12 hours of minute trends where the winds where over 5 meters/second(mps), [~11mph] are attached. First is the means of the three sensors overlayed. Looks okay with some rough times.
Second plot has the three sensors plotted separately and the differences between the sensors; ideally the difference is zero. Some areas of larger (~1mps) diffs and other times where they are very small. The bimodal trend may be from shifty directions, gustiness, or maybe the trends themselves. It's a long stretch of time but I'll zoom in and check second trends and full data where able.
Second trends show the same thing. I suspect the direction and speed of the wind combined to either cause physical interference between the anemometers or maybe the wind itself was altered.... Given the sharp On/Off nature of the differences, I suspect the former. I'll try to get out to examine the anemometers for impact evidence.
I also noticed the discrepancy between the wind sensors for higher wind speeds. However, I think that this is because of the differences in wind direction during this time period/the placement of the wind sensors. Within this windy period, I focused on 3:00-9:00 PM on August 11. I attached plots of the wind speed for each sensors, the wind direction, and the differences between the sensors. From 3:00-5:30 PM, the average wind direction is about 100 degrees. This corresponds to a much larger difference between wind sensor 2 and the other sensors. After 5:30 PM, the wind changes direction. This corresponds to a relatively uniform difference between the three sensors. Based on this, I don't think we need to be concerned about sensor 2 as this discrepancy seems to be because of changes of wind direction. I think that we are ready to move the sensors to the fence.
