Once windows were selected, Geopsy was used to generate an H/V plot for each test, considering only the selected windows. An example plot is displayed in Figure 4. The period at which the plot peaks was taken to be the fundamental period for that site. In some test, peaks were less obvious, or there were multiple possibilities. In these situations, the peak was chosen by comparing the results of the test in question to nearby tests, under the assumption that it is unlikely for the site period to change drastically over only 500 - 800 meters. Fig. 4 H/V Ratio Plot in Geopsy 3.2 Data Analysed by the FDD Technique One shortcoming in the use of Nakamura‟s method is an inability to determine if peaks in an H/V ratio are associated with horizontal or vertical motion. Assuming microtremor data is treated as output only modal data, it should be possible to apply the Frequency Domain Decomposition (FDD) technique to identify fundamental periods and their associated ground motion principal component [3]. The use of the FDD technique to identify site periods was first proposed by Ventura et. al. in 2004 [3]. Utilizing the commercial software package ARTeMIS Extractor [9], peaks of the FDD plot were chosen, as in Figure 5, and the corresponding dominant motion of each identified. For the purpose of determining site periods, peaks exhibiting horizontal motion were assumed to be representative of the site, while those with vertical motion were discarded. FDD was used heavily in tests performed on the North Shore as a way of confirming the results achieved through Nakamura‟s method. In most cases, these two methods agreed upon the dominant period for a site. Testing in Richmond relied almost exclusively upon Nakamura‟s method to identify the site period as FDD analysis of these tests provided inconsistent results. 455
RkJQdWJsaXNoZXIy MTMzNzEzMQ==