K. Venkateswara
As described by Sheila in 14416, high wind speeds were recorded at LHO on Saturday. BRS shows corresponding high ground tilts during this period.
The first plot shows the ASD of the GND STS, BRS RY and Stage 1 T240 X. There is excellent coherence between the GND seismometer and BRS RY below 60 mHz and the tilt-subtraction looks quite good - you can clearly make out the primary microseismic peak at 55-60 mHz. Also shown is the BRS reference mirror (~sensor noise). After Fridays calibration corrections, the fudge factor I had to use was smaller (0.79) to get the subtraction right.
Stage 1 shows slighlty better coherence with GND_STS and BRS, but it is still limited by some other source of noise below 0.1 Hz.
The second file shows a band-limited BRS RY (5-500 mHz) and wind speeds as measured by PEM anemometers and converted to X and Y directions using the 'direction' data. I might have the axes converted incorrectly, but visually it shows good correlation with the BRS data.
In looking for source of low-frequency noise in Stage 1, I found something interesting: The first file shows the X,Y,Z and RY motion measured by the T240s on Stage 1. The next file shows the coherence between X, Y and Z sensors and RY.
There is large coherence between Y and RY below ~50 mHz which wasn't too surprising to me. Small misalignments of sensor axes can produce that. However, there is large coherence between Z and RY at the microseism which is odd. To get coherence with Z, one needs sensor calibration mismatch or actuator alignment/strength mismatch. It is also surpsisingly large.
Jim said he'd help me look into this further. But it does explain one thing Jeff had noted in 14047 about why RY looks very similar to X but is not coherent at the microseism. Because RY is measuring Z at the microseism!