.
Race Route Testing
We arrived in Arlington that evening. Our plan was to only stay that night, and on the next day, begin a complete simulation of the first three days of the race. That meant that I had to leave for the day one finish line in Ada, Oklahoma, almost as soon as I arrived in Arlington. I got into Ada at about 2:00 am, which led to my decision to not personally do a complete simulation of day one, since that would involve waking up at 3:30 am. So I gave myself an extra couple of hours sleep, and got up at 5:00 am.
Since I was completely independent from the rest of team, I had a fair amount of leeway in how to do my job, so long as my job got done. As a result, the Spy race day schedule was not nearly so structured as the rest of the teams' schedules. I needed to give myself as much time in the morning as necessary so that I could produce a forecast for that day and an update for the rest of the race. The daily forecast had to be in to Eric Slimko by 8:15

am, so that he could run his simulations and decide how fast Maize & Blue should travel that day. Once I had made my input into the morning forecast, I needed to keep an eye out for any changes and provide updates if necessary. I was to continue this until Maize & Blue began to approach the city I was in. I needed to be at the finish line no less than an hour before Maize & Blue got there, so that I could pick a good charging site where the team could set up the array and recharge Maize & Blue's batteries. The Scout vehicle would usually arrive about fifteen minutes before Maize & Blue did to help me with this task. I also was usually responsible for guiding our tractor-trailer into the finish line area, and find a good spot for it to park in.
As was expected, things did not go all that smoothly the first day. But the biggest problem came from an unexpected source. About halfway through the run, Maize & Blue suffered a chassis failure. A member of the space frame literally broke away from the rest of the frame. A major catastrophe was luckily avoided as the broken piece did not puncture any other component of the car. But a broken chassis is still a minor catastrophe, no matter how you look at it. It took some time to get Maize & Blue drivable again, and it did not get into Ada until late in the afternoon.
On the weather end, there were also some technical problems. Attempts at transferring data files between Spy and Scout, and between Scout and Chase, almost always ended in failure. We planned on taking a long, hard look at the problem when we returned to Arlington in a few days, but we began to doubt that the problem the solvable. The weather, on the other hand, did go as planned; a morning with some light fog gave way to sunny skies in the late morning and partly cloudy with cumulus convection in the afternoon. Happy with the forecast, but unhappy with the data transfer problem, I headed to Tulsa, Oklahoma, that evening in preparation of our simulation of race day two.
Day two turned out to be a much calmer day than day one for most of the team. Maize & Blue ran flawlessly, and the run from Ada to Tulsa was uneventful. Except for the meteorologists. We missed this forecast by a significant amount. It was the worst kind of error, as well; we predicted more sunshine than we received. Joseph and I under forecast the amount of cumulus convection we would see in the afternoon near Tulsa. If this had been an actual race day, the run would have been very inefficient, since Maize & Blue would have slowed down late in the day to reduce the amoun1 of power it was using. An efficient run is one in which the solar car goes a constant speed, since energy used is proportional to velocity squared.
I already almost to the day three finish line of Fort Scott, Kansas, when the team called via the cellular phone to say that they had decided not to simulate the third race day. It seems that some of the crew members were; concerned about wear and tear we were putting on Maize & Blue, and wanted the extra day just to do maintenance on our car and equipment. So we spent Wednesday, June 16th, doing just that. That evening, the team headed back to Arlington, where we met Mark, as he had flown in from Detroit. There was an awful lot of doubt running through the entire team at
this stage as to _whether or not we had a car durable enough to last the entire race. .
Last Minute Preparations
Pre-race activities had actually started in Arlington on the 16th. The Arlington Civic Center was opened up as both a work area for the teams to use and a display area for the public to walk through and see the solar cars.
Those cars that had not successfully qualified at the qualifiers were allowed one final try during the week by running laps around Arlington Stadium until the required fifty miles was completed. We, of course, did not have to deal with that. Instead, we just needed to make sure Maize & Blue was in tiptop shape for race scrutineering, which for us was on Friday. We also made a point of checking out our competition. At this point, we were under the assumption that Cal State LA was going to be our toughest competitor.
On Thursday, we gave our communications system a final shakedown. It-revealed what we had feared for some time: data file transmission from cellular phone to cellular was next to impossible. The morning file could of course be given to the strategists in person, but any updates would have to be by voice, a cumbersome but now unavoidable process. I personally began to assemble data to make an initial long-term weather forecast of what race week would be like.
Friday was a very busy day for everyone. Maize & Blue scrutineered late in the morning, and the entire team was on hand for the occasion. The

