Mc Tavish and Burnside Wind Speed and Direction (Oct 21-22)

I looked at wind speeds and directions from thursday 21st to friday 22nd. The attached graphs show you what I have found out. First of all, I realised that the Burnside data (for whatever reason) is shifted by 4 hours. So the wind data are recorded with a time tb = tt + 4. To correct for this difference I simply subtracted 4 hours in the time columns for burnside. That is why the burnside data actually starts at t = -4. Note that the datasets end at 8pm local time on the 22nd October (that means at t = 44h, 4 hours before midnight).
If we correct for this shift the two data records correlate quite well. However, this is only based on qualitative estimation. As you can see from the Windrose plots, during relatively large time intervals Burnside and McTavish have more or less the same wind direction.
However, I tried some regression analysis and the correlation coefficients for the wind direction is always very low. That mean we don't have a quantitative proof yet to confirm our hypothesis. Maybe we can ask one of the profs to find out how we can use statistics to proof our finding quantitatively.

- Jan-erik