I was having problems with overseas shipping for some wind sensors so I tried used a cheap microphone as an airflow sensor instead. It sort of works and I am quite proud of it. At first I hooked up the microphone to an Arduino but it was not able to tell the difference between sound and wind and was too easy to trigger. So I took away the Arduino and fed the mic input directly into Processing and did a real time Fast Fourier Transform analysis to help differentiate a wind versus sound input. A sound input would usually cause a range of frequencies to be picked up whereas an airflow would cause a distinctive shape with the lowest frequency being the loudest, and tapering away with higher frequencies. I also used a smoothing filter by averaging the readings over a certain amount of time. The result can be seen in the video.
lohjianhui 