Detecting the Wind's Direction Is the Wind Predictable? by Nolan Reis Mentor: Rob Reis The intent of this project is to find out if the direction of the wind is predictable and if so, with that information, can a sailor more optimize his route around a racecourse. The first step before forecasting the wind is collecting and analyzing data for trends. Research shows that the wind has some predictability and if, for example, the wind shifts twenty degrees every ten minutes, the construction of a real-time wind predictor is useful. The technique that will be used to calculate the direction of the wind is to construct an apparatus containing cross-axis of ultrasonic transmitters and receivers. Sonic pulses will be sent along each axis. The magnitude of the wind ‘s effect on each pulse’s transit time will be logged and used to find the magnitude and direction of the wind, through the use of trigonometry. Tests will be done to verify that the wind has a linear relationship to the sonic pulse transit time. Also, through these tests the sensitivity of the temperature, a factor on the speed of sound, on the transit time will be determined. The final construction is the critical path at hand, and data collection should follow in the next couple of weeks. After, the data will be analyzed for trends and possible predictability. If there is a reasonable predictability of the wind, then real-time data will be provided to see if it optimizes a sailor’s path around a racecourse. The BASIC Stamp® microcontroller is connected to eight devices; the StampMem, the math co-processor, the LED AppMod, the Devantech compass and four modified SRF04 Ultrasonic Rangefinders. To start the program the user pushes the green button on the LED AppMod board. The first step is for the ultrasonic range finder to send out a pulse across a nine inch path to a ultrasonic receiver and also to find the average transit time. It repeats this for the other three range finders. Then the microprocessor takes the numbers from path 1 and path 3 (across from each other) to find the Vx component of the wind. Path 2 and 4 calculate Vy. With Vx and Vy, it uses the Pythagorean Theorem to find the magnitude of the wind and arctan to find the wind direction. This process is repeated every couple of seconds or (in my new version) 100 times a second and then stored, as well as the time and compass heading. Lastly, it uses the math coprocessor to display the computed wind magnitude, wind direction, and the actual compass heading. I use the trend line of the historic wind direction measurements to determine future wind trends. The award I won was the Grand Prize at the Silicon Valley Science Fair. I received a plaque, $250 cash, $1000 dollars for my school, and several specialty awards.  Nolan recently finished 3rd place in the Physics competition at the 2003 International Science Fair in Cleveland, Ohio.