72 J.-S. Lew 6800 6850 6900 6950 7000 7050 7100 7150 7200 12 14 16 18 20 Deformation (mm) 6800 6850 6900 6950 7000 7050 7100 7150 7200 -2 -1 0 1 2 Point Error (mm) 10/24/2006 11/28/2007 a b Fig. 8.5 Results of prediction deformation of data at NPL2Y90: (a)—y, ym; (b)—ye, ( ˙2 ),—( ˙3 ) 3. The previous study indicates that the deformation increase has a time delay corresponding to the quick increase of the water level. The quick water increase happens from 6,342 to 6,344 data points (from 7/17/2005 to 7/19/2005), where water increases from 153.84 to 163.8 m, and this cause the model error ye changes from 0.3904 to 0.8848 mm, which corresponds to the variation of 2.1302 . 4. Figure 8.7 show that there are five peaks of water levels between 6,341 and 6,440 data points with quick water level increases and decreases, which is from 7/16/2005 to 10/23/2005 in the typhoon season. This causes the significant changes of model error. The model error ye has a big increase from 1.2558 to 0.1826 mm from 6,437 to 6,438 data points (10/20/2005 to 10/21/2005), which corresponds to the change of 1.7986 . But the measured water level and temperature change insignificantly from 10/20/2005 to 10/21/2005. This abnormal condition requires further investigation. Figures 8.9–8.14 show the results of the analysis of the deformations at locations NPL2Y115, NPL2Y150, and NPL2Y172. To demonstrate the results, two metrics are defined to represent the magnitude of deformation at various measured locations as y Dmaxfy.di /g minfy.di /g; y Dstd fy.di /g: (8.8) In this example, these two metrics are computed from the measured deformations from 5,801 to 6,800 data points. Table 8.1 summarizes the results for the deformation analysis at various measured.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==