Hi Mark,

Thanks for the analyses and the rapid turnaround.  It's great working with 
you.

Attached is your spreasheet with a couple of summary plots that you might 
find interesting.  I have included the previous results you generated for the 
case where no filtering on wind direction was done.

The results indicate to me that the additional filtering of the data is 
overall improving the accuracy of the free-stream wind speed estimation 
(based on the mean nacelle wind speed and mean wind speed ratio), but at the 
cost of larger standard error because of the reduction in number of samples.  
Therefore, it seems that additional filtering might not be so valuable 
afterall.

Any comments?

Philippe