Ⓔ Suitability of Short-Period Sensors for Retrieving Reliable H/V Peaks for Frequencies Less Than 1 Hz by A. Strollo, S. Parolai, K.-H. Jäckel, S. Marzorati, and D. Bindi Abstract Using three different short-period electromagnetic sensors with reso- nance frequencies of 1 Hz (Mark L4C-3D), 2 Hz (Mark L-22D), and 4.5 Hz (I/O SM-6), coupled with three digital acquisition system, the portable data acquisition system (PDAS) Teledyne Geotech, the refraction technology (REFTEK) 72A, and the Earth Data Logger PR6-24 (EDL), the effect of the seismic instruments on the horizontal-to-vertical spectral ratio (H/V) using seismic noise for frequencies less than 1 Hz has been evaluated. For all possible sensors–acquisition system pairs, the back- ground seismic signal and instrumental self-noise power spectral densities have been calculated and compared. The results obtained when coupling the short-period sensors with different acquisition systems show that the performance of the considered instru- ments at frequencies < 1 Hz strongly depends upon the sensor–acquisition system combination and the gain used, with the best performance obtained for sensors with the lowest resonance frequency. For all acquisition systems, it was possible to retrieve correctly the H/V peak down to 0.1–0.2 Hz by using a high gain and a 1-Hz sensor. In contrast, biased H/V spectral ratios were retrieved when low-gain values were con- sidered. Particular care is required when using 4.5-Hz sensors, because they may not even allow the fundamental resonance frequency peak to be reproduced. Online Material: Results for REFTEK 72A and PDAS digital acquisition systems coupled with 1-, 2-, and 4.5-Hz sensors. Introduction It is well known that subsurface geology can strongly af- fect both the amplitude and lengthening of the earthquake- induced ground shaking recorded at the Earth’ s surface. The assessment and consideration of local site amplification effects is mandatory for studies aiming at seismic hazard assessment, such as the calibration of ground-motion pre- diction equations or the determination of ground-shaking scenarios computed for different levels of source and propa- gation complexity. Because local geology can vary signifi- cantly over distances of only a few hundreds meters, the assessment of site response within large urban areas may re- quire measurements at a large number of sites. Moreover, the chance of recording seismic events in earthquake-prone urban areas can be small, either due to a low signal-to-noise ratio or to a low rate of seismicity. There- fore, techniques that allow information about site response to be obtained without using earthquake recordings and at a low cost have recently gained popularity. Among these, the Nakamura technique (Nakamura, 1989) allows the esti- mation of the fundamental resonance frequency of a site from the peak in the horizontal-to-vertical (H/V) spectral ratio of microtremor measurements. The requirement of performing low-cost and rapid measurements therefore affects the choice of the equipment employed. For example, short-period sen- sors are often preferred to broadband ones because of their ease of installation, their robust nature, and their relatively low cost. Because of budget limitations, microzonation studies based on seismic noise measurements are also per- formed by using low-cost, but lower dynamic range, acquisi- tion systems. It follows that when microtremor measurements are used to infer information about the site response, it is neces- sary to evaluate the contribution to the recorded seismic sig- nal of both the ambient seismic vibrations, that is the signal that carries information about the source generating it and the medium through which it propagates, and of the self-noise of the instruments. In fact, high instrumental noise can limit the exploitability of the seismic noise at frequencies lower than the sensor corner frequency, affecting the estimate of the fundamental resonance frequency of a site from the peak in the H/V . Nowadays, the Nakamura technique is widely applied for estimating local site effects, especially in engineering seismology. While this technique is quite simple to under- 671 Bulletin of the Seismological Society of America, Vol. 98, No. 2, pp. 671–681, April 2008, doi: 10.1785/0120070055