car made it through without a hitch, bringing the team spirit up in the process. In the afternoon, Joseph, Mark, and myself went to the National Weather Service Forecast Office in Fort Worth to get a chance to look at their maps, some of which we could not receive through WSI. Upon return to Arlington late in the day, I constructed my first - and, as it would turn out, last - long range weather prediction.
F or once, all of the long-range computer models were in agreement. Only one mid-latitude cyclone would be affecting the central United States during the upcoming week. It looked like it would swing through the Upper Mississippi Valley during the fourth and fifth days of the race. Days six and seven of the race, which would be behind the surface cold front, looked like the best days as far as sunshine was concerned. Day three appeared to be a typical convection-in-the-afternoon summer day. Amazingly, my greatest uncertainties were concerning the first two days of the race. A tropical depression had been gathering strength in the western Gulf of Mexico, and was located just ESE of Brownsville. The storm had been drifting slowly northward, but was expected to stall out and remain relatively stationary over the next several days. However, the circulation around the storm, combined with the typical Bermuda High flow, was feeding a great deal of moisture into eastern Texas. The critical question was how far north the clouds would get. Counting on the storm being stationary, I decided that the clouds would be with us for most of day one, but that we would break into sunshine just before we got to Ada. On day two, I figured that we would again start with clouds and fog, but again we would break into sun again as we approached Tulsa. This completed forecast was presented to Eric Slimko, with the promise that I would provide updates and corrections if needed. They never were.
Saturday dawned a bright, sunny day. Overnight, the tropical depression in the Gulf had intensified, as expected, into the first named storm of the year, Tropical Storm Agnes. And the Dallas area slowly became overcast as the day progressed. On the team, it was an anxious day, as everyone hoped we would be ready for the start of the race tomorrow morning. By mid-afternoon, I was on my way to Ada with all of my computer and communications equipment. I wanted to get a good night's sleep before the start of Sunrayce '93. And amazingly, despite my personal nervousness, I did.
Day One
Start: Arlington, TX
Media Stop: Whitesboro, TX
Finish: Ada, OK
My day started promptly at 4:00 am. The slightly later wakeup time prompted by my realization during the race simulation that I did not need as much time as I was giving myself each morning. In Arlington it was a foggy morning with occasional light rain. The morning forecast was identical to the forecast I made two days previous: overcast over most of the route, but Maize and Blue will break into sunshine near Ada. The afternoon charging period looked to be pretty good, as I didn't think the clouds would roll in from the south until after sunset.
Unlike just about every day of practice we had, day one of Sunrayce '93 went like clockwork. Since my long-range forecast called for significant clouds for four of the next five days, Eric Slimko set what he thought would be conservative pace. Conservative it may have been, but by the mid-day media stop, we had moved from our starting position of tenth all the way first place. Cal State LA had all sorts of trouble with their motor, putting them far off the pace. We maintained our lead the rest of day, and Maize & Blue broke into sunshine for the first literally as she entered the Ada city limits. As the first arrival at the finish line, we the first shot at picking a charging site, so we of course got the best one available. A perfect end to a perfect day. The closest teams to us were California Polytechnic Institute at Pomona (Cal Poly-Pomona), and George Washington University (GW). I drove on to Tulsa under bright sunshine that afternoon, scarcely daring to believe that things were going as well as they were.
Day Two
Start: Ada, OK
Media Stop: Shawnee, OK
End: Tulsa, OK

