Pemodelan Estimasi Kecepatan Rambat Gelombang Geser Tanah (VS30) Berbasis Topografi, Geomorfologi dan Geologi

Abstract

The 30 m top layer of soil as the medium of propagation of the earthquake wave is the closest to the structure of the building, and could have different effects depending on the type of soil and topography. The Indonesian earthquake code for building and non building structures known as SNI 1726-2012 using the directly measured VS30 as the primary parameter to identify the stiffness effect of sediment. The VS30 can be measured using non invasive methods, such as multi-channel analysis of surface waves (MASW). Direct, invasive measurements of VS30 around Indonesia would be difficult to implement due to the vastness of the country and the high cost nature of the testing. To provide an alternative to the direct measurement, VS30 estimation models have been developed. VS30 estimates using topography in the form of slopes are commonly used in North America, while geomorphological units are used in Japan. This research was carried out by correlating VS30 direct measurements with topographical, geomorphological and geological attributes. The VS30 obtained from series of MASW tests, the topographic slope and elevation from the Shuttle Radar Topography Mission (SRTM) 30 arcsec data, the geomorphology units data which are Structural, Karst, Vulkanik, Fluvial and Marine from landsystem map and the geologic age data from Geological Survey Centre (PSG). Data were analyzed by linear regression and spatial analysis. VS30 estimation modeling produces models with four variables, namely elevation, slope, geomorphological unit and geological age in the regions of Palu. Four proxy based estimates provide values that are slightly higher or lower but in a range not so far from direct measurements. The results of data processing analysis shows that local conditions greatly affect VS30 estimates in Palu. The VS30 estimation model in the form of a logarithmic equation is as follows, Log VS30 = -3.925 + 0.062 log(s) + 0.069 log(Ev) + 0.665 log(G) + 1.824 log(A) gives a slightly higher or lower value but in a range not far from direct measurement. This VS30 estimation model is suitable for soil classification on a regional scale and can be adopted for microzonation maps or real-time shake map.