Long Term Behavior of an Accelerometer Station

In 2006 Cologne University installed a strong-motion network (SeFoNiB, Figure 1) with 19 stations in the Lower Rhine Embayment (Hinzen and Fleischer 2007). All stations are equipped with EpiSensors model FBA ES-T from Kinemetrics Inc. (2005). The three components of acceleration are recorded at a sampling rate of 250 samples/s at 2-g full-scale setting with 24Bit analog-digital converters. In addition to the acceleration waveform data, the rms amplitudes of each channel have been recorded in two-minute time windows. Recording these data with only one value per two minutes (2,160 values/day) provided a simple quality control of the data and station status. Some of the stations are only accessible by slow digital or even analog phone lines; therefore the small files with the noise-level values covering one day in 58 Kbytes allow a quick overview of station performance. Besides the typical fluctuations in daily noise levels in the densely populated area, the data exhibited quasi-periodic behavior. An initial hypothesis was that tilting of the instrument might cause the patterns in the signals, but this proved false. One of the stations, BA01, is located on the compound of a meteorological station of the Institute of Geography of the Technical University Aachen. Comparison to the climate data from this station revealed that we were observing temperature effects. In this contribution, we present a detailed description of the temperature dependency of the offset and long-term instrument drift of strong-motion stations over a period of three and a half years. While the temperature sensitivity of different kinds of accelerometers for technical applications has been tested in the laboratory ( i.e. , Chau et al. 1996; Allen et al. 1998; Kruglick et al. 1998; Acar and Shkel 2003), we are not aware of a long-term test of buried seismic sensors. Figure 1C shows the …