Introduction

An examination of some local earthquakes and large teleseisms indicated  a problem with the orientation of the borehole accelerometer at USIN. The repaired accelerometer was installed May 25, 2017 and metadata was updated.  However with the Mexican Mw=8.1 earthquake of 2017/09/08, the issue because apparent.

In order to compare the output of a broadband seismometer to an accelerometer there must be significant signal-to-noise in a common passband. For very large earthquakes, this means that comparisons can be made at periods as long as 20 seconds. For small nearby earthquakes, the comparison must be made at higher frequencies.

On 2017/09/19 11:47:28 there was an Mw=3.8 earthquake 55.25 km from this station. The back azimuth was 337.3 degrees.
Fortunately we can directly compare the ZNE channels without having to rotate the HH broadband and HN accelerometer channels because of the installation problem.

Fig. 1. Instrument corrected traces to ground velocity in m/s

Figure 1 compares the filtered ground velocities for the Illinois earthquake. If it obvious that the vertical agree, that the HHE and HNN look the same and that the HHN and HNE are reversed. Note we want HHE and HNE to be the same and HHN and HNN to be the same.

Channel Comparison

Fig. 2. Comparison of filtered ground motions for the vertical component. This is excellent.

Fig. 3. Comparison of HHE to HNN.

Fig. 4. Comparison of HHN to HNE. Notice the reversal

Fig. 5. Comparison of HHN to HNE

Resolution

Fortunately the accelerometer was installed  exactly 90 degrees off of alignment. This is what made it possible to have the successful overlays.  If the angle had not been 90 degrees, then it would have been necessary to go through a sequence of rotations to determine the true orientation with respect to the broadband accelerometer.

Action

Change the metadata for the USIN acclerometer for the epoch starting May 25, 2017 to present for all recorded horizontal channels, e.g., BNZ HNZ etc, so that we have the correct orientation.
Note we cannot change the channel names. Perhaps in the future we could call the accelerometer channels HN1 and HN2

Channel HNZ  ( and all NZ's)  - No change
Channel HNN  ( and all NN's)  - currently azimuth = 0
                                                     CHANGE TO AZIMUTH 90
Channel HNE  (and all NE's )   - currently azimuth = 90
                                                     CHANGE to AZIMUTH 180