Minor cool stuff to show off.

For the Wind Turbine project, we need to design it such that it will actually operate, right? It needs to be designed around a windspeed which is reasonable for the area it's going to be operated. If we pick a design speed which is too high, then the ambient wind won't be enough to actually turn the generator over - and the whole thing will be useless. On the other hand, if we design it for a wind speed which is too low, then a huge amount of available energy will go untapped and we'll probably lose the competition. Long story short - we need to know what kind of wind we can expect in the WL area.

So I contacted the National Climatic Data Center (NCDC), a subdivision of the NOAA, for some data. Apparently, at every operating airport in the United States, they take hourly weather data measurements which are stored. This is AWESOME because the intramural playing fields, which is where the turbine competition will be held, is less than 1/10th of a mile from the Purdue Airport. So, I got hourly measurements on windspeed and direction at the site for the past 30 years. Pretty cool, huh? Well, it turns out that's way too large a data file for EXCEL or MATLAB to handle all at once, so I chopped it down and only looked at the past 6 years (January 2000 to December 2005 still represents about 65000 data points). Naturally, I wrote a code that does some statistics and makes some sweet graphs which reveal the results:

1) Polar Plot. This shows data for one year, windspeed and direction. It's a little difficult to read because there are many data points on top of each other so you don't really know the frequency that one windspeed may occur in one direction.


2) Statistics Plot. This shows the average (mean) windspeed and standard deviation of the windspeed value as a function of direction. Now you can start to see that there are definitely some directions where wind is more likely to come from.

3) Mass Data Summary. This one is really cool. This surface plot shows the information that was used to calculate the numbers in the statistics. This is raw data and it's awesome:


Keep in mind this is especially cool because of how accurate this information is for our group. These graphs are based on public information and basic statistic analysis, but the code to make this stuff took a while to get right. Basically, I'm guessing the other team doesn't have this. So, if you know anybody in the other group, don't tell them about this, k? I'm just really pleased with how clear the trends came out and wanted to show off......