Evaluation of Predictive Modeling Based on N-VS Modeling of Geological Features
摘要
This study presents the Standard Penetration Test (SPT) N and cross-wave velocity (VS) models related to the modeling of geological properties in central Java, Republic of Indonesia. Multiple empirical N-VS models developed using datasets collected from different areas of central Java, Republic of Indonesia, produced different N-VS cross-modeling trends. In order to validate the predictive ability of each model, sounding records containing SPT N and VS measurements were collected in the Java region and then fully analyzed in this study. Java is located at the demise boundary between the Asian-European and Indian Ocean plates and is geologically active with numerous earthquakes and volcanoes. Despite the relatively limited geographic extent, we found that the model developed using data from measures such as reclamation, foundation treatment, and piling performed well, while the model constructed using datasets within the hilly, plain terrain performed better in the area of karstic hilly terrain (Semarang City, Central Java Province, Java Island). This comparative study confirms the dependence of N-VS on VS in modeling geologic attributes and provides N-VS that can be used to estimate VS with similar geologic attributes to model VS based on SPT N values.