RESEARCH COMMUNICATIONS CURRENT SCIENCE, VOL. 82, NO. 11, 10 JUNE 2002 1366 Strong ground motion at Bhuj city during the Kutch earthquake R. N. Iyengar* and S. T. G. Raghu Kanth Department of Civil Engineering, Indian Institute of Science, Bangalore 560 012, India In the absence of near field strong motion records, the level of ground motion during the devastating 26 January 2001 earthquake has to be found by indirect means. For the city of Bhuj, three broad band velocity time histories have been recorded by India Meteoro- logical Department. In this paper these data are processed to obtain an estimate of strong ground motion at Bhuj. It is estimated that the peak ground acceleration at Bhuj was of the order of 0.38 g. Ground motion in the surrounding region is indirectly found using available spectral response recorder (SRR) data. These instrument-based results are compared with analytical results obtained from a half-space regional model. BHUJ city (23.25°N, 69.65°E) in India was severely damaged during the Kutch, Gujarat earthquake of 26 January 2001; the event itself is widely referred to as Bhuj earthquake. According to the India Meteorological Department (IMD) 1 , which officially maintains a series of observatories all over India, the epicentre of this quake was at 23.40°N, 70.28°E. Bhuj city is close to the epi- centre, located at a distance of about 65 km to its south- west. An important feature of this earthquake was the widespread liquefaction and ground deformation 2 in the epicentral tracts. In nearby places such as Bachau (23.3°N, 70.30°E) and Manfera (23.46°N, 70.38°E) and even in some far-away locations like Kandla (23°N, 70.1°E), structural damage could be attributed as much to ground failure as to vibratory ground motion. From the engineering point of view, the most sought-after data are the accelerograms recorded digitally on a strong motion accelerograph (SMA). Since the time and place of occur- rence of a strong earthquake are not known before the event, the present practice is to place several self- triggering SMAs with in-built clocks in a seismically- active region to acquire data as and when an earthquake occurs. Unfortunately, for the Kutch earthquake no such near source strong motion records are available. There was an SMA at the IMD observatory at Bhuj, but this did not produce a record due to cable failure. A three- component record on a strong motion instrument has been obtained by the University of Roorkee 3 at the basement of a ten-storey building in Ahmedabad. The record appears to be affected by the vibrations of the building in which the SMA was located, and moreover the station was some 250 km away from the epicentre. Thus, apart from making educated guesses based on the damage pattern, very little has been understood about the strong ground motion during the Kutch earthquake. The only near source data available are the broad band velocity records from the Bhuj observatory, operated by IMD 1 . The velocity data are processed in this paper, to estimate the ground motion time history at Bhuj. Even though no SMAs were operating in the region, an array of spectral response recorders (SRR) was maintained by the Department of Earthquake Engineering, University of Roorkee 3 . These recorders directly give the peak res- ponse of damped simple harmonic oscillators, tuned to specific frequencies, under the earthquake. Here, we use these data to carry out an indirect estimation of peak ground motion at the respective stations. Variation of peak ground acceleration (PGA) with respect to distance gives first-hand information on regional characteristics and also possible spread of damage during future earth- quakes. Accordingly, the attenuation of PGA with respect to distance is presented. The ground motion results obtained here are not directly recorded, but are derived indirectly from other instrumental data. Thus, the ques- tion arises as to the possibility of validating these results by other means. With this in view, the estimated dis- placement time history at Bhuj is compared with analytical results based on the source mechanism proposed by Yagi and Kikuchi 4 , coupled with a half-space regional model. The only near source instrumental data available for the main shock are a set of three velocity records from the Bhuj observatory of IMD recorded on a force balance seismometer (STS 20-Q680). As described by Bhatta- charya and Dattatrayam 5 , this instrument can be used for measuring ground velocity in the range of 0.08–50 Hz, in which interval the instrument has a strictly flat frequency response. The basic three-component velocity signal recorded at the rate of 20 samples per second at Bhuj, is shown in Figure 1. The original data in counts have been *For correspondence. (e-mail: rni@civil.iisc.ernet.in) Figure 1. Velocity record (unfiltered) for the main shock of 26 January 2001 recorded at the IMD broadband station in Bhuj: a, EW; b, NS; c, Vertical. Time (s) a b c