Data discovery on liquefaction-induced lateral ground deformations.

摘要

Liquefaction-induced deformations during earthquake pose considerable risks to lifeline systems in urban areas. Empirical models, benefiting from case histories, are preferable for regionally assessing those deformations. Case histories of liquefaction grow rapidly and require new information technologies in data management. Based on pure regression analysis, those existing empirical models explain little physical mechanism. The objectives of this research are to improve data management of liquefaction-related case histories, and to develop the new generation of models that corroborate physical mechanism of liquefaction-induced lateral ground deformations. This goal is approached in four major steps.;The first step provides guidelines on gathering perishable information and automating Web reports of Post-Earthquake Reconnaissance (PER). Embedded metadata formats are applied to an example of the 2008 Wenchuan, China earthquake when documenting PER data. This application identifies the potential of inline metadata formats in establishing the Virtual Earthquake, a global-wide information system for archiving and sharing PER data in a spatial context.;The second step presents a new information system useful for distributing case histories of liquefaction-induced ground deformations over the Word Wide Web. Benefiting from AJAX (Asynchronous JavaScript and XML), and database technologies, this system is an interactive and lightweight application supported by external data from multiple data vendors.;The third step scrutinizes existing empirical models using a subset of case histories that are gathered and managed in the previous steps. This case study develops maps of liquefaction severity and liquefaction-induced lateral deformations from those existing models. The disagreement between the simulation results and the field observations infers a knowledge insufficiency in the regional evaluation of liquefaction-induced lateral ground deformations and therefore raises a demand for the next generation of models.;The fourth step is towards a new model that corroborates physical mechanism of liquefaction-induced deformations. Using a physical model and probabilistic analysis, PGV (Peak Ground Velocity) is demonstrated to be a significant parameter that characterizes earthquake contributions to the motions of mass that temporarily loses shear strength due to liquefaction. A PGV-dependent model is derived based on case histories to account for the importance of PGV for liquefaction-induced lateral ground deformations.