• 沒有找到結果。

Chapter 5 Result

5.3 Clustering result for each k

The statistic results indicate that Taiwan can be separated into four domains.

However, according to the geological survey and geodetic measurements, we can confirm that the tectonic settings of Taiwan are more complicated than four blocks. Therefore, I tried different values as k, which is given by the user-defined. Figure 29 to Figure 58 demonstrates the clustering result for k = 4 to 34, each figure will use red dot represent as the new group appears when the given number, k, which represents the number of the groups. Table 2 is the clustering result for the whole of Taiwan and each color corresponds to a different cluster for each k (number of clusters).

For k = 5, there is a new cluster appears at the front of the Western Foothills where are the collision edge of the Philippine sea plate and the Eurasian continental shelf (color 5 / Figure 29). For k = 6, a new cluster is introduced in the west of Longitudinal Valley Fault (color 6 / Figure 30).

For k = 7, a new cluster appears near the Chishan fault and the Lungchuan fault (color 7 / Figure 31). For k = 8, a new cluster is introduced in the west of Pingtung Plain and separated with other clusters by Fengshan fault (color 8 / Figure 32). For k = 9, a new cluster is introduced in the Ilan Plain where is located at northeastern Taiwan (color 9 / Figure 33). For k = 10, the cluster which color is orange (color B) is divided into cyan (color 10) and orange (color B) clusters along the Shuangtung fault and the Chukou fault

(Figure 34). For k = 11, the Eurasian continental shelf (color A) is be separated into Taipei Basin (color A) and the coastal plain (color 11) along the Yangmei structure (Figure 35).

For k = 12, a new cluster is introduced in the east of Hengchun fault (color 12 / Figure 36). For k = 13, a new cluster is appeared in the north of Chishan fault and the east of Hsiaokangshan fault (color 13 / Figure 37). For k = 14, the Western Foothills (color 11) is divided into two parallel structures are respectively the new cluster (color 14) and the Western Foothills (color 11), north to the Yangmei structure and south to the Tachia fault (Figure 38). For k = 15, there is a new cluster appeared in the east of Houchiali fault and the west of Loungchuan fault (color 15 / Figure 39). For k = 16, Lutao and Lanyu are be separated from the Taiwan island and are be assigned as a new color (color 16 / Figure 40). For k = 17, there is a new parallel structure appear at the Western Foothills and be marked as a new cluster (color 17 / Figure 41). For k = 18, a new cluster is introduced near the Chungchou fault (color 18 / Figure 42). For k = 19, Longitudinal Valley fault is divided into two clusters along the Chihshang fault, the north one is the new cluster (color 19 / Figure 43) and the south one is the old cluster which we got when k = 6.

For k = 20, a new cluster is introduced in the front of the Western Foothills which is between Tamaopu-Shuangtung fault and Chelungpu fault (color 20 / Figure 44). For k = 21, there are two stations been separated from surrounding stations at Ilan plain (color 21 / Figure 45). For k = 22, there is a new cluster is introduced near the central of Central

Range (color 22 / Figure 46). For k = 23, a new cluster is appeared in east of Hengchun fault and the south of Central Range (color 23 / Figure 47). For k = 24, several stations are been separated from the Costal range and form a new cluster (color 24 / Figure 48).

For k = 25, there is appear from the Chukou 8fault to Muchiliao fault and be marked as a new cluster (color 25 / Figure 49). For k = 26, a new cluster is introduced in the Hsinchu area and part of this cluster are located in the Taipei basin (color 26 / Figure 50). For k = 27, there is a new cluster is introduced near the Yuli fault (color 27 / Figure 51).

For k = 28, part of stations is been separated from the Taipei basin and group as a new cluster (color 28 / Figure 52). For k = 29, a new cluster is introduced in the east of Costal Range (color 29 / Figure 53). For k = 30, there is a new cluster appear near Liuchia fault and Hsinhua fault (color 30 / Figure 54). For k = 31, a new cluster is introduced in the Youchang fault (color 31 / Figure 55). For k = 32, there are some stations been separated from the central of Western Foothills (color 10) and group as a new cluster (color 32 / Figure 56). For k = 33, a new cluster is introduced in the Chungchou structure (color 33 / Figure 57). For k = 34, part of stations is been separated from Taipei basin again and group as a new cluster (color 34 / Figure 58).

For k > 34, not only the clusters seem too scatter and discontinuous in the geographic space but the number of stations of each cluster is too few to be representative. That is the principle when cluster analysis stopped at k = 34. This study can identify some

significant clusters, which can be correlated with the known active structures, but some of them do not match very well.

Table 2 RGB value of each cluster.

k RGB Active Structure k RGB Active Structure

14 (153, 0, 51)

Miaoli frontal structure to Surrounding of Dajia fault

31 (194, 194, 0) Youchang fault

15 (219, 179, 255)

East of Houchiali fault, West of Chungchou fault

32 (77, 77, 255)

Part of Western Foothills 16 (49, 73, 60) Lutao、Lanyu 33 (176, 92, 255) Chungchou structure

17 (255, 255, 255)

East of Changhua fault, West of Chelungpu fault

34 (150, 202, 202) Part of Taipei basin

Figure 29 Clustering result for Taiwan in k = 5.

Figure 30 Clustering result for Taiwan in k = 6.

Figure 31 Clustering result for Taiwan in k = 7.

Figure 32 Clustering result for Taiwan in k = 8.

Figure 33 Clustering result for Taiwan in k = 9.

Figure 34 Clustering result for Taiwan in k = 10.

Figure 35 Clustering result for Taiwan in k = 11.

Figure 36 Clustering result for Taiwan in k = 12.

Figure 37 Clustering result for Taiwan in k = 13.

Figure 38 Clustering result for Taiwan in k = 14.

Figure 39 Clustering result for Taiwan in k = 15.

Figure 40 Clustering result for Taiwan in k = 16.

Figure 41 Clustering result for Taiwan in k = 17.

Figure 42 Clustering result for Taiwan in k = 18.

Figure 43 Clustering result for Taiwan in k = 19.

Figure 44 Clustering result for Taiwan in k = 20.

Figure 45 Clustering result for Taiwan in k = 21.

Figure 46 Clustering result for Taiwan in k = 22.

Figure 47 Clustering result for Taiwan in k = 23.

Figure 48 Clustering result for Taiwan in k = 24.

Figure 49 Clustering result for Taiwan in k = 25.

Figure 50 Clustering result for Taiwan in k = 26.

Figure 51 Clustering result for Taiwan in k = 27.

Figure 52 Clustering result for Taiwan in k = 28.

Figure 53 Clustering result for Taiwan in k = 29.

Figure 54 Clustering result for Taiwan in k = 30.

Figure 55 Clustering result for Taiwan in k = 31.

Figure 56 Clustering result for Taiwan in k = 32.

Figure 57 Clustering result for Taiwan in k = 33.

Figure 58 Clustering result for Taiwan in k = 34.

相關文件