Ada on the morning of day two was much like Arlington on day one, except there was no precipitation. The clouds and fog were back, however. Not to worry, though, as they were expected. Again, we forecast the clouds to stick around through the morning. We thought they would dissipate a little earlier than they had the day before, as we put distance between ourselves and the moisture source, the Gulf of Mexico.
And again the forecast was right on target. I got a little nervous as the low clouds normal for the mornings hung around Tulsa a little longer than I thought they would, but they had broken up by 1 :00 p.m. Maize & Blue's lead was never threatened, and she got into Tulsa by mid-afternoon. We had a few more cumulus clouds during the charging period than we had anticipated, however. The difference was not very large, but it was there, and Mark, Joseph, and I knew that we could do a better job than that. Otherwise, day two was just as smooth as day one.
Day Three
Start: Tulsa, OK
Media Stop: Miami, OK
Finish: Fort Scott, KS
We were all surprised by the fog that formed in Tulsa overnight. Although it burned off very quickly, it still impacted the morning charging period in an adverse way. The morning forecast called for an otherwise mostly sunny morning, with some limited cumulus development in the afternoon.
This day- saw the first cracks in the Michigan armor. Although the car ran perfectly yet again, the rest of the caravan did not. Our Chase van, trying to keep up with Maize & Blue, ran a traffic light that was going from yellow to red in Joplin, MO. It was a very close call, but the official decided to hit us with a twenty minute penalty. Due to the penalty, we fell to third place behind GW and Cal Poly-Pomona. Our confidence was still high, however, as we thought we had the key to catching at least one of those two the next day. The forecast I had made two days before the race still looked good over the next two days, and it was becoming increasingly obvious that those two days would decide the race.
Day Four
Start: Fort Scott, KS
Media Stop: Kansas City, MO
Finish: Cameron, MO
F or the first time so far in the race, the starting line had a clear morning. With a somewhat clearer picture of what the frontal system was doing off to our west, I decided to improve my pre-race forecast for the afternoon somewhat. I called for a mostly sunny morning, with clouds gradually increasing towards a mostly cloudy evening.
This was by far my worst day of the entire race. The clouds did increase, but there was nothing gradual about them, as they rolled in very quickly early in the afternoon. A line of clouds ahead of the frontal system developed rapidly in the morning, and scattered showers and thunderstorms were in the Cameron area by the time Maize & Blue arrived. Joseph and Mark did an excellent job in the Scout vehicle this day, helping the solar car to dodge as much of the cloudy conditions as possible. Nonetheless, there was not nearly as much sun to be found as I had anticipated. If I had stuck with my pre-race forecast, I would have done a lot better.
The team had one strategy on day four. We knew that Cal Poly Pomona had decided to use our strategy as theirs~ they maintained the same speed at all times that we did. They were apparently listening in to the radio conversations between the Chase van and Maize & Blue. So, after studying and modeling Cal Poly-Pomona's car extensively overnight, we chose a pace that we knew we could keep, but that they probably could not. In this respect, my lousy forecast that day did not have as much impact as it could have, since on that day the speed we chose to travel was to some extent independent of the weather forecast. The strategy worked like a charm, as Pomona followed us for most of the day, and then abruptly fell off our pace as they simply ran out of battery charge. We had also gained a lot of time by our skillful navigation of the lunch-hour traffic in Kansas City.
Perhaps we were a bit too enthusiastic, then, as the day neared its end. Within site of the finish line in Cameron, we overtook GW. Unfortunately, while passing, the Chase van clipped in no-passing zone. We were given our second penalty in two days, this one ten minutes, due to that infraction. This one was especially tough to swallow. Before the day started, Pomona had an eighteen minute lead over us. We had virtually wiped that lead out by our strategy of running them into the ground, only to lose much of what we had gained by way of a stupid penalty. An extremely demoralized team was evident that afternoon; some thought that we had shot our last bullet that day. But they were wrong, as we still had some ammunition left.
Day Five
Start: Cameron, MO
Media Stop: Lineville, IA
Finish: Des Moines, IA
Day five began in a torrential downpour. Most teams did not even bother bringing their vehicles out to attempt to get a morning charge, as it was simply too dark to be worth it. We had been prepared for this weather for a full week, however, so it did not bother us. My forecast called for the rain to taper off through the clouds, but for the clouds to stick around until sometime between 5:00 and 6:00 p.m.
If day four was my worst day, then day five was my best. This was the day I had been telling and warning the team about for a week. We had played it conservatively all week long; now we planned on putting all of our stored energy to use. It seemed that most of the teams thought that the sun would come out early that afternoon. Indeed, upon talking to them afterwards, I found out that that was exactly the case. It was a fatal mistake. Most teams blew by Maize & Blue at the start that morning as we began the day at a twenty mile-per-hour pace. We passed every one of those teams on the side of the road later in the day, waiting for sunshine that never came until far too late. By mid-afternoon, we were the only team still out running, as every other team had trailered their vehicles to Des Moines, accepting the penalty involved in exchange for being able to charge their array at the finish instead of on the side of the road somewhere. And when the sun did finally come out at 5:45 p.m., it was loudly cheered.
Our team was doing more than its share of cheering. Maize & Blue's speed was down to fifteen MPH, and she having trouble getting up hills late in the day. The late afternoon sun was just enough to get the solar car across the finish line at 6: 15 p.m., just fifteen minutes before racing officially ended for the day. I had the honor that evening of accepting the team's pins from race director Richard King for finishing in second place that day. (Iowa State University had completely changed their battery pack to allow themselves to cross the line first in front of the home crowd. They incurred a ten hour penalty by doing so.)
What had been a ten minute deficit had been turned into a whopping two hour lead. The entire team was ecstatic. We knew that the race was all but over.
Day Six
Start: Des Moines, IA
Media Stop: Iowa Falls, IA
Finish: Albert Lea, MN
From here on in, the forecasting was easy. Day six dawned bright and sunny, and the sun wasn't going anywhere the next three days, as subsidence was hampering convective development. With a two hour lead, we could afford to play conservatively the rest of way . We kept an easy forty MPH pace on both day six and seven. I correctly forecast a slight
increase in puffy cumulus during the early afternoon hours, but to be honest, that was fairly immaterial at that point. We lost about fifteen minutes off of our lead that day, dropping the lead to 'only' one hour and forty-five minutes We could not wait for the next, and final, day.
Day Seven
Start: Albert Lea, MN
Media Stop: none
Finish: Minneapolis Zoo
Apple Valley, MN

