[1] We report the first consistent observations of rotational motions around a vertical axis induced by distant large earthquakes. It is standard in seismology to observe three components (up-down, N-S, E-W) of earthquake-induced translational ground motions using inertial seismometers. However, only recently ring laser technology has provided the required sensitivity for observations of the theoretically predicted rotational part of ground motion generated by seismic waves in a wide distance range and frequency band. Here we show that the rotations observed are consistent in waveform and amplitude with collocated recordings of transverse accelerations recorded by a standard seismometer. This suggests that rotations may become a new observable for seismology and related fields with the potential of providing complementary information on earthquake source processes, structural properties, and ground shaking.

S U M M A R YIt has been noted by theoretical seismologists for decades that-in addition to translations and strains-the rotational part of ground motions should also be recorded. It is expected that collocated measurements of translations and rotations may (1) allow transformation of translational seismograms to the complete ground motion of an observation point; (2) help to further constrain rupture processes and (3) provide additional hazard-relevant information to earthquake engineers. The lack of instrumental sensitivity used to be the main obstacle to observing rotational motions. Recently, ring laser technology has provided the means to develop instruments that allow in principle the observation of rotational motions in a wide frequency band and epicentral distance range. Here we investigate whether this technologyoriginally designed for geodesy-is capable of providing accurate and useful observations for seismology. We report observations of rotations around a vertical axis of several earthquakes obtained by a 4 × 4 m ring laser installed in SE-Germany and compare them to collocated broad-band translations. Assuming plane transverse wave propagation, acceleration and rotation rate should be in phase and their amplitude ratio proportional to horizontal phase velocity. Here we show that most of the observations can be explained under these assumptions and that the collocated observations allow the estimation of wavefield properties (e.g. phase velocities, propagation directions), otherwise only accessible through seismic array measurements, polarization analysis, or additional strain measurements.

Recently, ring laser technology has provided the first consistent observations of rotational ground motions around a vertical axis induced by earthquakes. "Consistent," in this context, implies that the observed waveforms and amplitudes are compatible with collocated recordings of translational ground motions. In particular, transverse accelerations should be in phase with rotation rate and their ratio proportional to local horizontal phase velocity assuming plane-wave propagation. The ring laser installed at the Fundamentalstation Wettzell in the Bavarian Forest, Southeast Germany, is recording the rotation rate around a vertical axis, theoretically a linear combination of the space derivatives of the horizontal components of motion. This suggests that, in principle, rotation can be derived from seismic-array experiments by "finite differencing." This has been attempted previously in several studies; however, the accuracy of these observations could never be tested in the absence of direct measurements. We installed a double cross-shaped array of nine stations from December 2003 to March 2004 around the ring laser instrument and observed several large earthquakes on both the ring laser and the seismic array. Here we present for the first time a comparison of array-derived rotations with direct measurements of rotations for ground motions induced by the M 6.3 Al Hoceima, Morocco, earthquake of 24 February 2004. With complete 3D synthetic seismograms calculated for this event we show that even low levels of noise may considerably influence the accuracy of the array-derived rotations when the minimum number of required stations (three) is used. Nevertheless, when using all nine stations, the overall fit between direct and array-derived measurements is surprisingly good (maximum correlation coefficient of 0.94).

[1] We report the first direct measurements of the very small effect of forced diurnal polar motion, successfully observed on three of our large ring lasers, which now measure the instantaneous direction of Earth's rotation axis to a precision of 1 part in 10 8 when averaged over a time interval of several hours. Ring laser gyroscopes provide a new viable technique for directly and continuously measuring the position of the instantaneous rotation axis of the Earth and the amplitudes of the Oppolzer modes. In contrast, the space geodetic techniques (very long baseline interferometry, side looking radar, GPS, etc.) contain no information about the position of the instantaneous axis of rotation of the Earth but are sensitive to the complete transformation matrix between the Earth-fixed and inertial reference frame. Further improvements of gyroscopes will provide a powerful new tool for studying the Earth's interior.

No abstract