A carbon copy of day six: bright sunshine, a few puffy cumulus, and an elated Michigan team. We covered the ninety or so miles that day in about two and a half hours. Again, we had plenty of solar energy to spare, but we could afford to be conservative. We had another fifteen minutes chipped off our lead, but we could not have cared less. At approximately 11 :30 am, we crossed the finish line to become back-to-back National Champions.
What I Could Have Done Better
There are two areas that I could have improved on. The first and most obvious was my day four forecast. I should have learned my lesson from the first two days, when persistent clouds tempted me, but did not convince me, to change my forecast. Instead, I changed slightly the forecast I had made before the race started the morning of day four, without really having a logical reason for doing, other than that a forecast made almost a week ago can not possibly still be accurate. And yet it was. Lesson number one: Be stubborn with your forecasts. Do not change them without a really good reason.
The second area involved my afternoon activities. For the first four days, I mostly stayed hunched over my computer, bringing in data through WSI. On day five, due to our late finish, I had more time on my hands than any of the previous days. I decided during the afternoon to take a drive out to the WNW to find the end of the clouds out to our west. I did find it, after an hour or so of driving that took me the better part of the way to Omaha. I was able to report back to the caravan that I had seen the clearing with my own eyes, and yes, it was on its way. Seeing the benefits of this "Spy"ing on the weather, so to speak, I did the same thing on day six, with less dramatic results, but it still allowed me to see increased cumulus clouds to our west, and moving in our direction. I now wish I had done some westward scouting every day of the race. Lesson number two: No piece of technology is more valuable to weather forecasting than your own eyes and
your own mind.

What I Could Not Have Done Better
I quite honestly stunned myself with the accuracy of the long-range forecast I made two days before the race started. That forecast, covering a period from two to eight days into the future, was absolutely, one hundred percent perfect. I have never made a better forecast, and I quite possibly never will. And it helped us win.
What I Got Out of It
The opportunity to use my knowledge in a real world, non-classroom situation was something to be grateful for. The weather forecasting contests within AOSS offer no comparison to the pressure and exhilaration of contributing to a team effort to achieve a common goal. The very opportunity to be on the team and to race with them is something I will always treasure.
I also learned a great deal that I can apply to my forecasting elsewhere. Besides the two lessons I listed before, I can give another piece of advice: Do not trust processed or enhanced data blindly. The satellite pictures we received from WSI were so processed to be user-friendly and easily understandable that they lost much of their value to us. Indeed, there was even a negative impact, as low clouds were often filtered out of the images by WSI processing.
I gained my first practical experience with tropical cyclones through this race. I gained my first experience with a desert climate through this race. I first applied FORTRAN to a real world problem through this race. And most importantly, I think: I first worked extensively on my weather forecasting with students outside of the AOSS department through this race. The AOSS curriculum is so isolated from the rest of Engineering that the chance to work with non-meteorologists in not often found. Perhaps this isolation is unavoidable, but if the curriculum of AOSS is so different than that of the rest of the College of Engineering, then maybe AOSS should not be in the College of Engineering.
I also greatly improved my leadership skills through this race. These new skills - the ability to see a problem and to fix it yourself without someone telling you to do so, or the ability to speak up when a problem is bothering you - would be greatly tested as the World Solar Challenge
Approached in the fall. But for now, I sit back and enjoy a National Championship that I helped attain for the University of Michigan, for the Solar Car Team, and for myself.

7509 words | fnazeeri

How informative is this Article?



	Related Content (2)

	Maize & Blue, Weather Forecasting and Australia (by Jeff Zoltowski) Article by Furqan Nazeeri created 11/18/1993 submitted 1/18/2004 07:26:47 AM This 20-page paper is a final report prepared by Jeff Zoltowski and relates the 1993 Maize & Blue team's experience racing in Australia with a focus on weather forecasting and it's affect on the team's overall performance. A very interest report. I encourage you to also read the Jeff's paper on the team's experience in Sunrayce 93 (also told from a weather-forecasting perspective)

	Weather Forecasting for Maize & Blue (by Joe Bartlo) Article by Furqan Nazeeri created 6/26/1993 submitted 1/18/2004 08:59:15 AM This is a paper written by Joseph Bartlo about Maize & Blue's use of weather in its race strategy in Sunrayce 93 and the World Solar Challenge '93.